Ministry of Higher Education and Scientific Research University of Sulaimani Faculty of Engineering Irrigation Engineering Department

WATER QUALITY ASSESSMENT FOR DOKAN LAKE USING LANDSAT 8 OLI SATELLITE IMAGES

A Thesis Submitted to the Faculty of Engineering of the University of Sulaimani in Partial Fulfilment of the Requirements for the Degree of Master in Science of Water Resources Engineering

By Hasti Shwan Abdullah BSc. Irrigation Engineering - 2010

October 2015 AD

Gelarezan 2715 KR

Moharam 1437 AH

Ministry of Higher Education and Scientific Research University of Sulaimani Faculty of Engineering Irrigation Engineering Department

WATER QUALITY ASSESSMENT FOR DOKAN LAKE USING LANDSAT 8 OLI SATELLITE IMAGES

A Thesis Submitted to the Faculty of Engineering of the University of Sulaimani in Partial Fulfilment of the Requirements for the Degree of Master in Science of Water Resources Engineering By Hasti Shwan Abdullah BSc. Irrigation Engineering - 2010

Supervised By Assist. Prof. Dr. Mahmoud S. Mahdi

October 2015 AD

Gelarezan 2715 KR

Dr. Hekmat M. Ibrahim

Moharam 1437 AH

Linguistic Certification This is to certify that I, Dr. Shokhan Rasool Ahmed, have proofread this thesis entitled “Water Quality Assessment for Dokan Lake Using Landsat 8 OLI Satellite Images” prepared by (Hasti Shwan Abdullah). After making and correcting the mistakes, the thesis was handed again to the researcher to make the correction in this last copy.

Signature: Proofreader: Dr. Shokhan Rasool Ahmed Position:

English Department, School of Language, University of Sulaimani

Date:

15 / 11 / 2015

Supervisors Certification We certify that the preparation of this thesis entitled “Water Quality Assessment for Dokan Lake Using Landsat 8 OLI Satellite Images” was presented by (Hasti Shwan Abdullah) under our supervision at the Irrigation Engineering Department of the Faculty of Engineering in the University of Sulaimani, as partial fulfilment of the requirements for the degree of Master in Science of Water Resources Engineering.

Signature: Co-Supervisor: Lecturer Dr. Hekmat M. Ibrahim Date:

/

/ 2015

Signature: Supervisor: Assist. Prof. Dr. Mahmoud S. Mahdi Date:

/

/ 2015

In view of the available recommendation, we forward this thesis for debate by the Examining Committee.

Signature: Name: Assist. Prof. Dr. Azad Abdulkadir Mohammed Head of Postgraduate Studies Committee Date:

/

/ 2015

Examining Committee Certification We certify that we have read this thesis entitled “Water Quality Assessment for Dokan Lake Using Landsat 8 OLI Satellite Images” and as the Examining Committee, examined the student (Hasti Shwan Abdullah) in its content and in what is connected with it and that in our opinion; it meets the standard of a thesis for the degree of Master in Science of Water Resources Engineering.

Signature:

Signature:

Name: Dr. Khalid Ibrahim Hassoon

Name: Dr. Ako Rashed Hama

Date:

Date:

/

/ 2016

/

/ 2016

Member

Member

Signature: Name: Prof. Dr. Abdul Razzak T. Zaboon Date:

/

/ 2016 Chairman

Signature:

Signature:

Name: Lecturer Dr. Hekmat M. Ibrahim

Name: Assist. Prof. Dr. Mahmood S. Mahdi

Date:

Date:

/

/ 2016

/

/ 2016 Supervisor

Co-Supervisor

Approved by the dean of the Faculty of Engineering of the University of Sulaimani. Signature: Name: Assist. Prof. Dr. Asso Raouf Majeed Dean of Faculty of Engineering Date:

/

/ 2016

‫ِب ْس ِم ل ّ‬ ‫الر ِح ِيم‬ ‫الر ْح لم ِن َّ‬ ‫َلاِ َّ‬

‫ي﴾‬ ‫اء ُك َّل ل‬ ‫﴿ لو لجعل ْلنلا ِم لن ْال لم ِ‬ ‫ش ْيء لح ّ‬ ‫)األنبياء‪(٣٠‬‬

“In memory of Dr. Salah Abdi Al-Hameed Saleh who left fingerprints of grace on our lives. Never be forgotten”

Dedication My father, who encouraged me to be the best I can be, to have high expectations and to fight hard for what I believe. He always provided me with best opportunities in life. Thanks for his supporting and guidance. My Mother, who has been a great tower of strength during my life. Her practical and prayerful support, her confidence in me made me realize that I could not have done it without her. She shares in my success and I will be forever grateful for her loving support. This thesis is as much hers as mine. My brothers and sisters, my heroes, have always been an inspiration and “the wind beneath my wings”. My love, whose love, understanding and encouragement have helped me struggle through the most difficult times. Her presence and unwavering support have been a continuous source of my motivation and inspiration throughout this study.

Hasti Shwan Abdullah, October, 2015

Acknowledgements

Acknowledgements Praise be to God, the creator who grant me the power of thinking and working. Firstly, I would like to express my sincere gratitude to my advisor Prof Dr. Salah Abdi Al-Hameed (RIP) for his support, patience, motivation, and immense knowledge in writing this thesis. Dr. Salah helped me in the very beginning of this study through his guidance. I could not have imagined that he would leave us so early. Then, I would like to express my sincere gratitude to my supervisors Asst. Prof. Dr. Mahmoud S. Mahdi and Dr. Hekmat M. Ibrahim for their guidance, support and help through writing this study. Sincere appreciation to Prof. Dr. Salahaddin Saeed Ali, The president of University of Sulaimani for his viable suggestions and support through my study; Prof. Dr. Ayad M. Fadhil, a lecturer in College of Agricultural, Salahaddin University, for his advices; Mr. Karwan Qaradaghi, a lecturer at Faculty of Engineering in University of Sulaimani; and Mr. Adil Othman, a lecturer at Faculty of Science in University of Sulaimani. Many thanks are also extended to Mr. Omed Ismael, the lecturer in Faculty of Science in University of Sulaimani, Sulaimani Environmental Department and Erbil Environmental Department, and especially Mr. Sarkar Syamend who helped me a lot, even in his weekend tests some samples for me. Thanks also to Mr. Hardy Abdulqader, Mr. Huner Ahmed and Mr. Rozhgar Ahmed for their helps and assistance in my sampling days. Thanks and gratitude to Dr. Ako H. Rasheed, the head of Irrigation Department for his valuable notice and discussions during my study in my master degree. Thanks are also maintained to lecturer Mr. Farouk Abdullah and staff of Irrigation Engineering Department in Faculty of Engineering / University of Sulaimani for their favor. Finally, I would like to thank my family, my parents and siblings for supporting me spiritually throughout my life.

[I]

Abstract

Abstract It is impractical to monitor water quality more than a small fraction of lakes by conventional field methods because of expense and time requirements. High resolution satellite image is more convenient to be applied to collect the required data for monitoring and assessing water quality in the lakes. Therefore, this study aims to estimate the water quality indices and concentration of some parameters (Temperature, DO, BOD, pH, Turbidity, TSS, TDS, EC, NO3, PO4 and E. coli) through applying developed water quality estimation models based on the remote sensing and GIS techniques on the Landsat 8 OLI satellite image using twenty points in Dokan Lake, Kurdistan Region, Iraq at two different seasons. Four standard mathematical methods (NSFWQI, CCMEWQI, OWQI and AWWQI) are used to find the water quality indices at the twenty stations in Dokan Lake. Results of NSFWQI method are found as medium class for all stations except station 11, 12 and 13 which are classified as bad class for Spring season. The second method (CCMEWQI), all stations are classified as good for Autumn Season except station 1 and 2 as marginal and 16 as fair. In third method (OWQI) are classified as good for all stations except station 1 and 8 which classified as fair. However, it has shown very poor for Spring season. Finally, in the fourth method (AWWQI), all stations are classified as poor except station 15 and 19 which classified as very poor while station 16 is classified as unsuitable for drinking. The Secchi Disk Transparency (SDT) and the Trophic State Index (TSI) was come up with high variance for all stations. Multiple linear regression is used to obtain mathematical models for estimating the water quality indices and concentration of some parameters depending on spectral reflectance of Landsat 8 OLI. In this study, new band (coastal blue) of Landsat 8 OLI has been undertaken in developing of models. Moreover, new Independent Component Analysis (ICA) and new 7 band ratios with 16 band combinations have been used. The best model is the AWWQI which has the highest coefficient of determination (R 2) of 0.993

[ II ]

Abstract

for Autumn season and slightly low (0.612) for Spring season. The highest determination coefficient for SDT and TSI is 0.982 and 0.873 for Autumn Season and (0.951, 0.973) for Spring season respectively. However, high R2 of 0.982, 0.982, 0.832 for TSS, Turbidity and DO are resulted respectively. Generally, for Spring season, the performance of all models is reduced due to seasonal change, variance of parameters and other factors. However, high R2 of 0.862 has been shown for Temperature. Once the developed models applied in order to have maps with a variation of colors. This facilitates to predict how the results of WQPs can be distributed within the lake and all the results are reasonable. The conclusions present that correlation of all bands of Landsat 8 OLI is appropriate to water quality indices and parameters. This study suggests further researching about how remote sensing of water quality index and parameters at different depths in Dokan Lake can be detected.

[ III ]

Table of Contents

List of Contents Title

Page

Acknowledgements ...............................................................................................I Abstract ............................................................................................................... II List of Contents.................................................................................................. IV List of Figures .................................................................................................. VII List of Tables....................................................................................................... X List of Abbreviations ....................................................................................... XII List of Notations .............................................................................................. XV Chapter One......................................................................................................... 1 Introduction ....................................................................................................... 1 1.1 General ........................................................................................................ 1 1.2 The Problem Statement ............................................................................... 3 1.3 Aim of Study ............................................................................................... 4 1.4 Methodology ............................................................................................... 4 1.5 Thesis Layout .............................................................................................. 7 Chapter Two ........................................................................................................ 8 Literature Review .............................................................................................. 8 2.1 Introduction ................................................................................................. 8 2.2 Previous Studies .......................................................................................... 9 2.3 Summary ................................................................................................... 23 Chapter Three ................................................................................................... 25 Theoretical Background .................................................................................. 25 3.1 Introduction ............................................................................................... 25 3.2 Water Quality ............................................................................................ 25 3.2.1 When to Sample ................................................................................. 26 3.2.2 Selecting Sampling Locations ............................................................ 26 3.2.3 Selection Water Quality Parameters .................................................. 27 3.2.4 Water Quality Index ........................................................................... 34 3.2.5 Determining a Lake's Trophic State (TSI) ......................................... 39 [ IV ]

Table of Contents

3.3 Remote Sensing ......................................................................................... 42 3.3.1 Brief History of Remote Sensing ....................................................... 43 3.3.2 Concepts of Remote Sensing System................................................. 45 3.3.2.1 Energy Sources and Radiation Principle..................................... 45 3.3.2.2 Energy Interactions in the Atmosphere....................................... 45 3.3.2.3 Energy Interaction with Earth Surface Features ......................... 47 3.3.2.4 Data Acquisition and Resolutions ............................................... 48 3.3.2.5 Reference Data and the Global Positioning System (GPS) ........ 49 3.3.2.6 Satellite Date Imagery System .................................................... 50 3.3.2.7 Data Image Process and Analysis ............................................... 51 3.4 Statistical Process (Correlation and Regression) ...................................... 58 3.4.1 Covariance and Correlation................................................................ 59 3.4.2 Regression Models ............................................................................. 60 3.4.3 Selecting Variables Technique........................................................... 62 3.4.4 Criteria for Choice of Subset Size in Regression............................... 63 Chapter Four ..................................................................................................... 65 Experiment Works and Processes ................................................................... 65 4.1 Introduction ............................................................................................... 65 4.2 The Study Area.......................................................................................... 65 4.3 Collecting and Analyzing in Situ Data ..................................................... 69 4.3.1 Autumn Season Sampling Conditions ............................................... 69 4.3.2 Spring Season Sampling Conditions .................................................. 73 4.4 Water Quality Tests and Laboratory Analyses ......................................... 73 4.5 Water Quality Index Calculation .............................................................. 79 4.5.1 NSF WQI Calculation ........................................................................ 79 4.5.2 Arithmetic Weight WQI Calculation ................................................. 85 4.5.3 CCMEWQI Calculation ..................................................................... 88 4.5.4 Oregon Water Quality Index .............................................................. 89 4.6 Water Clarity and TSI Calculation............................................................ 92 4.7 Remote Sensing Processing ...................................................................... 95 [V]

Table of Contents

4.7.1 Pre-Processing (Atmospheric Correction) ......................................... 95 4.7.2 Image Subset and Classification ........................................................ 97 4.7.3 Image Transformation ........................................................................ 98 4.7.4 Band Ratios, Combination and Spectral Indices ................................ 99 4.8 Correlation and Regression Models ........................................................ 100 Chapter Five .................................................................................................... 102 Result and Discussion ................................................................................... 102 5.1 Introduction ............................................................................................. 102 5.2 Estimating Water Quality Parameters ..................................................... 102 5.2.1 pH, Acidity and Alkalinity ............................................................... 107 5.2.2 Electrical Conductivity (EC) ............................................................ 107 5.2.3 Total Dissolved Solids (TDS) .......................................................... 108 5.2.4 Temperature ..................................................................................... 112 5.2.5 Nitrate ............................................................................................... 112 5.2.6 Phosphate ......................................................................................... 117 5.2.7 Dissolved Oxygen and Biochemical Oxygen Demand .................... 117 5.2.8 Total Suspended Solid and Turbidity............................................... 121 5.3 Estimating Water Quality Index ............................................................. 124 5.3.1 Autumn Season Models ................................................................... 124 5.3.2 Spring Season Models ...................................................................... 126 5.4 Water Clarity Estimating Models ........................................................... 127 5.5 Generality of Models............................................................................... 130 Chapter Six ...................................................................................................... 133 Conclusions and Recommendations ............................................................. 133 6.1 Conclusions ............................................................................................. 133 6.2 Recommendations ................................................................................... 134 References ........................................................................................................ 135 Appendix A ....................................................................................................... A1

[ VI ]

List of Figures

List of Figures Figure

Title

Page

Fig. (1-1): Flowchart of carrying out the methodology of the present study........ 6 Fig. (3-1): Two oscillating components of EM radiation: an electric and a magnetic field (Tempfli, et al., 2009). ........................................... 46 Fig. (3-2): The diagram shows the wavelength and frequency ranges of EM radiation (Tempfli, et al., 2009). .................................................... 46 Fig. (3-3): Interaction of Energy with the earth’s surface (Lillesand, et al., 2007). .............................................................................................. 48 Fig. (4-1): Map of Northern Iraq showing the study area cited in (UNESCWA & BGR , 2013). ................................................................ 66 Fig. (4-2): Landsat 8 image represent the study area of Dokan Lake on October 24th, 2014 (left) and April 2nd, 2015 (right). ..................... 66 Fig. (4-3): Water sampling from the lake and type of bottles used for samples. ........................................................................................... 70 Fig. (4-4): Station points of sampling on October 24th, 2014 in Dokan Lake. ... 71 Fig. (4-5): Station points of sampling on April 2nd, 2015 in Dokan Lake. ......... 74 Fig. (4-6): Measurement of Secchi disk in Dokan Lake. .................................... 92 Fig. (4-7): Landsat 8 OLI image full scene RGB true captured on October 24th, 2014 (left) and April 2nd, 2015 (right). ................................... 95 Fig. (4-8): Landsat 8 OLI image on October 24th, 2014, atmospheic and radiometric correction, (left) before pre-procccesing, (right) after pre-processing......................................................................... 96 Fig. (4-9): Landsat 8 OLI image on April 2nd, 2015, atmospheic and radiometric correction, (left) before pre-procccesing, (right) after pre-processing......................................................................... 97 Fig. (4-10): Landsat 8 OLI image on October 24th, 2014, after subset (left), unsupervised classification (middle), after extraction of wetted area (right). ...................................................................................... 98

[ VII ]

List of Figures

Fig. (4-11 ): Landsat 8 OLI image on April 2nd, 2015, after subset (left), unsupervised classification (middle), after extraction of wetted area (right). ...................................................................................... 98 Fig. (5-1): Surface reflectance of Landsat 8 OLI bands for selected stations in Dokan Lake on October 24th, 2014. .......................................... 104 Fig. (5-2): Surface reflectance of Landsat 8 OLI bands for selected station in Dokan Lake on April 2nd, 2015................................................. 106 Fig. (5-3): Measured and computed pH values in Dokan Lake for Autumn and Spring Seasons. ...................................................................... 109 Fig. (5-4): Computed pH values in Dokan Lake for Autumn Season (left) and Spring Season (right). ............................................................ 109 Fig. (5-5): Measured and computed EC values in Dokan Lake for Autumn and Spring Seasons. ...................................................................... 110 Fig. (5-6): Computed EC values in Dokan Lake for Autumn Season (left) and Spring Season (right). ............................................................ 110 Fig. (5-7): Measured and computed TDS values in Dokan Lake for Autumn Season. .......................................................................................... 111 Fig. (5-8): Computed TDS values in Dokan Lake for Autumn Season on October 24th, 2014......................................................................... 111 Fig. (5-9): Measured and computed Temperature in Dokan Lake for Autumn and Spring Seasons. ........................................................ 114 Fig. (5-10): Computed Temperature in Dokan Lake for Autumn Season (left) and Spring Season (right). ................................................... 114 Fig. (5-11): Measured and computed values of NO3 for Autumn and Spring Seasons. ......................................................................................... 115 Fig. (5-12): Measured and computed values of NO3 − N in Dokan Lake for Autumn and Spring Seasons. ........................................................ 115 Fig. (5-13): Computed NO3 values in Dokan Lake for Autumn Season (left) and Spring Season (right). ................................................... 116

[ VIII ]

List of Figures

Fig. (5-14): Computed NO3 − N values in Dokan Lake for Autumn Season (left) and Spring Season (right). ................................................... 116 Fig. (5-15): Measured and computed values of PO4 in Dokan Lake for Autumn and Spring Seasons. ........................................................ 118 Fig. (5-16): Measured and computed values of TP in Dokan Lake for Autumn and Spring Seasons. ........................................................ 119 Fig. (5-17): Computed PO4 values in Dokan Lake for Autumn Season (left) and Spring Season (right). ............................................................ 119 Fig. (5-18): Computed TP values in Dokan Lake for Autumn Season (left) and Spring Season (right). ............................................................ 120 Fig. (5-19): Measured and computed values of DO and BOD in Dokan Lake for Autumn and Spring Seasons respectively. .............................. 120 Fig. (5-20): Computed DO values for Autumn Season (left) and BOD values for Spring Season (right) in Dokan Lake. ..................................... 121 Fig. (5-21): Measured and computed values of TSS in Dokan Lake for Autumn and Spring Seasons. ........................................................ 122 Fig. (5-22): Measured and computed values of Turbidity in Dokan Lake for Autumn and Spring Seasons. ........................................................ 123 Fig. (5-23): Computed TSS values in Dokan Lake for Autumn Season (left) and Spring Season (right). ............................................................ 123 Fig. (5-24): Computed Turbidity values in Dokan Lake for Autumn Season (left) and Spring Season (right). ................................................... 124 Fig. (5-25): Computed AWWQI values in Dokan Lake for Autumn Season (left) and Spring Season (right). ................................................... 126 Fig. (5-26): Computed SDT values in Dokan Lake for Autumn Season (left) and Spring Season (right). ................................................... 129 Fig. (5-27): Predicted TSI values in Dokan Lake for Autumn Season (left) and Spring Season (right). ............................................................ 130

[ IX ]

List of Tables

List of Tables Table

Title

Page

Table (3-1): Water quality rating as per different water quality index methods. .......................................................................................... 40 Table (3-2) : Comparison of trophic state index to water quality parameters and lake productivity. ..................................................................... 42 Table (3-3): Landsat sattelite mission dates with sensors specifications............ 50 Table (3-4): Operational land imager (OLI) and thermal infrared sensor (TIRS) band designations. .............................................................. 51 Table (4-1): The Climate Data recorded by observation station of Dokan dam’s Department. .......................................................................... 68 Table (4-2): The water balance and hydrological data recorded by observation station of Dokan dam’s department. ........................... 68 Table (4-3): Sampling time and station points information on October 24 th, 2014 in Dokan Lake. ....................................................................... 72 Table (4-4): Sampling time and station points information on April 2nd, 2015 in Dokan lake. ........................................................................ 75 Table (4-5): Tests of water quality parameters correspond to the stations numbers on October 24th, 2014. ...................................................... 77 Table (4-6): Tests of water quality parameters correspond to the stations numbers on April 2nd, 2015............................................................. 78 Table (4-7): Water quality parameters and their weight. .................................... 82 Table (4-8): Sensitivity of WQPs that used to calculated NSF water quality index for October 24th 2014. ........................................................... 83 Table (4-9): Sensitivity of WQPs that used to calculated NSF water quality index for April 2nd, 2015. ................................................................ 84 Table (4-10): Recommended WHO standard of the water quality parameter......................................................................................... 85 Table (4-11): Sensitivity of WQPs that used to calculated AW water quality index for October 24th, 2014. .......................................................... 86 [X]

List of Tables

Table (4-12): Sensitivity of WQPs that used to calculated AW water quality index for April 2nd, 2015. ................................................................ 87 Table (4-13): Calculation of CCMEWQI for Autumn Season on October 24th, 2014......................................................................................... 88 Table (4-14): Calculation of CCMEWQI for Spring Season on April 2nd, 2015................................................................................................. 89 Table (4-15): OWQI calculation for Autumn Season on October 24th, 2014. .... 91 Table (4-16): OWQI calculation for Spring Season on April 2nd, 2015. ............ 91 Table (4-17): TSI Calculation for samples on October 24th, 2014. .................... 93 Table (4-18): TSI calculation for samples on April 2nd, 2015. ........................... 94 Table (5-1): WQPs models for Autumn Season on October 24th, 2014. .......... 104 Table (5-2): WQP, WQI & TSI models for Spring season on April 2nd, 2015............................................................................................... 106 Table (5-3): Water quality index models for Autumn Season data on October 24th, 2014......................................................................... 125 Table (5-4): Water quality index models for Spring Season data on April 2nd, 2015. ....................................................................................... 127 Table (5-5): Water clarity models for Autumn Season on October 24th, 2014............................................................................................... 128 Table (5-6): Water clarity models for Spring Season on April 2nd, 2015. ........ 129 Table (5-7): Generality of Autumn Season models on October 24th, 2014. ..... 131 Table (5-8): Generality of Spring Season models on April 2nd, 2015. ............. 132 Table (A-1): Pearson Correlation between Water Quality (WQI, WQPs, SDT and TSI) and Spectral Bands for Autumn Season…………. A1 Table (A-2): Pearson Correlation between Water Quality (WQI, WQPs, SDT and TSI) and Spectral Bands for Spring Season……….…. A9 Table (A-3): Selection of variables for assessment of water quality in relation to non-industrial water use (Chapman, 1996)…........…. A17

[ XI ]

List of Abbreviation

List of Abbreviations Abbreviation

Description

a.m.

Ante Meridiem, "before midday".

AIC

Akaike Information Criterion.

ANN

Artificial Neural Networks.

AVHRR

AWEI

Advanced Very High Resolution Radiometer instruments are a type of space-borne sensor that measure the reflectance of the earth. Automated Water Extraction Index.

AWWQI

Arithmetic Weight Water Quality Index.

B

Blue Band of Landsat 8 OLI.

BOD

Biochemical Oxygen Demand.

C

Coastal Blue band of Landsat 8 OLI.

CCME

Canadian Council of Ministers of the Environment.

CDOM

Colored Dissolved Organic Matter.

Chl

Chlorophyll.

Chl-a

Chlorophyll-a “a specific form of chlorophyll”.

COD

Chemical Oxygen Demand.

CWQI

Canadian Water Quality Index.

CZCS

Coastal Zone Color Scanner.

DEM

Digital Elevation Model.

DNs

Digital Numbers.

DO

Dissolved Oxygen.

DS

Dissolved Solid.

DVI

Difference Vegetation Index.

EC

Electrical Conductivity.

EM

Electromagnetic Radiation.

ENVI

Environment for Visualizing Images.

ERDAS

Earth Resources Data Analysis System.

ETM

Enhanced Thematic Mapper “sensor of Landsat 7”.

ETM+

Enhanced Thematic Mapper Plus “sensor of Landsat 7”.

FOV

Field of View.

G

Green Band of Landsat 8 OLI.

GIS

Geographical Information System. [ XII ]

List of Abbreviation

GPS

Global Positioning System.

GSD

Ground Sampling Distance.

IBM ICA

International Business Machines “American multinational technology and consulting corporation”. Independent Component Analysis.

IPVI

Infrared Percentage Vegetation Index.

IR

Infrared Radiation.

IRS LISS III JTU

Indian Remote Sensing Satellite Linear Imaging Self Scanner. Jackson Turbidity Unit

LSWI

Land Surface Water Index.

MIR

Mid-Infrared.

MLR

Multiple Linear Regression.

MNDWI

Modification of Normalised Difference Water Index.

MNF

Noise Fraction Transform.

MODIS

Moderate Resolution Imaging Spectroradiometer.

MR

Multiple Regression.

MS/TM

Multispectral Scanner / Thermal Mapper of Landsat.

MSI

Moisture Stress Index.

MSS

Landsat Multispectral Scanner.

MSS RBV

Return Beam Vidicon Camera.

NASA

National Aeronautics and Space Administration.

NBR

Normalized Burn Ratio.

NDMI

Normalized Difference Moisture Index.

NDVI

Normalized Difference Vegetation Index.

NDWI

Normalized Difference Water Index.

NIR

Near Infrared Band “Sensor of Landsat 8".

NSFWQI

National Sanitation Foundation.

NTU

Nephelometric Turbidity Units.

OLI

Landsat Operational Land Imager "Sensor of Landsat 8".

p.m.

Post Meridiem, "After Midday".

PCA

Principal Component Analysis.

pH

pH of Water: p representing German Potenz ‘Power’, H, the symbol for hydrogen. Particulate Organic Carbon.

POC

[ XIII ]

List of Abbreviation

psu

Practical Salinity Scale.

R

Red Band of Landsat 8 OLI.

R2

Coefficient of determination of regression models.

Rrs

Reflectance of remote sensing.

RS

Remote Sensing.

RVI

Ratio-Vegetation-Index.

SBC

Schwarz Bayesian Criterion.

SDD

Secchi Disk Depth.

SDT

Secchi Disk Transparency.

SeaWIFS

Sea-Viewing Wide Field-Of-View Sensor, “satellite-borne sensor works on global ocean biological data”. Sediment Oxygen Demand.

SOD

SPSS

(French: Satellite Pour d’Observation de la Terre) "Satellite for Observation of Earth". Statistical Package for the Social Sciences.

SS

Suspended Solids.

SWIR1

First Short-Wave Infrared band “Landsat 8 OLI”.

SWIR2

Second Short-Wave Infrared band “Landsat 8 OLI”.

TDS

Total Dissolved Solids.

TH

Total Hardness.

TIRS

Thermal Infrared Sensor.

TM

Landsat Thematic Mapper.

TN

Total Nitrogen.

TNR

Total Nonfilter-able Residue

TP

Total Phosphorus.

TSI

Trophic State Index.

TSS

Total Suspended Solid.

UK

United Kingdom.

USGS

United States Geological Survey.

UTM

Universal Transverse Mercator “cylindrical projection”.

UV

Ultraviolet Light.

WHO

World Health Organization.

WMO

World Meteorological Organization.

WQI

Water Quality Index.

WRI

Water Ratio Index.

SPOT

[ XIV ]

List of Notations

List of Notations Notation

Description

𝑞𝑖

Represents the rating for the 𝑖 𝑡ℎ determinant.

𝑤𝑖

Represents the weighting for the 𝑖 𝑡ℎ determinant.

𝑛

Number of determinants.

𝑄𝑖

𝑊𝑖

Quality rating of 𝑖 𝑡ℎ parameter for a total of 𝑛 water quality parameters. Actual value of the water quality parameter obtained from laboratory analysis. Ideal value of that water quality parameter can be obtained from the standard tables. Recommended WHO standard of the water quality parameter. Relative unit weight for 𝑛𝑡ℎ parameter.

𝑆𝑖

Standard permissible value for nth parameter.

𝑖

Proportionality constant.

𝑄𝑖

Quality rating.

𝑊𝑖

Relative weight.

𝑛

Number of sub-indices.

𝑆𝐼

Sub index of 𝑖 𝑡ℎ parameter.

Factor 1: F1

𝑛𝑠𝑒

This factor is called scope because it assesses the extent of water quality guideline non-compliance over the time period of interest. This factor is called frequency and represents the percentage of individual tests that do not meet the objectives (failed tests). This factor is called amplitude and represents the amount by which the failed test values do not meet their objectives, and is calculated in three steps. Sum of the excursions from objectives.

ℓ𝑛 𝑆𝐷

Logarithmic of Secchi Disk depth in (m).

E 𝐼 (λ)

Incident energy.

E 𝑅 (λ)

Reflected energy.

E 𝐴 (λ)

Absorbed energy.

𝑉𝑎𝑐𝑡𝑢𝑎𝑙 𝑉𝑖𝑑𝑒𝑎𝑙 𝑉𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑

Factor 2: F2

Factor 3: F3

[ XV ]

List of Notations

E 𝑇 (λ)

Transmitted energy.

𝐿𝜆

Spectral Radiance at the sensor's aperture in watts/(meter squared * ster * µm). Rescaled gain (the data product "gain" contained in the Level 1 product header or ancillary data record) in watts/(meter squared *ster * µm). The quantized calibrated pixel value in 𝐷𝑁.

𝐺𝑎𝑖𝑛

𝑄𝐶𝑎𝑙 𝑏𝑖𝑎𝑠

𝐿𝜆

Rescaled bias (the data product "offset" contained in the Level 1 product header or ancillary data record) in watts/(meter squared * ster * µm). Radiance in units of W/(m2 * sr * µm).

𝑑

Earth-sun distance, in astronomical units.

𝐸𝑆𝑈𝑁𝜆

Solar irradiance in units of W/(m2 * µm).

𝜃

Sun elevation in degrees.

(𝑦𝑖 − 𝑦̅)

(𝑥𝑖 − 𝑥̅ )(𝑦𝑖 − 𝑦̅)

Deviation of each observation 𝑦𝑖 from the mean of the response variable. Deviation of each observation 𝑥𝑖 from the mean of the predictor variable. Product of the above two quantities.

𝑠𝑦

Standard deviation of 𝑌.

𝑍

Z variable.

𝑧𝑖

𝑖 𝑡ℎ Value of the independent variable 𝑍.

𝑠𝑥

Standard deviation of 𝑋.

𝑟(𝑥, 𝑦) = 𝑐𝑜𝑟(𝑥, 𝑦)

Correlation coefficient between 𝑌 and 𝑋.

𝑟𝑃 𝑋̅

Pearson product-moment correlation coefficient.

𝑌̅

Sample mean of 𝑌1 ,𝑌2 , . . . ,𝑌𝑛 .

𝑦𝑖

Dependent variable.

𝛽0 , 𝛽1 , 𝛽2 , … , 𝛽𝑛

Regression coefficients.

𝑥1 , 𝑥2 , … , 𝑥𝑛

Independent variables in the models.

𝑅2

Coefficient of determination.

𝑥𝑖

𝑖 𝑡ℎ value of the independent variable 𝑥.

𝑥̅

Sample mean value of the dependent variable 𝑥.

𝑦𝑖

𝑖 𝑡ℎ value of the dependent variable 𝑦.

(𝑥𝑖 − 𝑥̅ )

Sample mean of 𝑋1 ,𝑋2 , . . . , 𝑋𝑛 .

[ XVI ]

List of Notations

𝑦̅

Sample mean value of the dependent variable 𝑦.

𝑛 − 𝑝′

Residual.

𝜌𝐻

Equivalent horizontal surface.

𝜌𝑇

Inclined surface.

NH4

Chemical formula of Ammonia.

NO3

Chemical formula of Nitrate.

NO3-N

Chemical formula of “Nitrate as Nitrogen”.

PO4

Chemical formula of phosphate.

CaCO3

Chemical formula of calcium carbonate.

T

Temperature.

[ XVII ]

CHAPTER ONE INTRODUCTION

Chapter One

Introduction

Chapter One Introduction 1.1 General Water quality monitoring is the systematic collection and evaluation of data about the chemical, physical, and biological quality of the water bodies, and assesses how external changes, both natural and anthropogenic, affect that quality. The monitoring strategy identifies various assessment needs and forms a basis for setting resource priorities to ensure their best use in achieving strategic water quality management goals. Monitoring and assessment is a fundamental need of the Water Quality Program and an integral component of protecting human health and the environment. A well designed monitoring and assessment program defines water quality problems, characterizes existing and emerging problems, determines the magnitude and geographical extent of water conditions, provides the basis for designing and operating pollution prevention and abatement programs, evaluates the effectiveness and compliance of water quality programs, and identifies trends in water quality over time. As the communities grow, the water quality will be more significant to lives of humankind, whether it is used for drinking, irrigation, or recreational purposes. Water quality will be a majority of impacts on health to the outbreaks of waterborne disease and by contributing to the background rates of disease. Accordingly, countries developed water quality standards to protect public health as guidelines to present an authoritative assessment of the health risks associated with exposure to health hazards through water and of the effectiveness of approaches to their control. A water quality impairment which adversely affects the lake water to an extent which inhibits or prevents some beneficial water use. Since a water quality issue normally results from the deleterious effects of one or more human uses, major conflicts between users or uses may occur in lake systems subjected to [1]

Chapter One

Introduction

multiple use. These kinds of water quality conflicts and associated human interactions bring considerable complexity to lake management. This often results in continuous arbitration between user-groups, and failure to take the necessary control actions to restore or maintain lake water quality. Since the problems of water quality relate to water use, water quality assessment strategies should be depending on adequate knowledge of lake physics and structure, lake uses and associated water quality requirements, and the legislative powers or authorities which may be used to enforce and ensure compliance with standards. Monitoring data and the associated assessments are thus the basis upon which sound lake management should be conducted. The issues affect lake water quality have mostly identified and described in industrialized regions, which have progressed from one issue to another in a sequence parallel to social and industrial development. This means that, in the developing world, multi-issue water quality problems must be faced with greater cost and complexity in assessment design, implementation and data interpretation. One of the current issues that facing lake water quality is Eutrophication, which is the biological response to excess nutrient inputs to a lake. Numerous indices have been developed to measure the degree of eutrophication of water bodies. Many have been based on phytoplankton species composition, but are not recommended since they are complicated to undertake, difficult to interpret and are affected by local conditions. The most reliable methods are based on the classifications and suitable alternatives for biological indicators are total particulate organic carbon (POC) and chlorophyll-a, since they represent total biomass To get a true picture about the nature of the river and lake water, it is necessarily required to measure the quantity and quality of water through water quality monitoring, which implemented in many methods and techniques. The traditional method of water quality monitoring, is collecting and analysing water samples after testing them in laboratory, but the method requires more times and it costs. Recently, with advance and increasing role of technology, new techniques [2]

Chapter One

Introduction

and methods are developed for assessing water quality such as Remote Sensing (RS) and Geographical Information System (GIS) that overcomes through using satellite data to monitor water quality to reduce time and cost for the process, and to increase accuracy of results. Remote sensing and GIS have a potential to monitor spatial variation in water quality over large areas. Remote sensing of lake water is often limited to high spatial resolution satellites such as Landsat, which have limited spectral resolution. This thesis presents the results of an investigation into satellite monitoring of lake water quality. The aim of this investigation is to ascertain the feasibility of estimating water quality and its spatial distribution using Landsat 8 OLI imagery combined with in situ data from Dokan lakes. 1.2 The Problem Statement As developing of countries, the dams have been constructed on rivers for storage of waters in lakes and reservoirs to prevent the risk of flooding and to be used for multipurpose such as (drinking, irrigation, industries, generating powers, etc.). In Iraq, several dams were built on the Euphrates and Tigris rivers and their tributaries, and Dokan Dam is one these dams, which was built in 1959. Dokan Dam is one of the main sources of water for drinking, irrigation and power generation in the province of Kurdistan Region and especially to the province of Sulaimani. The water storage process in the lakes behind the dams as well as the formation of large bodies of water often leads to significant changes to the quality of water. The change of river water properties is due to the hydrological properties changes of the lake or as well as the presence of pollutants that reach the lake one way or another. Therefore, it is necessary to monitor and follow up the quality of water stored in lakes or reservoirs behind the dams. It is important to investigate the water quality index of Dokan Lake and implement and apply a remote sensing model to monitor and evaluate the spatial distribution of the water quality index within the lake.

[3]

Chapter One

Introduction

1.3 Aim of Study Dokan Lake is the primary source of water for irrigation, power generation and drinking water of Sulaimani city and it is necessary to study, monitor and evaluate the quality of water in the lake and its suitability for different uses. Therefore, the aim of this study is to measure some Water Quality Parameters (WQPs) such as (pH, EC, TDS, Temperature, turbidity, 𝑁𝑂3− , 𝑃𝑂4 , DO and BOD) in order to determine the Water Quality Index (WQI) and Secchi Disk to develop a Trophic State Index (TSI) for two seasons (Autumn and Spring) by using mathematical relationship such as regression models between water quality parameters as independent variables and spectral bands of Landsat 8 OLI as dependent variables. Finally, the spatial distribution of the WQPs, (WQI), and (TSI) for Dokan Lake have been estimated using Remote Sensing and Geographic Information Systems (GIS) techniques. 1.4 Methodology Generally, the methodology of carrying out this study can be divided into two parts: 1. Theoretical approach: it concerns the national and international historical background researches that used different techniques to find and evaluate the water quality indices (WQIs) for local and international rivers and lakes and that related to the use of remote sensing and GIS techniques in this issue. In addition to all the related water quality index, remote sensing, GIS, statistical evaluation and regression analysis theories and techniques can be adopted and applied for achieving the aims of this study. 2. Experimental Approach: it concerns the field work, laboratory test and data analysis which can be divided as follow: i.

Field work: twenty water samples were collected at specific dates and the location of these samples were specified. The tools and devices that used for carrying out this works are; (GPS Receiver GARMIN 62S, Secchi disk, Senz pH meter and laboratory thermometer). [4]

Chapter One

ii.

Introduction

Laboratory test: these tests were conducted in laboratory of (Public Environmental Departments in both Erbil and Sulaimani city) and (University of Sulaimani - Faculty of Science) or using the 3540 Bench Combined Conductivity/pH Meter to measure (pH, EC, TDS and Temperature) and (Loviband Model: TB 300 IR) to measure turbidity as NTU and finally for the (𝑁𝑂3− ) is measured by (APEL PD-303UV), (Hanna PO4 meter) for ( 𝑃𝑂4 ) and chemical process such as (Winkler Method) to tests (DO and BOD).

iii.

Data analysis: analysis of field and laboratory measurement, estimation of WQI, image processing and GIS, statistical evaluation and regression analysis and applying the implemented model were achieved using computer software, Exelis ENVI 5.3, ESRI ArcGIS 10.3, Erdas Imagine 2014 and IBM SPSS 22. Fig. (1-1) shows the flow chart of carrying out the present study. Briefly, the

steps can be summarized as the follow: 1. Selection the study area (Dokan Lake). 2. Selection of satellite (LANDSAT 8 OLI)’s image at specified dates. 3. Selection of water quality parameters (12 WQPs) to determine WQI and Secchi Disk Transparency (SDT) for estimation of Trophic State Index (TSI). 4. Selection of sampling date when it should be at the same acquisition date of the selected satellite Image. 5. The satellite image should undergo some pre-processing such as atmospheric corrections, cloud and haze removals, Principal Component Analysis (PCA) and Independent Component Analysis (ICA) Transformations …etc. 6. Developing several regression models to estimate WQI (AWWQI, CWQI, NSFWQI and OWQI) and TSI and other WQPs depending on the satellite image data using SPSS software and selection of the suitable models according to the results of statistical evaluation of these models. 7. The selected best models have been applied through using the ENVI software to get the values of parameters and indices. Finally, the ArcGIS software is used to map the spatial distribution of the water quality parameters and indices. [5]

Chapter One

Introduction

Literature Review

Experimental Works

Methodology

Selection of Date

Selection of Satellite Image

Using ENVI Software

Selection of Date Correspond to Acquisition Date

Convert ROA to Radiance

Selection of WQPs, WQI

Sampling Date

Convert Radiance to Surface Reflectance

Transform Reflectance to PCA, ICA, EM, MNF

SDT Measurement

WQPs Measurement

TSI Calculation

WQI Calculation

Band Ratios, Combinations

Using ArcGIS Software

Use GPS Coordinate to Extract Image Value for Each Station Point

Results

Using SPSS Software

Correlation Between Spectral Bands and ( WQI, WQPs, TSI)

Using ENVI Software

Regression Models (WQPs, WQI, TSI)

Appling Models to Setllite Image Choosing Models with Highest Coefficient of Determination Using ArcGIS Software

Mapping and Presentation of Models

Fig. (1-1): Flowchart of carrying out the methodology of the present study. [6]

Chapter One

Introduction

1.5 Thesis Layout The thesis is divided into six chapters. The first chapter is the introduction. While the second chapter is literature review, provides the basic introduction for previous studies in general. It includes main studies that concern the thesis subject and differences between these studies and the aim of the present study. Finally, a summary of scientific findings of some earlier researches and what has to be studied in the present study and fill the existing gap in these studies. The three main theories that are used in this study to achieve its objectives are summarized in chapter three. The first theory includes water quality principles, water quality monitoring, sampling and evaluation, selection of parameters and calculation of water quality index. The next theory includes remote-sensing principles, knowledge and theory concerns satellite data, pre-processing, indices and model maker tools. The statistical analysis with its algorithms and theoretical bases related to correlations and regression models for WQP, WQI, TSI and spectral bands are existed as a last theory. Subsequently, the first part of chapter four, experimental work, describes the study area and characteristics and process of the available satellite image data. Afterwards, experimental work which includes all the performed field works (determination of sampling station, and sampling), tools, apparatus and equipment that used to perform the field and laboratory tests and analysis are explained. In chapter five, all the satellite image data and statistical analyses with discussion of the results are included. The Satellite image data includes preprocessing (correction, layer stacks, the layer subset, converting digital numbers to reflectance and PCA). On the other hand, the statistical analysis includes correlations and estimation of the WQPs, WQI Models. The final part of this chapter includes the assessment of turbidity, TSS and TSI and shows the results of applying the estimated models on Dokan Lake. Finally, conclusions of all findings from this study and recommendations for future works are included in chapter six.

[7]

CHAPTER TWO LITERATURE REVIEW

Chapter Two

Literature Review

Chapter Two Literature Review 2.1 Introduction Water quality monitoring consists of Chemical, physical and biological measurements. Chemical measurements include levels of dissolved oxygen, nutrients, metals and sediments while physical measurements include the general conditions of the water, such as temperature, conductivity, salinity, current speed and direction, water depths, and water clarity. On the other hand, biological measurements include abundance, variety and growth rates of aquatic plant and animal life in a water body or the ability of aquatic organisms to survive in water sample. The traditional methods which were widely used in measuring water quality parameter’s concentration depended on collection of samples in site and laboratory measuring instruments which have several limits, one of these limits is like these measurements occur usually in limited positions, does not give applied result, and needs to time and highly cost to be done. Therefore, discoveries have found a new way in testing by using remote sensing method for estimating water quality parameter’s concentration in water surface. This method is characterizing by accuracy, low cost, widely coverage and applied results, repetition and speed. Therefore, many studies have used this method to explain and representations of water quality parameters. Most of them used and tested many kind of water quality parameters such as chlorophyll – a, Secchi disk and phosphorous by using Landsat TM, ETM or Spot Data and a high coefficient of determination (R2) have obtained in those studies. The most computer software used in the studies are ENVI, Erdas Imagine and ArcGIS software. To facilitate the object of water quality of lake, this review is organized in a convenient manner such that the reader can easily follow what kinds of works concerning water quality are being done and what sort of gaps still exist such that

[8]

Chapter Two

Literature Review

one could contribute to this important topic. Accordingly, this chapter reviews some of the water quality models usually used with remote sensing and GIS. 2.2 Previous Studies The development of remote sensing techniques during 1970s and 1980s in particular the launch of space instruments dedicated to the observation of earth’s environments such as Landsat MSS and, later the AVHRR series of instruments, stimulated research and promoted new applications. Progress in the monitoring and characterisation of plant canopies and ecosystems on the basis of space observations has been achieved in various directions. One of the remote sensing’s applications is water quality which has been studied since the 1980s with different sensors in almost every inland water, Here are several main recent previous studies which have considered the issue of this study earlier by using the remote sensing and GIS techniques for the water quality parameters over 20 years. (Ming, et al., 1996) have established an integrated water quality models with GIS techniques in central Taiwan. This model is developed as a water quality monitoring system data which obtained from SPOT data to get the distribution of chlorophyll-a, Secchi depth (SD), and phosphorus using ERDAS IMAGINE software. Moreover, an approach simulation system is used to estimate the complexity of water quality. The model is based on the band ratio regression to show the water quality conditions for remote sensing monitoring. Consistent with the theoretical basis, red-band infrared band correlation matrix shows the highest rate of chlorophyll correlation and green / red (XS1 / XS2) ratio have a significant correlation with Secchi depth and phosphorus concentrations. A natural logarithmic regression model is also developed for chlorophyll-a, the Secchi depth (SD) and phosphate (PO4) and the results of the regression model shows a statistically significant (R2 = 0.95, P = 0.005) between the band ratios with chlorophyll-a and Secchi depth with a lower correlation (R2= 0.827, P = 0.032) for the phosphorus variable.

[9]

Chapter Two

Literature Review

(Hedger, et al., 2001) have carried out a study on two lakes in the United Kingdom, Loch Awe (surface area of 38.4 km2 and a maximum length of 41 km) and Loch Ness (surface area of 56.4 km2 and a maximum length of 39 km) by exploring the design of sampling strategies with reference to spatial distributions of chlorophyll. The first objective of this study is to show that systematic sampling regimes using geo-statistics to interpolate regional mean water quality lead to more accurate estimates than the random sampling regimes using classical statistics that are more commonly implemented. The second and main objective is to show that it is necessary to sample lakes spatially with sampling regimes (in particular sampling intensity) dependent upon the spatial variation in water quality. At the end, it has been tried to object and highlight the role that remote sensing may play in helping to optimize sampling. The result of this study explores that the systematic scheme is optimal and always more efficient than a random sampling scheme and a relative advantage of the geo-statistical approach over the classical approach. (Olmanson, et al., 2002) have developed an image processing and classification procedure based on a strong relationship between Landsat TM bands 1 and 3 and Secchi disk transparency and applied it to 40 Landsat images to classify the water clarity of over 10,000 lakes in the State of Minnesota for the two different time periods. The resulting atlas of water clarity is being used to assess spatial and temporal patterns in lake water clarity based on surrounding land use and cover by using a geographical information system (GIS) to link the lake clarity data with land-use features. (Carlson & Ecker, 2002) have statistically compared the water quality in two lakes (Silver and Casey, USA) and examined whether each lake had changed or not, in terms of water quality variables, from year 1999 to 2000. In addition, the most effective variables are explored phosphorus levels in each lake in 2000 and tried to make a regression analyses to show that, in Silver Lake, phosphorus levels increased during the summer of 2000 while they decreased with increasing levels of surface dissolved oxygen and decreased as the water became less clear. [ 10 ]

Chapter Two

Literature Review

At final, phosphorus levels are shown in Lake Casey decreased as the water became less clear and a significant relationship between phosphorus in the sediment and depth exists in Lake Casey. On the other hand, the study concludes that a significant 2-dimensional spatial correlation cannot be shown in Silver Lake, while the spatial analyses does not show the existence of a significant 3dimensional spatial correlation in Lake Casey. (Hellwegera, et al., 2004) have investigated utility of satellite imagery for water quality studies in New York Harbour. Ground data from a routine sampling program (New York Harbour Water Quality Survey) are compared to imagery from the Landsat Thematic Mapper (TM) and Terra Moderate Resolution Imaging Spectroradiometer (MODIS) sensors. Using a time-averaged spatial analysis, it is shown that turbidity as determined from Secchi depth correlates with Landsat TM red reflectance in regions affected by the Hudson River sediments (N=21, R2=0.85). Based on this correlation, the estuarine turbidity maximum of the Hudson River is mapped. Landsat TM red reflectance is also used to identify and map plumes of increased turbidity caused by rainfall runoff and/or Spring tide resuspension in Newark Bay. Chlorophyll-a concentration correlates with the ratio of Landsat TM green to red reflectance in the eutrophic East River and Long Island Sound (N=16, R2=0.78). Terra MODIS estimates of chlorophyll-a show no correlation with ground observations in New York Harbour. (Hua, et al., 2004) have explored the potential of Tampa Bay, FL for using MODIS medium-resolution bands (250- and 500-m data at 469-, 555-, and 645nm) for estuarine monitoring. Field surveys during 21–22 October 2003 has shown that Tampa Bay has a quality of waters, for the salinity range of 24–32 psu; chlorophyll concentration (11 to 23 mg m3), coloured dissolved organic matter (CDOM) absorption coefficient at 400 nm (0.9 to 2.5 m1), and total suspended sediment concentration (2 to 11 mg L-1). CDOM is the only constituent that showed a linear, inverse relationship with surface salinity, although the slope of the relationship changed with location within the bay. The MODIS mediumresolution bands, although designed for land use, are 4–5 times more sensitive [ 11 ]

Chapter Two

Literature Review

than Landsat-7/ETM+ data and are comparable to or higher than those of Coastal Zone Color Scanner (CZCS). Several approaches are taken into account to derive synoptic maps of water constituents from concurrent MODIS medium-resolution data. They have found that application of various atmospheric-correction algorithms yielded no significant differences, due primarily to uncertainties in the sensor radiometric calibration and other sensor artefacts. However, where each scene could be ground trothed, simple regressions between in situ observations of constituents and at-sensor radiances provided reasonable synoptic maps. They have addressed the need for improvements of sensor calibration/characterization, atmospheric correction, and bio-optical algorithms to make operational and quantitative use of these medium-resolution bands. (Doxaran, et al., 2005) have measured and related the concentrations of coloured dissolved organic and suspended (total, organic and inorganic) matter to in situ hyperspectral remote-sensing Reflectance (Rrs) measurements in the Tamar estuary (south-west UK). A simple method has used to determine the Rrs signal from underwater optical measurements, in order to avoid any surface reflection effects. They have obtained linear relationships and a high correlation between the Rrs (850 nm)/Rrs (550 nm) ratio and total suspended matter concentration and inorganic suspended matter concentration. Also they found that the Rrs (400 nm)/ Rrs (600 nm) ratio correlates with the coloured dissolved organic matter concentration, according to a power law regression. These relationships, which appeared to be invariant during the summer period (June– September) and valid for the whole estuary, may be applied to airborne remote sensing data to map the tidal movements of turbidity in the estuary. (Sudheer, et al., 2006) have employed the artificial neural network (ANN) technique to decorrelate satellite data in order to assess water quality in Beaver Reservoir in Arkansas, USA. In their study, they have proposed a method that establishes the output sensitivity toward changes in the individual input reflectance channels while modelling water quality from remote sensing data collected by Landsat thematic mapper (TM). From the sensitivity, a hypothesis [ 12 ]

Chapter Two

Literature Review

about the importance of each band can be made and used as a guideline to select appropriate input variables (band combination) for ANN models based on the principle of parsimony for water quality retrieval. The approach has been explored through a case study of Beaver Reservoir in Arkansas, USA. The results of the case study are highly promising and validate the input selection procedure outlined in their study. The results indicate that this approach could significantly reduce the effort and computational time required to develop an ANN water quality model. (Alparslan, et al., 2007) have worked on water quality at Omerli dam, which is a vital potable water resource of Istanbul city, Turkey. To assess the water quality, they use the first four bands of Landsat 7-ETM satellite data, acquired in May 2001 and water quality parameters, such as chlorophyll-a, suspended solid matter, Secchi disk transparency and total phosphate that measured at several measurement stations at Omerli dam during satellite image acquisition time and archived. They have tried to establish a relationship between this data, and the pixel reflectance values in the satellite image, chlorophyll-a, suspended solid matter, Secchi disk and total phosphate. First, they concerted satellite image to digital number values to unit less planetary reflectance to eliminate the effects of high local variability on remote sensing observation values then atmospheric and geometric correction of the satellite image have done. Finally, the regression model between the pixel reflectance values and the water quality parameters have been established. At the end of their research a new water quality has been derived with their regression numbers for chlorophyll-a, suspended solid matter, Secchi disk and total phosphate. The result reveals very high accuracy (R 2 values), for the suspended solid matter, Secchi disk and total phosphate parameters except for chlorophyll-a. (Nas, et al., 2008) have studied the spatial distribution for water quality parameter in Beysehir Lake which it is a drinking and irrigation water source for [ 13 ]

Chapter Two

Literature Review

the Central Anatolia, Italy. They have attempted to use the Terra ASTER satellite image as remote sensing data source for water quality mapping. The water samples have been collected from 40 stations and the physical and chemical parameters (pH, DO, Secchi Disk Depth (SDD), Turbidity, Conductivity, TSS, Alkalinity, COD, BOD, TN, TP, NO3, NH4 and chlorophyll-a) values are determined. According to the water quality values (TP, SOD, chi-a) the trophic level of the lake is determined. By using multiple regression, ordinary kriging interpolation technique and aid of software packages, they have produced the spatial distribution of water quality parameters over the lake. They conclude that the simultaneous ground and satellite remote sensing data are highly correlated. (Kabbaraa, et al., 2008) have investigated water quality in the coastal area of Tripoli, Lebanon using Landsat 7 ETM+ data to obtain a baseline for coastal resources management. Their data are geometrically rectified to a standard geographical grid and brightness values are converted to reflectance through radiometric correction. Sea-truth data, collected in the field within 6 hours before/after the time of the satellite overpass, are used to derive empirical algorithms for chlorophyll-a concentration, Secchi disk depth and turbidity. The maps of the distribution of the selected water quality parameters are generated for the entire area of interest, and compared with analogous results obtained from SeaWIFS data. The maps indicate that the Tripoli coastal area is exposed to moderate eutrophic conditions, along most of its shoreline (in particular along the northern stretch), in correspondence with fluvial and wastewater runoff sources. They proved that Landsat 7 ETM+ data is useful for the intended application, and can be used to start a national database on water quality in the Lebanese coastal environment. (Almdny, et al., 2008) have used the application of remote sensing and GIS approaches to monitoring the changes in plant cover under degradation in water quality in coastal areas. The study area has been selected in high intensity of plant cover located in the north western part of Libya. A variety of data including filed [ 14 ]

Chapter Two

Literature Review

investigation, groundwater quality for different period, and land cover classification from satellite images for different period have been collected. The water quality data has been collected for two periods, years (1970-1980) and year (2006). Satellite images database for the period of 1980 and 2006 are selected for the study area. Supervised classification is carried out by using ERDAS imaging software and the Classified and Normalized Difference Vegetation Index (NDVI) outputs of two time periods are compared to derive information on changes that occurred over a period of time in the study area. Water quality database and Satellite images database are integrated and analysed in GIS to evaluate the changes in plant cover under degradation in water quality in the study area. Their approaches very good results that help in monitoring environmental changes in the plant cover that effected by degradation in water quality in the coastal area. (HE, et al., 2008) have developed and analysed water quality retrieval models for Guanting Reservoir, which is the auxiliary drinking water source of Beijing, China, by using Remote Sensing (RS) approach using Landsat 5 Thematic Mapper (TM) data. Eight common parameters are concerned water quality variables, algae content, turbidity, and concentrations of chemical oxygen demand, total nitrogen, ammonia nitrogen, nitrate nitrogen, total phosphorus, and dissolved phosphorus, are taken in to consideration in their study. The results show that there is a statistical significant correlation between each water quality variable and remote sensing data in the slightly-polluted inland water body with fairly weak spectral radiation. With appropriate method of sampling pixel digital numbers and multiple regression algorithms, algae content, turbidity, and nitrate nitrogen concentration could be retrieved within 10% mean relative error, concentrations of total nitrogen and dissolved phosphorus within 20%, concentrations of ammonia nitrogen and total phosphorus within 30% while chemical oxygen demand had no effective retrieval method. (Mancino, et al., 2009) have worked on bio-physical parameters with association of water quality by a calibration of model which is atmospherically [ 15 ]

Chapter Two

Literature Review

corrected. They have used stepwise multiple regression to develop models for the Secchi disk depth and chlorophyll concentration parameters for very small lakes at Monticchio, Italy, using Landsat TM data. The results show that the values of water transparency are strongly correlated with chlorophyll-a concentration by linear relationship between the two parameters. They have found that the ratios among visible bands and, most significantly, between TM2 and TM3 are good predictors both of transparency and chlorophyll concentration because the TM2 and TM3 bands have opposing trends in reflectance as a function of chlorophyll concentration in which an increase of chlorophyll-a concentration markedly increases reflectance in band TM2 whereas it results in a stronger absorbance in band TM3. The study present that the application of this approach on lakes with a small surface area is effective and the developed model well describes the water quality parameters. (Mahdi, et al., 2009) have used remote sensing techniques to monitor the water depth and water clarity within the Iraqi marshlands, considering Al Huweizah Marsh as a case study, and to estimate the discharges of Al Karkheh River into Al Huweizah Marsh with satellite images of Landsat-7 TM for 1990 and Landsat-7 ETM for 2000, 2002, and 2006. In their research, water depth measurements are performed in 982 points along four profiles and 50 points distributed within the marsh area. Water samples collection and spectral reflectance measurements have been carried out at the same 50 points where the water depths are measured. At first the collected water samples have been tested in the laboratory to specify some of the physical and chemical characteristics for the water of these samples. A laboratory simulation system is used in the study to evaluate the relation between the water depth and spectral reflectance in the Blue and Green spectrum bands by measuring the spectral reflection of a water body with different water depths and qualities. The water quality effect on the estimated water depths has been studied by using the ratio transform algorithm based on the results of the field and laboratory measurements. Their study shows that the marsh water [ 16 ]

Chapter Two

Literature Review

quality has a little effect on the relation between the estimated water depths and the spectral reflectance of the Blue, Green and Red spectrum bands. The relation between the reflectance of these bands and the values or concentrations of the considered water quality parameters cannot be detected with these ranges of values or concentrations of the marsh water. Finally, a linear regression model with acceptable correlation are developed with R2=0.78 which was obtained considering the measured water depths in the water area of the marsh and the corresponding values of the ratio Ln(Blue)/Ln(Green) which are extracted from the satellite image of Landsat-7 ETM for 2/2/2006. On the other hand, their water clarity estimation model, using the trophic state index (TSI) which is an indicator for the water clarity, is adopted to estimate the TSI distribution within the marsh area. A set of measured Secchi disk (SD) at 20 points and the obtained spectrum reflectance at these points by using the Landsat ETM images of 2/2/2006 has been used to perform the calibration process. The calibrated model is applied on the marsh area using the Landsat ETM images of 2/2/2006 and 6/3/2006. (Bilgehan, et al., 2010) have presented an application of water quality mapping by satellite Landsat 5TM image and ground data for Lake Beysehir, Turkey. A satellite image data source is collected for water quality mapping based on simultaneously obtained in situ lake water quality measurements, Suspended Sediment (SS), turbidity, Secchi disk depth (SDD), and chlorophyll-a data for 40 stations in August, 2006. Spatial patterns in these parameters are estimated using bivariate and multiple regression (MR) techniques based on Landsat-5TM multispectral data and water quality sampling data. Their study aims to estimate single TM bands, band ratios, and combinations of TM bands and correlate them with the measured water quality parameters. The results show that the measured and estimated values of water quality parameters are in good agreement and TM3 provided a significant relationship with SS concentration while TM1, TM2, and

[ 17 ]

Chapter Two

Literature Review

TM4 are strongly correlated with measured chlorophyll-a concentrations. On the other hand, TM1, TM2, and TM3 are moderately correlated to turbidity. (Alparslan, et al., 2010) have investigated multi-temporal changes of water quality parameters in Darlik dam, Istanbul, Turkey. The multi-temporal Landsat 5 TM satellite images of the Dam and its watershed acquired in years 2004, 2005, and 2006 were employed in the study. Multiple regression models are developed for chlorophyll-a, total deposited matter, Secchi disk, total phosphate, and total nitrate parameters for seven stations in the lake. In his study, a high correlation detected between estimated and observed values for water quality parameters and accurate maps have been obtained for parameters in the lake. (Somvanshi, et al., 2011) have estimated water turbidity for Gomti River Lucknow and parts of Sitapur and Barabanki districts lies between 27012'13"to 27043'19" N latitude 800 46'40" to 81012' 59" E longitude with the aim of retrieval of turbidity, from the Google Earth’s Quickbird satellite data. The water turbidity has been estimated by Normalized Difference Turbidity Index (NDTI) and classified on the basis of mean and standard deviation of image using Erdas Imagine 9.1 software. Accordingly, water turbidity is classified into three classes i.e. low, moderate and high having area of 3.22 ha, 774.85 ha and 4.96 ha respectively. The information generated in the study can be used as an important input in the water quality management. (Somvanshi, et al., 2012) have developed a decision making tool for mapping of water quality parameters for Gomti River, Lucknow, Sitapur and Barabanki districts of Uttar Pradesh, India. Mapping has been done by using IRS LISS III data combined with measurement of selected sample points. Radiance value of each band of IRS LISS III data has been calculated and observed radiance value on those sample points of each band along with band ratios and principal components were compared with in situ measurements of water quality parameters. The water quality parameters included in their study are TS, DS, SS, pH, COD, BOD, DO, Chloride and TH. Appropriate band combinations are [ 18 ]

Chapter Two

Literature Review

selected to develop multiple linear regression models and the results show that the IRS P6 LISS III radiance data can be successfully used to map some surface water quality parameters. (Al-Bahrani, et al., 2012) have studied the water quality indices and their classifications for irrigation use at many stations along the Euphrates River, inside the Iraqi lands. They have correlated the results of the water quality indices with the satellite images from LANDSAT satellite and modified at November, 2009 for bands 1, 2 and 3. They have analyzed their data to develop a colored water quality model which can be used to predict irrigation usability of the water at any point along the river within Iraq. In their study fifteen irrigation water quality parameters (chlorides, sulfates, total dissolved solids, total nitrates, electrical conductivity, pH, sodium adsorption ratio, iron, lead, zinc, cadmium, copper, boron, chromium, and coliform bacteria) has been used during the period from April, 2007 to December, 2010 for sixteen stations along the Euphrates River from its entrance to the Iraqi lands at Al-Qaim in Anbar governorate to its union with Tigris River at Qurna in Basrah governorate. They have built a regression models for water quality indices with a correlate of Coordinates of the sixteen stations of the Euphrates River which have been projected at the mosaic of Iraq satellite image. Regression models at significant 0.01 are built between the water quality index of the sixteen stations and their digital number at band 2 during the study period. The best models for Bhargava and the Canadian methods have been used to develop a colored model can be used to reflect the water quality classifications of the two mathematical methods for irrigation usability on the satellite image. These colored models can be used to estimate the water quality classification at any point along the river. (Kutser, et al., 2012) have studied optically very variable lakes in order to test both physics based methods and conventional band-ratio type algorithms in retrieval of water parameters. The study shows that the spectral library model in physics based approach provides very good results for chlorophyll-a retrieval [ 19 ]

Chapter Two

Literature Review

while the number of different concentrations of CDOM and suspended matter used simultaneously is too low to provide good estimates of these parameters. (Ghazal & Hassoon, 2012) have calculated the ground temperature for a part of the province of Sulaimani-Northern Iraq, using the thermal bands (low acquisition, and high acquisition) of the ETM+ sensor mounted in the LANDSAT 7 satellite and the process implemented by ERDAS Imagine 9.2 software package. A mathematical model is developed and the results indicate that there is a significant convergence between the values of temperature calculated from both thermal bands (low acquisition, and high acquisition). The second part of the work includes the distribution of temperature degree over the study area which is extracted as a raster map by averaging both low gain and high gain thermal bands. (Gardino, et al., 2014) have given an overview of optical remote sensing techniques applied to water quality monitoring in Maggiore Lake, Italy. They have explored the techniques by presenting the temporal trend of chlorophyll-a, suspended particulate matter (SPM), coloured dissolved organic matter (CDOM) and the z90 signal depth (depth from which 90% of the reflected light comes from) as estimated from the images acquired by the Medium Resolution Imaging Spectrometer (MERIS) in the pelagic area of the lake from year 2003 to 2011. Concerning chlorophyll-a trend, the results are in agreement with the concentration values measured during field surveys, confirming the good status of Lake Maggiore. They assume that the recently launched (e.g., Landsat-8) and the future satellite missions (e.g., Sentinel-3) carrying sensors with improved spectral and spatial resolution are going to lead to a larger use of remote sensing for the assessment and monitoring of water quality parameters and also allowing further applications (e.g., classification of phytoplankton functional types) to be developed. (Peng Li, 2014) have focused on the substantial differences between instruments on-board of Landsat-7 Enhanced Thematic Mapper Plus (ETM+) and Landsat-8 Operational Land Imager (OLI) and Thermal Infrared Sensor (TIRS) which are currently operating for routine earth observation. Their enhancements [ 20 ]

Chapter Two

Literature Review

have achieved with Landsat-8 refer to the scanning technology (replacing of whisk-broom scanners with two separate push-broom OLI and TIRS scanners), an extended number of spectral bands (two additional bands provided) and narrower bandwidths. Therefore, cross-comparative analysis is very necessary for the combined use of multi-decadal Landsat imagery. In the study, 3311 independent sample points of four major land cover types (primary forest, unplanted cropland, swidden cultivation and water body) are used to compare the spectral bands of ETM+ and OLI. Eight sample plots with different land cover types are manually selected for comparison with the Normalized Difference Vegetation Index (NDVI), the Modified Normalized Difference Water Index (MNDWI), the Land Surface Water Index (LSWI) and the Normalized Burn Ratio (NBR). So these indices are calculated with six pairs of ETM+ and OLI cloudfree images, which were acquired over the border area of Myanmar, Laos and Thailand just two days apart, when Landsat-8 achieved operational obit. Their final comparative results show that the average surface reflectance of each band differed slightly, the subtle differences of vegetation indices derived from both sensors demonstrated that ETM+ and OLI imagery can be used as complementary data and the LSWI and NBR performed better than NDVI and MNDWI for crosscomparison analysis of satellite sensors, due to the spectral band difference effects. (Akbar, et al., 2014) have developed, evaluated, and applied the remote sensing based models for Canadian Water Quality Index (CWQI) and turbidity for the Bow River of Alberta. They have considered 31 scenes of Landsat-5 TM satellite data to establish the relationship between the planetary reflectance and the monthly ground measured data for the period of 5 years (2006 - 2010). The four spectral bands (blue, green, red, and near infrared) are used to obtain the most suitable models from 26 different band combinations. The best-fit models are validated with ground measured data and found that 72% of the data showed 100% matching for the CWQI model and 83% of the data for the turbidity model. The Landsat-5 TM based CWQI and turbidity models are applied on all the scenes [ 21 ]

Chapter Two

Literature Review

to obtain five CWQI classes (excellent, good, fair, marginal and poor), and six classes of turbidity (i.e. 0–10 NTU, 10–20 NTU, 20–30 NTU, 30–40 NTU, 40– 50 NTU, >50 NTU). The variation of river water quality in different years of interest is associated with the climatic changes. The most deteriorated water quality noted in two natural sub-regions included mixed grass and dry mixed grass, which could be related to irrigation-based farming. (Fan, 2014) has developed algorithms for hyper spectral remote sensing of water quality based on in situ spectral measurement of water reflectance. The water reflectance spectra R(λ) are acquired by a pair of Ocean Optic 2000 Spectroradiometer during the summers from year 2008 to 2011 at Patuxent River, a tributary of Chesapeake Bay, USA. Simultaneously, concentrations of chlorophyll-a and total suspended solids (TSS), as well as absorption of coloured dissolved organic matter (CDOM) are measured. Empirical models that based on spectral features of water reflectance generally showed good correlations with water quality parameters. The retrieval model that uses spectral bands at red/NIR showed a high correlation with chlorophyll-a concentration and CDOM absorption is best correlated with spectral features at blue and NIR regions. On the other hand, the results showed that the ratio of green to blue spectral bands is the best predictor for TSS. (Rokni, et al., 2014) have modelled the spatiotemporal changes of Urmia Lake, Middle East in the period 2000–2013 using the multi-temporal Landsat 5TM, 7-ETM+ and 8-OLI images. In their study, the applicability of different satellite-derived indexes including Normalized Difference Water Index (NDWI), Modified NDWI (MNDWI), Normalized Difference Moisture Index (NDMI), Water Ratio Index (WRI), Normalized Difference Vegetation Index (NDVI), and Automated Water Extraction Index (AWEI) are being investigated for the extraction of surface water from Landsat data. They were found that the NDWI is superior to other indexes and hence it was used to model the spatiotemporal changes of the lake. In addition, a new approach based on principal components of multi-temporal NDWI (NDWI-PCs) has been proposed and evaluated for [ 22 ]

Chapter Two

Literature Review

surface water change detection. The results indicate an intense decreasing trend in Urmia Lake surface area in the period 2000–2013, especially between 2010 and 2013 when the lake lost about one third of its surface area compared to the year 2000. The results illustrated the effectiveness of the NDWI-PCs approach for surface water change detection, especially in detecting the changes between two and three different times, simultaneously. (Giardino, et al., 2014) have evaluated the capabilities of three satellite sensors for assessing water composition and bottom depth in Garda Lake, Italy. A consistent physics-based processing chain is applied to Moderate Resolution Imaging Spectro-radiometer (MODIS), Landsat-8 Operational Land Imager (OLI) and Rapid-Eye. The computed remote sensing reflectance (Rrs) from MODIS and OLI are converted into water quality parameters by adopting a spectral inversion procedure based on a bio-optical model calibrated with optical properties of the lake. The same spectral inversion procedure is applied to RapidEye and to OLI data to map bottom depth. In situ measurements of Rrs and of concentrations of water quality parameters collected in five locations are used to evaluate the models. The bottom depth maps from OLI and Rapid-Eye showed similar gradients up to 7 m. The results indicate that the spatial and radiometric resolutions of OLI enabled mapping water constituents and bottom properties; MODIS is appropriate for assessing water quality in the pelagic areas at a coarser spatial resolution; and Rapid-Eye had the capability to retrieve bottom depth at high spatial resolution. 2.3 Summary From all the reviewed researches in the previous section, the followings can be concluded: 

All related research tested many kind of water quality parameters but mostly is chlorophyll – a, Secchi disk and phosphorous.



Most studied used correlations between water quality parameters and Landsat TM, Landsat ETM or Spot Data.

[ 23 ]

Chapter Two



Literature Review

The most computer software used in previous studies are ENVI, Erdas Imagine and ArcGIS software.



The best models obtained in previous studies are regression models but a few researchers use the combination of traditional methods and artificial neural networks (ANN) for developing their models. Therefore, the present study focuses on the modelling of more important

water quality parameters using remote sensing and GIS as follow: 1. The study uses Landsat 8 OLI which have not used before in previous studies. 2. The water quality parameters that used is this study are pH, Total Dissolved Solids (TDS), Electrical Conductivity (EC), Dissolved Oxygen (DO), Biochemical Oxygen Demand (BOD), Nitrate (NO3), Nitrate Nitrogen (NO3N), Phosphate (PO4), Total Phosphorous (TP), Temperature, Total Suspended Solids (TSS), Turbidity and finally Secchi Disk transparency (SDT). 3. The water quality index that used in this study are National Sanitation Foundation (NSFWQI), Arithmetic Weighted Formula (AWWQI), Canadian Council of Ministers of the Environment (CCMEWQI) Oregon Water Quality Index (OWQI). 4. The Secchi Disk Transparency (SDT) measured to calculate Trophic State Index (TSI) and developing models from Landsat 8 OLI reflectance bands. 5. A new band such as Coastal Blue of Landsat 8, which it is not used before, are used in this study. 6. A new ratios and combinations of spectral bands of Landsat 8 OLI and their transforms such as Independent Component Analysis (ICA), Minimum Noise Fraction (MNF), are used in this study.

[ 24 ]

CHAPTER THREE THEORETICAL CONCEPTS

Chapter Three

Theoretical Background

Chapter Three Theoretical Background 3.1 Introduction Knowledge of the spatial distribution of different biological, chemical and physical variables is essential in environmental water studies as well as for resource management (Meer & Jong, 2006). Hence, coupled with advanced processing methods and improved sensor capabilities, recent years have seen increasing interest and research in remote sensing of the quality of inland and coastal waters. Unless a water body is sufficiently instrumented by in situ sensors, remote sensing is the only satisfactory method to detect the quality of remote and large inland waters. Thermal remote sensing, for example, can be used to define groundwater discharge zones. Remote sensing technology provides an emerging capability that can significantly augment or replace traditional in situ methods but the field is relatively new, especially in addressing optically complex waters. Satellite image archives exist, dating back to 1970, enabling change detection of previously unmeasured or unmonitored water bodies. Remote sensing is a suitable technique for coarse scale monitoring of inland and coastal water quality and provides a synoptic view of the spatial distribution of different biological, chemical and physical variables of both the water column and, if visible, the substrate (WMO, 2013). Therefore, this chapter provides the basic and principle theories related to water quality, remote sensing, satellite images and statistical analysis that depended in the present study as described in the next sections. 3.2 Water Quality Water quality monitoring provides the practical and methodological details whereas water quality assessments gives the overall strategy for assessments of the quality of the main types of water body. Together they cover all the major aspects of water quality, its measurement and its evaluation. This section concentrates on the theoretical process of setting up monitoring programs for the [ 25 ]

Chapter Three

Theoretical Background

purpose of providing a valid data base for water quality assessments. The choice of variables to be measured in the water and the common procedures for data handling and presentation in theoretical way. Finally, it focuses on monitoring strategies, requirements for water quality and quantity data and interpretative techniques. 3.2.1 When to Sample The water quality assessment’s objectives are to set a primary factor that influencing sampling which is when to start. The most recommended critical time period in a lake is typically during the growing season. For general water quality assessment purposes, it is sufficient to monitor from April or May through September or October, either monthly or preferably, every two weeks. There is little benefit in monitoring more than every two weeks and, in fact, for some, parameters there are statistical reasons for not doing so. If the samples have taken on a lake throughout the year, even research professionals typically take samples only monthly during the winter. Samples also should be collected at about the same time of day each time of sampling. This allows for some consistency in daylight hours and in all the indirect effects daylight has on the different lake processes (Michaud, 1991). When the water quality assessment is provided with remote sensing the time should be on exact date of acquisition date satellite image that captured in that date. 3.2.2 Selecting Sampling Locations The selection of sampling stations is directly dependent upon the monitoring objectives. If the objective is to characterize the entire lake, it is likely to place a number of stations on lake to provide an adequate characterization, while when the objective is to learn how lakes function; the stations should be in physically diverse locations and at different depths should be selected. Conversely, if tracking water quality trends to goals, it could be argued that one station could be used to represent any lake.

[ 26 ]

Chapter Three

Theoretical Background

Sampling should be avoided near shore, near inflows, or in the downwind direction. The prevailing wind is blowing algae, plankton, and debris toward the bottom of the lake and the shoreline; samples collected in these areas are less representative of the lake’s overall water quality. If a boat is not available, a location chosen in about midway down the shoreline and sample off a long pier, using a pole to collect the sample as far from the shoreline as possible. In deep lakes, sampling at two depths (near surface and near bottom) is a good idea and that depends on which kind of depth tries to be studied. In a large lake, if there are sufficient volunteers for more than one station, choose additional stations according to the objective or physical aspects of the lake. Addition of a station at the input of the lake helps in figuring out how much they contribute to pollution in the lake (Michaud, 1991). 3.2.3 Selection Water Quality Parameters The selection of variables to be included in a water quality assessment must be related to the objectives of the water quality assessment program and the factors that depend on it. Broadly, assessments can be divided into two categories, use-orientated and impact-orientated. In addition, operational surveillance can be used to check the efficiency of water treatment processes by monitoring the quality of effluents or treated waters, but this is not discussed here. Use-oriented assessment tests and evaluate water quality whether it is satisfactory for specific purposes, such as drinking water supply, industrial use or irrigation. Many uses of water have specific requirements with respect to physical and chemical variables or contaminants. Guidelines and criteria list identify the minimum set of variables to be included in the evaluation such as water quality assessment programs. Table 3 in Appendix A indicates that the appropriate variables for specific uses of water can be used where the guidelines are not available. Other variables can also be monitored, if necessary, according to the special conditions related to the use intended (Chapman, 1996). The use-oriented monitoring includes the followings:

[ 27 ]

Chapter Three

Theoretical Background

1. Background monitoring: a quality knowledge of the background necessary to assess the suitability of water for use and disclosure of human impacts in the future. 2. Aquatic life and fisheries: individual aquatic organisms have requirements, which differ with respect to the physical and chemical properties of the water bodies. Therefore, various important guidelines for water have been proposed for fisheries or aquaculture. 3. Drinking water sources: the water that is used for drinking should be monitored for variables that may pose a potential threat to human health. It has been developed guidelines for maximum concentrations of these variables in drinking water by the regional and national authorities, such as World Health Organization (WHO, 1993). 4. Recreation and health: using water for cleaning and recreation have an associated health risk due to poor quality because of possibility of eating small amounts, or pathogens enter directly in the eyes, nose, ears or open wounds. It is mostly recommended the variables in association of entertainment and health with pathogens or the aesthetic quality of the water in guidelines and standards. 5. Agricultural use: irrigation of food crops represents a danger to the health of consumers of food and it is possible if the quality of irrigation water is not enough, particularly with regard to pathogens and toxic compounds. It has been appointed that the recommendation guidelines for some of the variables in the irrigation water, but may be tolerated with the highest levels if the water was scarce. 6. Industrial uses: industry requirements for a variety of water quality, depending on the nature of the industry and individual processes using water within the industry. So water quality parameters, too, should be selected to meet project objectives, number of volunteers, and available money. Field measurements such as pH, temperature, dissolved oxygen, and Secchi depth are inexpensive to [ 28 ]

Chapter Three

Theoretical Background

measure once the initial equipment or chemical reagents have been purchased. This information alone is enough to do a general water quality assessment, determine trophic state, provide plenty of educational information, and even describe water quality trends if data are collected for a long enough period. Including additional parameters such as nutrient analyses and chlorophyll-a just provides more in-depth information for meeting these same objectives. Since these parameters can be more expensive to analyse depending upon the measurement method used, a decision on whether to measure them and at how many stations will be money-dependent. If only a few, nutrient or chlorophyll-a samples can be collected, stations that best meet monitoring objectives should be picked (Michaud, 1991). The water quality variables have been considered in the present study with a brief description for each one are listed below: A. Temperature The temperature of surface waters influenced by latitude, altitude, and season, time of day, air circulation, cloud cover and the flow and depth of the water body. Temperature measured in situ, using a thermometer or thermistor. The temperature affects physical, chemical, biological processes and most important changes in concentration of many variables in lake water (Michaud, 1991). As temperature has an influence on so many aquatic variables and processes, it is important always to include it in a sampling regime, and to take and record it at the time of collecting water samples (Chapman, 1996). B. Dissolved Oxygen (DO) Oxygen is essential to all aquatic organism, including their responsibility for the self-purification processes in water bodies that varies with temperature, salinity, turbulence, the photosynthetic activity of algae and plants, and atmospheric pressure of water. Determination of DO concentrations is important factor for water quality assessment that influences nearly all chemical and biological processes within water bodies. Variations in DO can occur seasonally,

[ 29 ]

Chapter Three

Theoretical Background

or even over 24 hour periods, in relation to temperature and biological activity (photosynthesis and respiration). The measurement of DO can be used to indicate the degree of pollution by organic matter, the destruction of organic substances and the level of selfpurification of the water. Its determination is also used in the measurement of biochemical oxygen demand (BOD) (Chapman, 1996). C. Biochemical Oxygen Demand (BOD) The biochemical oxygen demand (BOD) is an approximate measure of the amount of biochemically degradable organic matter present in a water sample. It is defined by the amount of oxygen required for the aerobic microorganisms present in the sample to oxidize the organic matter to a stable inorganic form (Chapman, 1996). BOD data are normally required to know the strength of a waste, which has to be treated by biological means, as in an oxidation ditch or percolating filter and where wastes are being discharged to receiving waters (EPA, 2001). D. Electrical Conductivity (EC) Conductivity of a water is an expression of its ability to conduct an electric current and expressed as micro Siemens per centimetre (μhos/cm). It has related to the ionic content of the sample, which is in turn a function of the dissolved (ionisable) solids concentration. In itself, conductivity is a property of little interest to a water analyst but it is an invaluable indicator of the range into which hardness and alkalinity values are likely to fall, and also of the order of the dissolved solids content of the water. Conductivity is usually measured in situ with a conductivity meter, and may be continuously measured and recorded. Such continuous measurements are particularly useful in rivers for the management of temporal variations in TDS and major ions (Chapman, 1996). E. Total Dissolved Solid (TDS) Dissolved solids in water samples include soluble salts that yield ions such as sodium, calcium, magnesium… etc. High TDS levels, especially due to [ 30 ]

Chapter Three

Theoretical Background

dissolved salts, affect many forms of aquatic life, add a laxative effect to water or cause the water to have an unpleasant mineral taste and change the pH of water. Total dissolved solids (mg/l) can be determined by evaporating a pre-filtered sample to dryness, and then finding the mass of the dry residue per liter of sample. A second method may be obtained by multiplying the conductance by a factor, which is commonly between 0.55 and 0.75. It is often convenient and acceptable to use the very rapid determination of conductivity to give an estimation of the total dissolved solids (Chapman, 1996). F. pH, Acidity and Alkalinity pH is the negative logarithm of the hydrogen ion concentration of a solution and it is thus a measure of whether the liquid is acid or alkaline. The pH scale ranges from 0 (very acid) to 14 (very alkaline) and with pH 7 representing a neutral condition (Michaud, 1991). In unpolluted waters, pH is principally controlled by the balance between the carbon dioxide, carbonate and bicarbonate ions as well as other natural compounds such as humic and fulvic acids. The natural acid-base balance of a water body can be affected by industrial effluents and atmospheric deposition of acid-forming substances. Changes in pH can indicate the presence of certain effluents, particularly when continuously measured and recorded, together with the conductivity of a water body. Diel variations in pH can be caused by the photosynthesis and respiration cycles of algae in eutrophic waters (Chapman, 1996). G. Phosphate (PO4) Phosphorus is an essential nutrient for living organisms and exists in water bodies as both dissolved and particulate species. It is generally the limiting nutrient for algal growth and, therefore, controls the primary productivity of a water body. The significance of phosphorus is principally concerning the phenomenon of eutrophication (over-enrichment) of lakes and, to a lesser extent, rivers. Phosphorus is rarely found in high concentrations in freshwaters as plants [ 31 ]

Chapter Three

Theoretical Background

actively take it up. As a result, there can be considerable seasonal fluctuations in concentrations in surface waters. In most natural surface waters, phosphorus ranges from 0.005 to 0.020 mg/l. Concentrations as low as 0.001 mg/l may be found in some pristine waters and as high as 200 mg/l in some enclosed saline waters (Chapman, 1996). H. Nitrate (𝑵𝑶− 𝟑) The nitrate ion (𝑁𝑂3− ) is the common form of combined nitrogen present in surface and ground waters, because it is the end product of the aerobic decomposition of organic nitrogenous matter. Nitrate is an essential nutrient for aquatic plants and seasonal fluctuations can be caused by plant growth and decay. Natural concentrations, which seldom exceed 0.1 mg/l, may be enhanced by municipal and industrial wastewaters, including leachates from waste disposal sites and sanitary landfills. In rural and suburban areas, the use of inorganic nitrate fertilisers can be a significant source. The determination of nitrate helps the assessment of the character and degree of oxidation in surface waters, in groundwater penetrating through soil layers, in biological processes and in the advanced treatment of wastewater (Bartram & Ballance, 1996 ). In lakes, concentrations of nitrate in excess of 0.2 mg/l tend to stimulate algal growth and indicate possible eutrophic conditions (Chapman, 1996). I. Total Suspended Solid (TSS) Suspended solids is the matter that suspended in quiescent water and consist of finely divided light solids, which may never settle or do so only very slowly. Indeed, the net effect may be one of apparent turbidity without any discernible solids. In flowing water, on the other hand, the solids which are kept in suspension by the turbulence may be settleable if the water is let stand. While the latter would be determinable as "Settleable Solids" and the former could possibly be assessed as "Turbidity". The term total suspended solids (TSS) applies to the dry weight of the material that is removed from a measured volume of water sample by filtration [ 32 ]

Chapter Three

Theoretical Background

through a standard filter. The test is basically empirical and is not subject to the usual criteria of accuracy (Bartram & Ballance, 1996 ). Total suspended solids (TSS) concentrations and turbidity both indicate the amount of solids suspended in the water, whether mineral (e.g., soil particles) or organic (e.g., algae). However, the TSS test measures an actual weight of material per volume of water, while turbidity measures the amount of light scattered from a water sample (more suspended particles cause greater scattering). This difference becomes important when trying to calculate total quantities of material within or entering a lake. High concentrations of particulate matter affect light penetration and productivity, recreational values, and habitat quality, and cause lakes to fill in faster. Particles also provide attachment places for other pollutants, notably metals and bacteria. (Michaud, 1991) J. Turbidity Turbidity in water arises from the presence of very finely divided solids (which are not filterable by routine methods). The existence of turbidity in water will affect its acceptability to consumers and it will also affect markedly its utility in certain industries. As turbidity can be caused by sewage matter in a water there is a risk that pathogenic organisms could be shielded by the turbidity particles and hence escape the action of the disinfectant. The type and concentration of suspended matter controls the turbidity and transparency of the water. Turbidity results from the scattering and absorption of incident light by the particles, and the transparency is the limit of visibility in the water. Both can vary seasonally according to biological activity in the water column and surface runoff carrying soil particles. Heavy rainfall can also result in hourly variations in turbidity. Turbidity should be measured in the field but, if necessary, samples can be stored in the dark for not more than 24 hours. The most reliable method of determination uses nephelometry (light scattering by suspended particles) by means of a turbidity meter which gives values in Nephelometric Turbidity Units (NTU) (Chapman, 1996). [ 33 ]

Chapter Three

Theoretical Background

K. Secchi Disk Transparency and Clarity This parameter gives an indication of the presence or absence of suspended matter, living or inert, and hence it is a reflection of the overall quality of a water. However, it must be remembered that the presence of any undesirable substances in solution will not be indicated by transparency. It is expressed as the maximum depth in metres at which it is possible to distinguish the markings of a Secchi disc, and it is widely used in studies on lakes to assess the abundance of algae. Clarity is affected by algae, soil particles, and other materials suspended in the water. However, Secchi disk depth is primarily used as an indicator of algal abundance and general lake productivity (Michaud, 1991). L. Faecal Coliform Bacteria Faecal coliform bacteria are microscopic animals that live in the intestines of warm-blooded animals. They also live in the waste material. When faecal coliform bacteria are present in high numbers in a water sample, it means that the water may have received faecal matter from one source or another. Although not necessarily agents of disease, faecal coliform bacteria indicate the potential presence of disease-carrying organisms, which live in the same environment as the faecal coliform bacteria. Most states have strict standards for faecal coliform bacteria concentrations, primarily for reasons of public health. The abundance of faecal coliform bacteria is measured as the number of colonies in 100 ml of water. To date the universal indicator organisms have been the coliforms, specifically Escherichia coli (Michaud, 1991). 3.2.4 Water Quality Index Water-quality indices aim at giving a single value to the water quality of a source on the basis of one or the other system which translates the list of constituents and their concentrations present in a sample into a single value. One can then compare different samples for quality on the basis of the index value of each sample. Rather than assigning a numerical value to represent water quality, [ 34 ]

Chapter Three

Theoretical Background

these classification systems categorised water bodies into one of several pollution classes or levels. By contrast, indices that use a numerical scale to represent gradations in water-quality levels are a recent phenomenon (Abbasi & Abbasi, 2012). Therefore, the common indices that use a numerical scale are used in this study and detailed below: A. Brown’s or National Sanitation Foundation Index (NSFWQI) Known as Arithmetic weighted formula, Brown and other researchers in 1970 developed a water-quality index similar in structure to Horton’s index but with much greater rigour in selecting parameters, developing a common scale and assigning weights for which elaborate Delphic exercises are performed. This effort was supported by the National Sanitation Foundation (NSF). (Abbasi & Abbasi, 2012). For this reason, Brown’s index is also referred to as NSF-WQI and was given as: 𝑛

Eq. ( 3-1)

𝑊𝑄𝐼 = ∑ 𝑞𝑖 . 𝑤𝑖 𝑖=1

Where: 𝑞𝑖 : represents the rating for the 𝑖 𝑡ℎ determinant, this value varies from (0-100). 𝑤𝑖 : Represents the weighting for the 𝑖 𝑡ℎ determinant and this value varies from (0-1) and ∑ 𝑤𝑖 =1. 𝑛: Number of determinants. B. Modified Arithmetic Weighted Formula In this method, different water quality components are multiplied by a weighting factor and are then aggregated using simple arithmetic mean. For assessing the quality of water in this study, firstly, the quality rating scale (𝑄𝑖 ) for each parameter is calculated: 𝑄𝑖 =

(𝑉𝑎𝑐𝑡𝑢𝑎𝑙 – 𝑉𝑖𝑑𝑒𝑎𝑙) × 100 (𝑉𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑 – 𝑉𝑖𝑑𝑒𝑎𝑙)

Where:

[ 35 ]

Eq. ( 3-2)

Chapter Three

Theoretical Background

𝑄𝑖 = Quality rating of ith parameter for a total of n water quality parameters. 𝑉𝑎𝑐𝑡𝑢𝑎𝑙 = Actual value of the water quality parameter obtained from laboratory analysis. 𝑉𝑖𝑑𝑒𝑎𝑙 = Ideal value of that water quality parameter can be obtained from the standard tables. 𝑉𝑖𝑑𝑒𝑎𝑙 for pH = 7 and for other parameters it is equaling to zero, but for DO 𝑉𝑖𝑑𝑒𝑎𝑙 = 14.6 mg/L at 0 C0. 𝑉𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑 = Recommended WHO standard of the water quality parameter. Then, after calculating the quality rating scale (𝑄𝑖 ), the relative unit weight (𝑊𝑖 ) was calculated by a value inversely proportional to the recommended standard (𝑆𝑖 ) for the corresponding parameter using the following expression; Eq. ( 3-3)

𝑊𝑖 = 1/ 𝑆𝑖 Where: 𝑊𝑖 = Relative unit weight for nth parameter. 𝑆𝑖 = Standard permissible unitless value for nth parameter. 𝑖 = Proportionality constant.

That means, the relative (unit) weight (𝑊𝑖 ) to various water Quality parameters are inversely proportional to the recommended standards for the corresponding parameters. Finally, the overall WQI is calculated by aggregating the quality rating with the unit weight linearly (Al-Bahrani, et al., 2012) by using the following equation:

𝑊𝑄𝐼 = 𝛴𝑄𝑖 𝑊𝑖 / 𝛴 𝑊𝑖

Eq. ( 3-4)

Where: 𝑄𝑖 = Quality rating. 𝑊𝑖 = Relative weight. C. Oregon Water Quality Index (OWQI) OWQI creates a score to evaluate the general water quality of Oregon’s stream and the application of this method to other geographic regions. The [ 36 ]

Chapter Three

Theoretical Background

original OWQI was designed by Dojlido J., Raniszewsk I. J., and Woyciechowska in 1994 after the NSFWQI where the Delphi method has been used for variable selection. It expresses water quality status and trends for the legislatively mandated water quality status assessment. The index is free from the arbitration in weighting the parameters and employs the concept of harmonic averaging. (Tyagi, et al., 2013) The mathematical expression of this WQI method is given by: 𝑛 𝑊𝑄𝐼 = √ 𝑛 ∑𝑖=1 1⁄ 2 𝑆𝐼𝑖

Eq. ( 3-5)

Where: 𝑛 : Number of sub-indices 𝑆𝐼: Sub index of 𝑖 𝑡ℎ parameter Furthermore, the rating scale of this OWQI has also been categorized in various classes see D. WQI of Relative Sub-Indices This index is defined as the formula that its solution depends on water quality standards. The most important formulas of this approach, which depend in their solution on water quality standards, are Centre St Laurent, Quebec Index, Maximum Operator Formula, Alberta Index, British Columbia Water Quality Index (BCWQI), Ontario Index and Canadian Council of Ministers of the Environment Index (CCME-WQI). Among these indices, the last one is most important formulas and commonly used by researchers, therefore, it is used in the present study. The CCME-WQI has adopted the conceptual model of BCWQI (based on relative sub-indices). There are three factors in the index, each of which has been scaled between 0 and 100. In the CCME-WQI, the values of the three measures of variance from selected objectives for water quality are combined to create a vector in an imaginary ‘objective exceedance’ space. In the index ‘objectives’ refer to Canada wide water-quality guidelines or site-specific water-quality objectives. The length [ 37 ]

Chapter Three

Theoretical Background

of the vector is then scaled to range between 0 and 100, and subtracted from 100 to produce an index which is 0 (or close to 0) for very poor water quality, and close to 100 for excellent water quality. The revised CWQI consists of three factors: a. Factor 1 (F1) This factor is called scope because it assesses the extent of water quality guideline non-compliance over the time period of interest. It has been adopted directly from the British Columbia Index: 𝐹1 =

𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑓𝑎𝑖𝑙𝑒𝑑 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒𝑠 × 100 𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒𝑠

Eq. ( 3-5)

Where variables indicate those water quality parameters with objectives which are tested during the time period for the index calculation. b. Factor 2 (F2) This factor is called frequency and represents the percentage of individual tests that do not meet the objectives (failed tests). 𝐹2 =

𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑓𝑎𝑖𝑙𝑒𝑑 𝑡𝑒𝑠𝑡𝑠 × 100 𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑡𝑒𝑠𝑡𝑠

Eq. ( 3-6)

The formulation of this factor is drawn directly from the British Columbia Water Quality Index. c. Factor 3 (F3) This factor is called amplitude and represents the amount by which the failed test values do not meet their objectives, and is calculated in three steps: The number of times by which an individual concentration is greater than (or less than, when the objective is a minimum) the objective is termed an excursion and is expressed as follows. When the test value must not exceed the objective: 𝑓𝑎𝑖𝑙𝑒𝑑 𝑡𝑒𝑠𝑡 𝑣𝑎𝑙𝑢𝑒𝑠𝑖 Eq. ( 3-7) )−1 𝑜𝑏𝑗𝑒𝑐𝑡𝑖𝑣𝑒𝑗 For the cases in which the test value must not fall below the objective: 𝑒𝑥𝑐𝑢𝑟𝑠𝑖𝑜𝑛1 = (

𝑓𝑎𝑖𝑙𝑒𝑑 𝑡𝑒𝑠𝑡 𝑣𝑎𝑙𝑢𝑒𝑠𝑗 𝑒𝑥𝑐𝑢𝑟𝑠𝑖𝑜𝑛1 = ( )−1 𝑜𝑏𝑗𝑒𝑐𝑡𝑖𝑣𝑒𝑖

[ 38 ]

Eq. ( 3-8)

Chapter Three

Theoretical Background

The collective amount by which individual tests are out of compliance is calculated by summing the excursions of individual tests from their objectives and dividing by the total number of tests (those which do and do not meet their objectives). This variable, referred to as the normalized sum of excursions, or 𝑛𝑠𝑒, is calculated as: ∑𝑛𝑖=1 𝑒𝑥𝑐𝑢𝑟𝑠𝑖𝑜𝑛𝑖 Eq. ( 3-9) 𝑛𝑠𝑒 = 𝑛𝑜 𝑜𝑓 𝑡𝑒𝑠𝑡𝑠 𝐹3 is then calculated by an asymptotic function that scales the normalized sum of the excursions from objectives (𝑛𝑠𝑒) to yield a range between 0 and 100. The CWQI is finally calculated as: 𝑛𝑠𝑒 Eq. ( 3-10) 𝐹3 = ( ) 0.01 𝑛𝑠𝑒 + 0.01 The factor of 1.732 arises because each of the three individual index factors can range as high as 100. This means that the vector length can reach √1002 + 1002 + 1002 = √30,000 =173.3 as a maximum. Division by 1.732 brings the vector length down to 100 as a maximum (Abbasi & Abbasi, 2012). For each indicator, the grading scale followed the "ranking" scale recommended by water quality index standard. That scale also uses five categories or levels that corresponds to specific levels of water quality impairment, which is shown in Table (3-1). The Water Quality Index is calculated as the "grade point average" of the component indicators, and is reported as a Grade (i.e., A-F) and a Score. √𝐹12 + 𝐹22 + 𝐹32 𝑊𝑄𝐼 = 100 − 173.3

Eq. ( 3-11)

3.2.5 Determining a Lake's Trophic State (TSI) This method is called the trophic state index (TSI) or the Carlson index. It is new approach that developed by Robert E. Carlson in 1977 to present the trophic classification of lakes. Because of frustration in communication to the public both the current nature or status of lake and their future condition after restoration when the traditional trophic classification system is used (Carlson, 1977). Since lake water quality has so much natural variation, it is not possible to set water quality [ 39 ]

Chapter Three

Theoretical Background

standards for lakes. Therefore, it can be much more valuable to compare changes in one lake’s quality over the years or to compare between lakes, than to have simple limits for “good” and “bad” lakes. Table (3-1): Water quality rating as per different water quality index methods. (Abbasi & Abbasi, 2012) WQI Method

AWWQI

NSFWQI

CCMEWQI

OWQI

WQI Value >100 76-100 51-75 26-50 0-25 91-100 71-90 51-70 26-50 0-25 95-100 80-94 60-79 45-59 0-44 90-100 85-89 80-84 60-79 0-59

Rating of Water Quality Unsuitable for drinking Very poor water Poor water Good water Excellent Excellent water quality Good water quality Medium water quality Bad water quality Very bad water quality Excellent water quality Good water quality Fair water quality Marginal water quality Poor water quality Excellent water quality Good water quality Fair water quality Poor water quality Very poor water quality

The method that devised for rating lakes is called the Trophic State Index (TSI) or the Carlson Index (after the scientist who devised it) (Michaud, 1991). TSI can be calculated by using the Secchi disk depth, the total phosphorus concentration at the surface of the lake, or the chlorophyll-a concentration at the surface. Either one day’s values or, preferably, average values over the summer can be used. The equation used in this study to calculate TSI is based on the Secchi disk depth as given below: Eq. ( 3-12)

𝑇𝑆𝐼 = 60 − 14.41 (ℓ𝑛 𝑆𝐷)

Where SD is the Secchi depth in meters, and ℓ𝑛 stands for the natural log of a number. [ 40 ]

Chapter Three

Theoretical Background

Once the TSI have been calculated, it is easy to compare the results to other lakes or recalculate the value each year to see whether there appears to be any upward or downward trend in the lake. Again, because of the large natural variation for these parameters, it takes a number of years of data to determine whether any trend existed. It should be aware that calculation of the same TSI value with each of the parameters. In other words, if TSI is calculated using Secchi disk depth, the same result may not be obtained when calculating it with TP. According to the scientists who developed this index, chlorophyll-a is the best indicator to use if using data from the summer months, while TP is the best during the rest of the year. Of course, if Secchi disk data is the only parameter that used thus the TSI calculated from SD data (Carlson, 1977). Table (3-2) provides a comparison of each parameter and the resultant TSI (Carlson, 1977). The higher the TSI value, the older or more productive the lake is. Roughly speaking, lakes with TSI values between 0 and 40 are considered to be oligotrophic, those between 40 and 60 are mesotrophic, and those between 60 and 100 are eutrophic. A trophic state index is not the same as a water quality index. The term quality implies a subjective judgment that is best kept separate from the concept of trophic state. A major point of confusion with the existing terminology is that eutrophic is often equated with poor water quality. Excellent, or poor, water quality depends on the use of that water and the local attitudes of the people. The definition of trophic state and its index should remain neutral to such subjective judgments, remaining a framework within which various evaluations of water quality can be made. The TSI can be a valuable tool for lake management, but it is also a valid scientific tool for investigations where an objective standard of trophic state is necessary (Michaud, 1991).

[ 41 ]

Chapter Three

Theoretical Background

Table (3-2) : Comparison of trophic state index to water quality parameters and lake productivity (Carlson, 1977). Chl (ug/L)

TSI

SD (m)

TP (ug/L)

<30

<0.95

>8

<6

30-40

0.95-2.6

8-4

6-12

40-50

2.6-7.3

4-2

12-24

50-60

7.3-20

2-1

24-48

60-70

20-56

0.5-1

Attributes Oligotrophy: Clear water, oxygen throughout the year in the hypolimnion Hypolimnia of shallower lakes may become anoxic Mesotrophy: Wate r moderately clear; increasing probability of hypolimnetic anoxia during summer Eutrophy: Anoxic hypolimnia, macrophyte problems possible

48-96

Blue-green algae dominate, algal scums and macrophyte problems

70-80

56-155

0.25-0.5

96-192

Hypereutrophy: (li ght limited productivity). Dens e algae and macrophyte

>80

>155

<0.25

192-384

Algal scums, few macrophyte

Water Supply Water may be suitable for an unfiltered water supply.

Fisheries and Recreation Salmonid fisheries dominate Salmonid fisheries in deep lakes only

Iron, manganese, taste, and odour problems worsen. Raw water turbidity requires filtration.

Hypolimnetic anoxia results in loss of salmonids. Walleye may predominate Warm-water fisheries only. Bass may dominate.

Episodes of severe taste and odour possible.

Nuisance macrophyte, algal scums, and low transparency may discourage swimming and boating.

Rough fish dominate; summer fish kills possible

3.3 Remote Sensing Generally, Remote sensing (RS) is the art, science, and technology of observing an object, scene, or phenomenon by instrument-based technique at a distance without physical contact with the object of interest (Tempfli, et al., 2009). But remote sensing for earth observation, refers to obtaining information about objects or areas at the Earth’s surface without being in direct contact with the object or area. Most sensing devices record information about an object by measuring an object’s transmission of electromagnetic energy from reflecting and radiating surfaces (Lillesand, et al., 2007). Remote sensing includes aerial photography and satellite imagery. The acquainted information about the object

[ 42 ]

Chapter Three

Theoretical Background

needs to be comprehended as a clear representation in order to be recognized to produce important information. Satellite data systems can now measure phenomena that change continuously over time and cover large, often inaccessible areas. The general processes and elements involved in remote sensing of earth resources are grouped into two basic processes: 1. Data Acquisition The data acquisition involves the following elements: a. Energy sources. b. Propagation of energy through the atmosphere. c. Energy interactions with earth surface features. d. Airborne and/or space borne sensors. e. Resulting in the generation of Sensors data in pictorial and/or numerical form. f. Using sensors to record variations in the way earth surface features reflect and emit electromagnetic energy. 2. Data Analysis Processes The data analyses processes involve the following elements: a. Examining the data using various viewing and interpretation devices to analyse pictorial data, and/or a computer analysis of numerical sensor data. b. Presentation of the information in the form of maps, tables, or reports. c. Applying the resulting information to user’s decision. 3.3.1 Brief History of Remote Sensing The progressive growth of remote sensing can be observed from four views: 1. The development of the remote sensors and their resolution. 2. The development of means, which bear the sensors or transport them to the proper altitudes. 3. The development of the techniques used to transmit the data. 4. The abilities of showing that data and data enhancement. [ 43 ]

Chapter Three

Theoretical Background

The photography was born in 1839 with the public disclosue of the pioneering of photographic processes of Nicephore Niepce, William Henry Fox Talbot, and Louis Jacques Mande Daguerre. As early as 1840, Argo, Director of the Paris observatory, advocated the use of photography for topographic surveying. The first known Aerial photograph is taken in 1858 by a Persian photographer named Gaspard-Félix Tournachon. Known as “Nadar,” he used a tethered balloon to obtain the photograph over Val de Bièvre, near Paris. With this picture the era of earth observation and remote sensing had started. It was known as a first aerial photograph. Then James Wallace Black took the earliest existing aerial photograph from a balloon over Boston in 1860. By the 1900’s: photographic technology had improved to the point that cameras were smaller and faster films and lenses were available. At this time, aerial photographs had been taken from kites and pigeons. The airplane, which had been invented in 1903, was not used as a camera platform until 1909 when a biosphere motion pictures photographer accompanied Wilbur Wright and took the first aerial motion picture (Lillesand, et al., 2007). After World War II, the photo interpretation techniques developed in wartime

became

tile-established

procedures

for

civilian

applications.

Topographic mapping, geologic mapping and engineering mapping were routinely done using aerial photography, as they are today. The 1960’s produced a new age for remote sensing when the development of earth-orbiting satellites able to obtain high-altitude images of the earth surface independent of political boundaries. The simple camera in the first try of remote sensing was developed to be more complex sensors which are acquiring the spectral data in multi bands of the electromagnetic spectral in the recent remote sensors and collect data from hundred kilometres above. The first non-military satellite designed to collect information about the earth’s land resources was the Earth Resources Technology Satellite (ERTS-1) launched in 1972 by the US. The satellite was later renamed Landsat 1 and was [ 44 ]

Chapter Three

Theoretical Background

followed by Landsat 2-9. A network of ground receiving stations around the world received the data (Campbell & Wynne, 2011). The SPOT (System Probatoirc d’Observation de la Terre) satellite system was launched by the French space agency in 1986. It had a resolution of 10 and 20 meters depending on the spectral resolution. The orbital altitude of SPOT was 832 km. Many other countries use programs to develop their abilities in remote sensing like United State, Japan, Russia, India, etc. 3.3.2 Concepts of Remote Sensing System Remote sensing is a wide field and has many concepts. So, some spatial terms, which may be essential, are discussed here: 3.3.2.1 Energy Sources and Radiation Principle Electromagnetic (EM) energy is the link between the components of remote sensing system which refers to all energy that moves with the velocity of light in a harmonic wave pattern. The word harmonic implies that the component waves are equally and repetitively spaced in time which is shown in Fig. (3-1). The wave concept explains the propagation of electromagnetic energy, but this energy is detectable only in terms of its interaction with matter. In this interaction, electromagnetic energy behaves as though it consists of many individual bodies called photons that have such particle-like properties as energy and momentum (Tempfli, et al., 2009). In remote sensing, the most common way to categorize EM waves is by their wavelength location within EM spectrum as shown in Fig. (3-2). 3.3.2.2 Energy Interactions in the Atmosphere Earth’s atmosphere, which is an essential component of our very existence, is a nuisance for remote sensing of the earth’s surface. The radiation falling on earth and the reflected/emitted radiation from earth’s surface reaching the sensor are modified spectrally and spatially by the intervening atmosphere (Joseph, 2005). Particles and gases in the atmosphere can affect the incoming light and radiation. These effects are caused by the mechanisms of scattering and absorption (Levin, 1999). [ 45 ]

Chapter Three

Theoretical Background

Fig. (3-1): Two oscillating components of EM radiation: an electric and a magnetic field (Tempfli, et al., 2009).

Fig. (3-2): The diagram shows the wavelength and frequency ranges of EM radiation (Tempfli, et al., 2009). a. Scattering Scattering occurs when particles or large gas molecules present in the atmosphere interact with and cause the electromagnetic radiation to be redirected from its original path. Scattering takes place and depends on several factors including the wavelength of the radiation, the abundance of particles or gases, and the distance the radiation travels through the atmosphere. There are three types of scattering (Levin, 1999): [ 46 ]

Chapter Three

Theoretical Background

1. Rayleigh scattering: this type of scattering occurs when particles are very small compared to the wavelength of the radiation and causes shorter wavelengths of energy to be scattered much more than longer wavelengths. 2. Mie Scattering: this type of scattering occurs when the particles are just about the same size as the wavelength of the radiation. 3. Nonselective Scattering: this type of scattering occurs when the particles are much larger than the wavelength of the radiation and caused by water droplets and large dust particles. b. Absorption Absorption is the other main mechanism at work when electromagnetic radiation interacts with the atmosphere. This phenomenon causes molecules in the atmosphere to absorb energy at various wavelengths. Ozone, carbon dioxide, and water vapor are the three main atmospheric constituents which absorb radiation. Any effort to measure the spectral properties of a material through a planetary atmosphere, must consider where the atmosphere absorbs. These wavelengths cannot usually use when remotely measuring spectra through the Earth's atmosphere. Those areas of the spectrum which are not severely influenced by atmospheric absorption and thus, are useful to remote sensors, are called atmospheric windows (Levin, 1999). 3.3.2.3 Energy Interaction with Earth Surface Features When electromagnetic energy is incident on any given earth surface feature, three fundamental energy interactions with the feature are possible. Various fractions of the energy incident on the element are reflected, absorbed, and/or transmitted. This is illustrated in Fig. (3-3) for an element of the volume of a water body. Applying the principle of conservation of energy, the interrelationship between these three interactions can be stated as (Lillesand, et al., 2007): E 𝐼 (λ) = E 𝑅 (λ) + E 𝐴 (λ) + E 𝑇 (λ)

[ 47 ]

Eq. ( 3-13)

Chapter Three

Theoretical Background

Where E 𝐼 (λ) is incident energy, E 𝑅 (λ) is reflected energy, E 𝐴 (λ) is absorbed energy and E 𝑇 (λ) is transmitted energy.

Fig. (3-3): Interaction of Energy with the earth’s surface (Lillesand, et al., 2007).

The spectral reflectance of the material is different from one to another material. The spectral characteristics of various earth surface features do not remain static; they change with geographic location and time. In remote sensing, change detection techniques can be used to monitor these temporal changes. In addition to the spectral and related temporal consideration, the spatial characteristics (tone, colour, and texture) are also important to be considered. The spectral response from a water body is complex, as water in any quantity is a medium that is semi-transparent to electromagnetic radiation. The spectral response also varies according to the wavelength, the nature of the water surface (calm or wavy), the angle of illumination and observation of reflected radiation from the surface and bottom of shallow water bodies (Gibson, 2000). 3.3.2.4 Data Acquisition and Resolutions The detection of electromagnetic energy can be performed either photographically or electronically. The process of photography uses chemical reactions on the surface of a light sensitive film to detect energy variations within a scene. Electronic sensors generate an electrical signal that corresponds to the energy variations in the original scene and multispectral images. Because of the [ 48 ]

Chapter Three

Theoretical Background

various aspect of remote sensed images, the end user should be familiar with in selection the right data source. Imagery can be expressed in four dimension (Navulur, 2007): 1. Spectral resolution: is the number of bands in multispectral image. It refers to the bandwidth and depends on the ability of the sensor to collect data in multi bands. A multispectral data set can be envisaged as of consisting a number of stacked layers in which each element of image is associated with a digital number for each layer (Gibson, 2000). 2. Temporal resolution: is a measure of how often data are obtained for the same area. The temporal resolution varies from less than one hour for some systems to approximately 20 days for others (Gibson, 2000). 3. Radiometric resolution: is a measure of how many grey levels are measured between pure black (which could represent no reflectance from the surface) and pure white (Gibson, 2000). 4. Spatial resolution: is refer to the area covered on the ground by a single pixel and expressed in term of ground sampling distance (GSD). Spatial resolution is based on various factor such as sensor field of view (FOV), altitude at which the senor flown, and the number of detector in the sensor (Navulur, 2007). 3.3.2.5 Reference Data and the Global Positioning System (GPS) The acquisition of reference data involves collecting measurement or observation about the objects, areas, or phenomena that are being sensed remotely. These data can take any of a number of different forms and may be derived from a number of sources. It may also involve filed measurement of temperature and other physical and/or chemical properties of various feature. The geographic positions at which such field measurements are made often noted on a map base to facilitate their location in a corresponding remote sensing image. Usually, Global Positioning System (GPS) receiver are used to determine the precise geographic position of filed observation and measurement (Lillesand, et al., 2007).

[ 49 ]

Chapter Three

Theoretical Background

3.3.2.6 Satellite Date Imagery System Since 1972, Landsat satellites have continuously acquired space-based images of the Earth’s land surface, coastal shallows, and coral reefs. The Landsat program, a joint effort of the U.S. Geological Survey (USGS) and the National Aeronautics and Space Administration (NASA), was established to routinely gather land imagery from space. NASA develops the remote sensing instruments and spacecraft, then launches and validates the performance of the instruments and satellites. The USGS then assumes ownership and operation of the satellites, in addition to managing all ground reception, data archiving, product generation, and distribution. The result of this program is a long-term record of natural and human-induced changes on the global landscape as shown in Table (3-3). Table (3-3): Landsat sattelite mission dates with sensors specifications (USGS, 2015). Satellite

Launch

Decommissioned

Sensors

Landsat 1

July 23, 1972

January 6, 1978

MSS/RBV

Landsat 2

January 22, 1975

July 27, 1983

MSS/RBV

Landsat 3

March 5, 1978

September 7, 1983

MSS/RBV

Landsat 4

July 16, 1982

June 15, 2001

MSS/TM

Landsat 5

March 1, 1984

2013

MSS/TM

Landsat 6

October 5, 1993

Did not achieve orbit

ETM

Landsat 7

April 15, 1999

Operational

ETM+

Landsat 8

February 11, 2013

Operational

OLI/TIRS

Landsat 8 orbit the earth at 705 kilometres altitude. It makes a complete orbit every 99 minutes, complete about 14 full orbits each day, and cross every point on earth once every 16 days. Although each satellite has a 16-day full-earthcoverage cycle, their orbits are offset to allow 8-day repeat coverage of any Landsat scene area on the globe. The centrepiece of the observatory of Landsat 8 is: 1. The Operational Land Imager (OLI): The OLI collects data in nine shortwave bands; eight spectral bands at 30 m resolution and one panchromatic band at 15 m. OLI data products have a 16-bit range.

[ 50 ]

Chapter Three

Theoretical Background

2. The Thermal Infrared Sensor (TIRS): The TIRS captures data in two long wave thermal bands with 100 m resolution, and is registered to and delivered with the OLI data as a single product as shown in Table (3-4) (USGS, 2015). TIRS data products have a 30 m resolution and a 16-bit range. Table (3-4): Operational land imager (OLI) and thermal infrared sensor (TIRS) band designations (USGS, 2015). Spectral Bands

Wavelength (µm)

Resolution (m)

Band 1–Coastal/aerosol

0.43–0.45

30

Band 2–blue

0.45–0.51

30

Band 3–green

0.53–0.59

30

Band 4–red

0.64–0.67

30

Band 5–near IR

0.85–0.88

30

Band 6–SWIR 1

1.57–1.65

30

Band 7–SWIR-1

2.11–2.29

30

Band 8 –panchromatic

0.50–0.68

15

Useful in ‘sharpening’ multispectral images.

Band 9–cirrus

1.36–1.38

30

Useful in detecting cirrus clouds.

Band 10–TIRS 1

10.60–11.19

100

Useful for mapping thermal differences in water currents, monitoring fires and other night studies, and estimating soil moisture.

Band 11–TIRS 2

11.50–12.51

100

Same as band 10.

Use Increased coastal zone observations. Bathymetric mapping; distinguishes soil from vegetation; deciduous from coniferous vegetation. Emphasizes peak vegetation, which is useful for assessing plant vigor. Emphasizes vegetation slopes. Emphasizes vegetation boundary between land and water, and landforms. Used in detecting plant drought stress and delineating burnt areas and fire-affected vegetation, and is also sensitive to the thermal radiation emitted by intense fires. Used in detecting drought stress, burnt and fire affected areas, and can be used to detect active fires, especially at nighttime.

Landsat 8 data are used by government, commercial, industrial, civilian, military, and educational communities throughout the United States and worldwide. The data support a wide range of applications in such areas as global change research, agriculture, forestry, geology, resource management, geography, mapping, water quality, and coastal studies (USGS, 2015). Therefore, the Landsat 8 data have been used in the present study to determine the water quality of Dokan Lake water. 3.3.2.7 Data Image Process and Analysis When remote sensing data are available in digital format, digital processing and analysis may be performed using a computer. Digital processing and analysis [ 51 ]

Chapter Three

Theoretical Background

may be used to enhance data as a prelude to visual interpretation and can also be carried out to automatically identify targets and extract information completely without manual intervention by a human interpreter (CCRS, 2014). Here is some process that the satellite data image undergoes: A. Pre-Processing Methods Preprocessing commonly comprises a series of sequential operations, including geometric correction, atmospheric correction or image registration, normalization, masking (e.g., for clouds, water, irrelevant features), image rectification, and image re-sampling (Baboo & Thirunavukkarasu, 2014). Pre-processing operations, sometimes referred to as image restoration and rectification, are intended to correct for sensor- and platform-specific radiometric and geometric distortions of data (CCRS, 2014). The full preprocessing correction consists in fact of the following steps: 1. Geometric correction and ortho-correction: geometric corrections are made to correct the inconsistency between the location coordinates of the raw image data, and the actual location coordinates on the ground or base image. Several types of geometric corrections include system, precision, and terrain corrections. 2. Relative radiometric correction: radiometric correction is a prerequisite for generating high-quality scientific data, making it possible to discriminate between product artefacts and real changes in earth processes as well as accurately produce land cover maps and detect changes. There are two formulas that can be used to convert Digital Numbers (DNs) to radiance; the method selection depends on the scene calibration data available in the header file(s). One method uses the Gain and Bias (or Offset) values from the header file. The longer method uses the LMin and LMax spectral radiance scaling factors (Goslee, 2011): I. Gain and bias method: the first step in processing is to convert DNs to atsensor spectral radiance 𝐿, also called top-of-atmosphere radiance. The conversion coefficients are available in the metadata accompanying the [ 52 ]

Chapter Three

Theoretical Background

images. Whenever possible the metadata values should be used, as coefficients vary by platform and over time: Eq. ( 3-14)

𝐿𝜆 = 𝐺𝑎𝑖𝑛 ∗ 𝑄𝐶𝑎𝑙 + 𝑏𝑖𝑎𝑠 Where:

𝐿𝜆 = Spectral Radiance at the sensor's aperture in watts/ (meter squared * ster * µm) 𝐺𝑎𝑖𝑛 = Rescaled gain (the data product "gain" contained in the Level 1 product header or ancillary data record) in watts/(meter squared * ster * µm) 𝑄𝐶𝑎𝑙 = the quantized calibrated pixel value in DN 𝑏𝑖𝑎𝑠 = Rescaled bias (the data product "offset" contained in the Level 1 product header or ancillary data record) in watts/(meter squared * ster * µm) II.

Spectral radiance scaling method: coefficients are provided in one of three band-specific formats: gain and offset; 𝐺𝑟𝑒𝑠𝑐𝑎𝑙𝑒 (also called gain) and 𝐵𝑟𝑒𝑠𝑐𝑎𝑙𝑒 (bias); or radiances associated with minimum and maximum DN values (𝐿𝑚𝑎𝑥 and 𝐿𝑚𝑖𝑛). Any of the three can be used to convert from DN to at-sensor radiance. The formula used in this process is as follows:

𝐿𝜆 =

𝐿𝑀𝐴𝑋𝜆 − 𝐿𝑀𝐼𝑁𝜆 × (𝑄𝐶𝐴𝐿 − 𝑄𝐶𝐴𝐿𝑀𝐼𝑁) + 𝐿𝑀𝐼𝑁 𝑄𝐶𝐴𝐿𝑀𝐴𝑋 − 𝑄𝐶𝐴𝐿𝑀𝐼𝑁

Eq. ( 3-15)

Where: 𝐿𝜆 = the cell value as radiance. 𝑄𝐶𝐴𝐿= digital number. 𝐿𝑀𝐼𝑁𝜆 = spectral radiance scales to QCALMIN. 𝐿𝑀𝐴𝑋𝜆 = spectral radiance scales to QCALMAX. 𝑄𝐶𝐴𝐿𝑀𝐼𝑁 = the minimum quantized calibrated pixel value (typically = 1).

[ 53 ]

Chapter Three

Theoretical Background

𝑄𝐶𝐴𝐿𝑀𝐴𝑋 = the maximum quantized calibrated pixel value (typically = 255). 3. Conversion to top of atmosphere reflectance units: surface reflectance is generated by a new approach developed from a previous simplified radiometric (atmospheric + topographic) correction. In this conversion, new characteristics are added to enhance and automatize ground reflectance retrieval (Goslee, 2011): 𝜋𝐿𝜆 𝑑 2 𝜌𝜆 = 𝐸𝑆𝑈𝑁𝜆 𝑠𝑖𝑛𝜃 Where:

Eq. ( 3-16)

𝐿𝜆 = Radiance in units of W/ (m2 * sr * µm). 𝑑= Earth-sun distance, in astronomical units. 𝐸𝑆𝑈𝑁𝜆 = Solar irradiance in units of W/ (m2 * µm). 𝜃= Sun elevation in degrees. 4. Absolute atmosphere correction: this class of methods attempts to deduce values for atmospheric parameters from information contained within the image itself rather than using externally-measured data. Each image is treated on its own: 𝜋𝑑 2 (𝐿 − 𝐿ℎ𝑎𝑧𝑒 ) 𝜌= 𝑇𝑣 (𝐸𝑠𝑢𝑛 𝑐𝑜𝑠𝜃𝑧 𝑇𝑧 + 𝐸𝑑𝑜𝑤𝑛 )

Eq. ( 3-17)

The conversion of sensor radiance to atmospherically-corrected surface reflectance is described in Eq. (3-16). Absolute atmospheric correction methods use measurements or atmospheric simulation models to determine the parameters 𝑇𝑧 ,𝑇𝑣 ,𝐸𝑑𝑜𝑤𝑛 , and 𝐿ℎ𝑎𝑧𝑒 by Chavez in 1989. The major difference between the relative atmospheric correction methods is the procedure for estimating these values. With default parameter choices, this simplifies to the equation for at-sensor reflectance (Goslee, 2011). 5. Topographic normalization: the interaction between sun angle, surface slope and satellite position produces variations in surface reflectance [ 54 ]

Chapter Three

Theoretical Background

unrelated to true reflectance. Signal if located on opposite sides of a hill, the intent of topographic correction is to remove this source of variation. It leaves only the portion of the reflectance signal actually due to ground cover, thus converting the reflectance from an inclined surface (𝜌𝑇) to that from an equivalent horizontal surface (𝜌𝐻) (Goslee, 2011). B. Image Processing Methods The objective of the second group of image processing functions are grouped under the term of: 1. Image enhancement: it is solely to improve the appearance of the imagery to assist in visual interpretation and analysis (CCRS, 2014). 2. Image transformations: image transformations are operations similar in concept to those for image enhancement. However, unlike image enhancement operations which are normally applied only to a single channel of data at a time, image transformations usually involve combined processing of data from multiple spectral bands. Arithmetic operations (i.e. subtraction, addition, multiplication, division) are performed to combine and transform the original bands into "new" images which better display or highlight certain features in the scene (CCRS, 2014). Some of image transformations that can be made by the ENVI software and used in this study are listed below: a. Minimum Noise Fraction transform (MNF): use the minimum noise fraction (MNF) transforms to determine the inherent dimensionality of image data, to segregate noise in the data, and to reduce the computational requirements for subsequent processing. b. Principal Component Analysis (PCA): principal component analysis is to produce uncorrelated output bands, to segregate noise components, and to reduce the dimensionality of data sets. This is done by finding a new set of orthogonal axes that have their origin at the data mean and that are rotated so the data variance is maximized (Kottegoda & Rosso, 2008).

[ 55 ]

Chapter Three

Theoretical Background

c. Independent Component Analysis (ICA): independent components analysis on multispectral or hyperspectral datasets is to transform a set of mixed, random signals into components that are mutually independent. d. Band ratios: band ratios are to enhance the spectral differences between bands and to reduce the effects of topography. Dividing one spectral band by another produces an image that provides relative band intensities. e. Indices: spectral indices are combinations of surface reflectance at two or more wavelengths that indicate relative abundance of features of interest. Vegetation indices are the most popular type, but other indices are available for burned areas, man-made (built-up) features, water, and geologic features (Ray, 1994). Here some spectral indices that used in the present study:

I. Normalized Difference Vegetation Index (NDVI) The Normalized Difference Vegetation Index (NDVI) is an index of plant greenness or photosynthetic activity, and is one of the most commonly used vegetation indices. 𝑁𝐷𝑉𝐼 =

𝑁𝐼𝑅 − 𝑅 𝑁𝐼𝑅 + 𝑅

Eq. ( 3-18)

II. Modification of Normalised Difference Water Index (MNDWI) The modified NDWI (MNDWI) can enhance open water features while efficiently suppressing and even removing built-up land noise as well as vegetation and soil noise. The enhanced water information using the NDWI is often mixed with built-up land noise and the area of extracted water is thus overestimated. 𝑀𝑁𝐷𝑊𝐼 =

𝐺 − 𝑆𝑊𝐼𝑅1 𝐺 − 𝑆𝑊𝐼𝑅2

Eq. ( 3-19)

III. Land Surface Water Index (LSWI) Response of LSWI to rainfall, observed in the rate of increase in LSWI in the subsequent fortnights, shows that this index could be used to monitor the [ 56 ]

Chapter Three

Theoretical Background

increase in soil and vegetation liquid water content, especially during the early part of the season. 𝑁𝐼𝑅 − 𝑆𝑊𝐼𝑅1 𝑁𝐼𝑅 − 𝑆𝑊𝐼𝑅1 IV. Normalized Burn Ratio (NBR)

Eq. ( 3-20)

𝐿𝑆𝑊𝐼 =

The NBR was defined to highlight areas that have burned and to index the severity of a burn using Landsat imagery. The formula for the NBR is very similar to that of NDVI except that it uses near-infrared band 4 and the short-wave infrared band 7. 𝑁𝐼𝑅 − 𝑆𝑊𝐼𝑅2 𝑁𝐼𝑅 + 𝑆𝑊𝐼𝑅2 V. Moisture Stress Index (MSI)

Eq. ( 3-21)

𝑁𝐵𝑅 =

The MSI is a reflectance measurement that is sensitive to increases in leaf water content. As the water content of leaves in vegetation canopies increases Applications of the MSI include canopy stress analysis, productivity prediction and modelling, fire hazard condition analysis, and studies of ecosystem physiology. The MSI is inverted relative to other water VIs; higher values indicate greater water stress and less water content. 𝐺 − 𝑁𝐼𝑅 𝐺 + 𝑁𝐼𝑅 VI. Normalized Difference Moisture Index (NDMI) 𝑀𝑆𝐼 =

Eq. ( 3-22)

This index contrasts the near-infrared (NIR) band 4, which is sensitive to the reflectance of leaf chlorophyll content to the mid-infrared (MIR) band 5, which is sensitive to the absorbance of leaf moisture. 𝑅 − 𝑁𝐼𝑅 𝑅 + 𝑁𝐼𝑅 VII. Ratio-Vegetation-Index (RVI)

Eq. ( 3-23)

𝑁𝐷𝑀𝐼 =

The simplest vegetation index is the so called Ratio-Vegetation-Index. 𝑅𝑉𝐼 =

𝑁𝐼𝑅 𝑅

Eq. ( 3-24)

[ 57 ]

Chapter Three

Theoretical Background

VIII. Infrared Percentage Vegetation Index (IPVI) IPVI was first described by Crippen in 1990. Crippen found that the subtraction of the red in the numerator was irrelevant, and proposed this index as a way of improving calculation speed. It also is restricted to values between 0 and 1, which eliminates the need for storing a sign for the vegetation index values, and it eliminates the conceptual strangeness of negative values for vegetation indices. 𝑁𝐼𝑅 𝑅 + 𝑁𝐼𝑅 IX. Difference Vegetation Index (DVI) 𝐼𝑃𝑉𝐼 =

Eq. ( 3-25)

DVI is the Difference Vegetation Index, appears as VI (vegetation index) in (Lillesand, et al., 2007). Lillesand and Kiefer refer to its common use, so it has been certainly introduced earlier, but they have not given a specific reference. 𝑆𝑊𝐼𝑅 1 𝑅 3. Image Classification and Analysis 𝐷𝑉𝐼 =

Eq. ( 3-26)

Image classification and analysis operations are used to digitally identify and classify pixels in the data. Classification is usually performed on multi-channel data sets (A) and this process assigns each pixel in an image to a particular class or theme (B) based on statistical characteristics of the pixel brightness values (CCRS, 2014). 3.4 Statistical Process (Correlation and Regression) This section, focuses at the nature of statistical relationship between water quality and spatial response of satellite image for predicting the parameters. This process undergoes for plotting a scatter diagram which is an important preliminary step prior to undertaking a formal statistical analysis of the relationship between two variables before the two main processes. The first statistical process, correlations, examines linear relationships founds in the data that used to construct a scatterplot in a symmetric manner. Then the correlation process focuses primarily on association to establish any cause and effect. While the second part, regression models, considers the relationship of a response variable as determined [ 58 ]

Chapter Three

Theoretical Background

by one or more explanatory variables. It is often used as a tool to establish causality or designed to help make predictions. 3.4.1 Covariance and Correlation The covariance measures the linear relationship between a pair of quantitative measures 𝑥1 ,𝑥2 ,…, 𝑥𝑛 (𝑋) and 𝑦1 , 𝑦2 , ……, 𝑦𝑛 (𝑌) on the same sample of 𝑛 individuals. 𝑛

1 𝑐𝑜𝑣(𝑥, 𝑦) = ∑(𝑥𝑖 − 𝑥̅ )(𝑦𝑖 − 𝑦̅) 𝑛−1

Eq. (3-27)

𝑖=1

Where: (𝑦𝑖 − 𝑦̅): the deviation of each observation 𝑦𝑖 from the mean of the response variable. (𝑥𝑖 − 𝑥̅ ): the deviation of each observation 𝑥𝑖 from the mean of the predictor variable. (𝑥𝑖 − 𝑥̅ )(𝑦𝑖 − 𝑦̅): the product of the above two quantities. A positive covariance means that the terms (𝑥𝑖 − 𝑥̅ )(𝑦𝑖 − 𝑦̅) in the sum are more likely to be positive than negative. This occurs whenever the 𝑥 and 𝑦 variables are more often both above and below the mean in tandem than not. To avoid this disadvantage of the covariance, the data standardized before computing the covariance. To standardize the 𝑌 data, the mean must subtract from each observation then divide by the standard deviation, that is, by computing: 𝑧𝑖 =

(𝑦𝑖 − 𝑦̅) 𝑠𝑦

Eq. ( 3-28)

∑𝑛𝑖=1(𝑦𝑖 − 𝑦̅)2 𝑠𝑦 = √ 𝑛−1

Eq. ( 3-29)

Where 𝑠𝑦 is the sample standard deviation of 𝑌. It can be shown that the standardized variable 𝑍 in equation of 𝑧𝑖 has mean zero and standard deviation one. The 𝑋 standardized in a similar way by subtracting the mean 𝑧 from each observation 𝑧𝑖 then divide by the standard deviation 𝑠𝑥 . The covariance between [ 59 ]

Chapter Three

Theoretical Background

the standardized 𝑋 and 𝑌 data is known as the correlation coefficient between 𝑌 and 𝑋 (Chatterjee & Hadi, 2006) is given by: 𝑛

1 𝑥𝑖 − 𝑥̅ 𝑦𝑖 − 𝑦̅ 𝑐𝑜𝑣(𝑥, 𝑦) 𝑟(𝑥, 𝑦) = 𝑐𝑜𝑟(𝑥, 𝑦) = ∑( )( )= 𝑛−1 𝑠𝑥 𝑠𝑦 𝑠𝑥 𝑠𝑦

Eq. ( 3-30)

𝑖=1

The observations 𝑥 and 𝑦 are called uncorrelated if 𝑐𝑜𝑟(𝑥, 𝑦) = 0. A statistic that quantifies the strength of the linear relationship between the two variables is the correlation coefficient. Care must be taken lest correlation is confused with causation. Correlation may, but does not necessarily, indicate causation. Observing that 𝑦 increases when 𝑥 increases does not mean that a change in 𝑥 causes the increase in 𝑦. Both 𝑥 and 𝑦 may change as a result of change in a third variable, 𝑧 (Berthouex & Brown, 2002). There are three different types of correlation coefficients which are in use. One is called the Pearson product-moment correlation coefficient, Kendell rank correlation and the other called the Spearman rank correlation coefficient, which is based on the rank relationship between variables. The Pearson product-moment correlation coefficient is more widely used in measuring the association between two variables, and it has been taken into account in this study as well. The linear correlation coefficient is sometimes referred to as the Pearson product moment correlation coefficient in honour of its developer Karl Pearson. Given paired measurements (𝑋1 , 𝑌1 ), (𝑋2 , 𝑌2 ). . . (𝑋𝑛 , 𝑌𝑛 ), the Pearson productmoment correlation coefficient is a measure of association given by: 𝑟𝑃 =

∑𝑛𝑖=1(𝑋𝑖 − 𝑋̅)(𝑌𝑖 − 𝑌̅) √∑𝑛𝑖=1(𝑋𝑖 − 𝑋̅)2 ∑𝑛𝑖=1(𝑌𝑖 − 𝑌̅ )2

Eq. ( 3-31)

Where 𝑋̅ And 𝑌̅ are the sample mean of 𝑋1 ,𝑋2 , . . . , 𝑋𝑛 and 𝑌1 ,𝑌2 , . . . ,𝑌𝑛 , respectively (Berthouex & Brown, 2002). 3.4.2 Regression Models Generally, Regression analysis concerns the study of relationships between quantitative variables with the object of identifying, estimating, and validating the [ 60 ]

Chapter Three

Theoretical Background

relationship (Johnson & Bhattacharyya, 2010). Regression is a statistical technique to determine the relationship between two or more variables. It is one of the most widely used statistical tools because it provides simple methods for establishing a functional relationship among variables (Chatterjee & Hadi, 2006). There are three types of regression, simple, multiple and nonlinear regressions. The regressions that used in this study are simple and multiple regressions as explained below: 1. Simple Regression: Linear regression model, in its simplest (bivariate) form, shows the relationship between one independent variable (𝑋) and a dependent variable (𝑌), as in the formula below: Eq. ( 3-32)

𝑌 = 𝛽0 + 𝛽1 𝑋 + 𝑢

An error term (𝑢) captures the amount of variation not predicted by the slope and intercept terms. The regression coefficient (𝑅2 ) shows how well the values fit the data (Campbell & Campbell, 2008). 2. Multiple Regression: Multiple linear regression which is a linear regression model with one dependent variable and more than one independent variables. The multiple linear regression assumes that the response variable is a linear function of the model parameters and there are more than one independent variables in the model. The general form of the multiple linear regression model is as follows: 𝑦𝑖 = 𝛽0 + 𝛽1 𝑥1 + 𝛽2 𝑥2 + ⋯ + 𝛽𝑛 𝑥𝑛 + 𝜀𝑖

𝑖 = 1,2,3, … , 𝑛

Eq. ( 3-33)

Where 𝑦𝑖 is dependent variable, 𝛽0 , 𝛽1 , 𝛽2 , … , 𝛽𝑛 are regression coefficients, and 𝑥1 , 𝑥2 , … , 𝑥𝑛 are independent variables in the model. In the classical regression setting it is usually assumed that the error term 𝜀 follows the normal distribution with E (𝜀) = 0 and a constant variance Var (𝜀) = 𝜎 2 (Yan & Su, 2009).

[ 61 ]

Chapter Three

Theoretical Background

3.4.3 Selecting Variables Technique The motivation for selecting a subset of available predictors is that using all of the available predictors would unduly inflate Var(𝑌̂). This should be intuitively apparent, because 𝑌̂ is a linear combination of the ̂𝛽𝑖 , which of course are random variables that have their own variances. Thus, the more ̂𝛽𝑖 that are in the model, the greater will be Var(𝑌̂) (Rawlings, et al., 1998). Alternative variable selection methods have been developed that identify good (although not necessarily the best) subset models, with considerably less computing than is required for all possible regressions. The subset models are identified sequentially by adding or deleting, depending on the method, the one variable that has the greatest impact on the residual sum of squares. These methods are not guaranteed to find the best subset for each subset size, and the results produced by different methods may not agree with each other (Ryan, 2007). These methods are: 1. Forward Selection This selection of variable has been termed forward selection by Draper and Smith in 1998 which has one simple approach to start adding variables, one at a time, to form a model, and then stop when some threshold value is not met (Ryan, 2007). 2. Backward Elimination of Variables Elimination of variables chooses the subset models by start-back warding with the full model and then eliminating at each step the one variable elimination whose deletion will cause the residual sum of squares to increase the least. This will be the variable in the current subset model that has the smallest partial sum of squares. Without a termination rule, backward elimination continues until the subset model contains only one variable (Rawlings, et al., 1998). 3. Stepwise Regression Stepwise regression, is a combination of forward and backward procedures. The procedure starts in forward mode, but unlike forward selection, testing is performed after each regressor has been entered to see if any of the other [ 62 ]

Chapter Three

Theoretical Background

regressors in the model can be deleted. The procedure terminates when no regressor can be added or deleted. Stepwise regression can be used when there are so many variables (say, more than 50) which are impractical to implicitly or explicitly consider all possible subsets (Ryan, 2007). 3.4.4 Criteria for Choice of Subset Size in Regression Many criteria for choice of subset size have been proposed. These criteria are based on the principle of parsimony which suggests selecting a model with small residual sum of squares with as few parameters as possible. Most of the criteria are monotone functions of the residual sum of squares for a given subset size and, consequently, give identical rankings of the subset models within each subset size. However, the choice of criteria may lead to different choices of subset size, and they may give different impressions of the magnitude of the differences among subset models (Rawlings, et al., 1998). The commonly used criteria are coefficient 2 of determination (𝑅2 ), adjusted R2 (𝑅𝑎𝑑𝑗 ), residual mean square 𝑀𝑆𝑟𝑒𝑠 , Mallows

Cp Statistic, Akaike Information Criterion (AIC), Schwarz Bayesian Criterion (SBC), …, etc. The following criterion were used in the present study: a. Coefficient of Determination The coefficient of determination (𝑅2 ) is the proportion of the total (corrected) sum of squares of the dependent variable explained by the independent variables in the model (Weisberg, 2005): 𝑅2 =

∑(𝑥𝑖 − 𝑥̅ )(𝑦𝑖 − 𝑦̅) ∑(𝑥𝑖 − 𝑥̅ )2 × ∑(𝑦𝑖 − 𝑦̅)2

Where: 𝑅2 : The coefficient of determination. 𝑥𝑖 : 𝑖 𝑡ℎ Value of the independent variable 𝑥̅ : sample mean Value of the dependent variable 𝑥 𝑦𝑖 : 𝑖 𝑡ℎ Value of the dependent variable 𝑦̅: sample mean Value of the dependent variable 𝑦

[ 63 ]

Eq. ( 3-34)

Chapter Three

Theoretical Background

b. Adjusted Coefficient of Determination 2 The adjusted R2, which is labeled as (𝑅𝑎𝑑𝑗 ), is a rescaling of R2 by degrees

of freedom so that it involves a ratio of mean squares rather than sums of square (Dowdy, et al., 2004), which expressed by:

2 𝑅𝑎𝑑𝑗

(1 − 𝑅2 )(𝑛 − 1) =1− ( 𝑛 − 𝑝′ )

Eq. ( 3-35)

Where: 𝑅2 : The coefficient of determination. 𝑛 − 𝑝′ : Residual

c. Significance Levels for Choice of Subset Size F-to-enter and F-to-stay, or the equivalent “significance levels,” in the stepwise variable selection methods serve as subset size selection criteria when they are chosen so as to terminate the selection process before all subset sizes have been considered. Bendel and Afifi in 1977 compared several stopping rules for forward selection and showed that the sequential 𝐹𝑡𝑒𝑠𝑡 based on a constant “significance level” compared very favorably. The optimum “significance level to enter” varied between SLE = 0.15 and 0.25. Although not the best of the criteria they studied, the sequential F-test with SLE = 0.15 allowed one to do “almost best” when n − p ≤ 20. When n−p ≥ 40, the 𝐶𝑝 statistic was preferred over the sequential F-test but by a very slight margin if SLE = 0.20 were used. (Rawlings, et al., 1998).

[ 64 ]

CHAPTER FOUR EXPERIMENT AND PROCESS

Chapter Four

Experiment Works and Processes

Chapter Four Experiment Works and Processes 4.1 Introduction Theoretical study of any phenomena, often, requires verification of its results. Especially the one which depends on many variables and driving or developing its theoretical relations may be not easy. Such relations can be developed empirically based on experimental measurement, field and laboratory, results. Then the accuracy of applying these relations can be evaluated by using actual values measured in field or laboratory. The experimental works in this study include collecting samples, testing and analyzing the process which covers calculating the water quality indices and Trophic State Index as described in the sections of 4.5 and 4.6. 4.2 The Study Area Dokan Lake is located in Iraq on the Lower Zab River approximately 295 km north of Baghdad and 65 km southeast of Sulaimani city as shown in Fig. (41) and Fig. (4-2). It is surrounded by mountains of Sara and Quasar to the southeast, Assos to the northeast, Kosrat to the southwest and Barda Rash to the northwest. There is a gorge that extends from the Turba Village to Bemusha Village. This gorge separates the larger northern part of the lake in Bitwen plain from the small southern part of the lake near Dokan Dam. Villages and towns with agricultural lands, such as Rania, Chwar Qurna, and Qala Dza, surround the lake, with the densest populations and agricultural development to the northwest of the large lake (Ararat, et al., 2009). The dam is a concrete arch with gravity abutment blocks located in a narrow steep sided gorge incised in the limestone and dolomite bedrock. The crest length of the dam is 350 m and has a maximum height of 116 m. The reservoir impounded by Dokan Dam has a total design capacity at normal operating level (511 m.a.s.l) of 6.870 BCM, of which 6.140 BCM is live storage and 0.730 BCM being dead storage. The current storages will be less than this due to over 50 years of sedimentation. [ 65 ]

Chapter Four

Experiment Works and Processes

Area of Study

Fig. (4-1): Map of Northern Iraq showing the study area cited in (UN-ESCWA & BGR , 2013).

Fig. (4-2): Landsat 8 image represent the study area of Dokan Lake on October 24th, 2014 (left) and April 2nd, 2015 (right). [ 66 ]

Chapter Four

Experiment Works and Processes

The dam was designed by Binnie, Deacon and Gurley of the UK for the Republic of Iraq. Construction of the dam was started in 1954 and was completed in 1959. The main civil works contractor was Dumez-Balloz of France. Construction of the powerhouse of the five 80 MW turbine units was commissioned in 1979 (Ali & Salley, 2003) as cited in (Bilbas, 2014). The climate of this study area is characterized by cold and snowy winters and warm dry summers. On the plains, typical semi-arid climatic conditions prevail. Precipitation occurs from October to May, decreasing from the NE to SW. The existing data at Sulaimani meteorological station (in the middle of the territory) show an annual average precipitation of 659 mm for the period 19412011 (Bilbas, 2014). Dokan Dam observation station records data of flowrates, humidity, pressure, elevation, hydrology and others. Table (4-1) shows the climate data recorded by observation station of Dokan dam’s department at the two dates of sampling of the present study. There are also two seasonal streams from northwest of Dokan Lake which fed the lake on a seasonal basis; The Shauwr stream flow at elevation of 1000 m.a.s.l from Dola Raqa village in Shawur valley northwest of Rania district and travels about 33 km through the valley before join the Dokan Lake. The other seasonal stream is Qashan stream which flow at elevation of 1400 m.a.s.l as three different small streams from Shekh Wasan, Bellawa and Darash villages in Balisan valley, they join together at Zikha village to form Qashan stream which travels about 40 km before joins Dokan Lake in northwestern direction in Bitwen plain. The mean annual discharge variability and anomalies of the Dokan Lake is characterized by regular oscillation of wet and dry periods. The peak flows generally occur earlier in Spring (April), mainly as a result of lower snowfall levels and earlier snow-melt. Table (4-2) shows the water balance and hydrological data recorded by observation station of Dokan dam’s department station at the two dates of sampling of the present study.

[ 67 ]

Chapter Four

Experiment Works and Processes

Date

Air Temperature (Co)

Humidity (%)

Average Dew Point (Co)

Average Air Pressure (Mbar)

Speed (m/s)

Direction

Rainfall (mm)

Pan Evaporation (mm)

Ground Temperature (C0)

Sun Direction (Min)

Sun Radiation

Table (4-1): The Climate Data recorded by observation station of Dokan dam’s Department.

October, 24th,2014

13-22

48-83

10.9

955.5

2.1

159

0

1.8

21

516

134.3

April, 2nd, 2015

8-19

42-88

8.1

948.1

1.6

146

0

2.7

16

416

166.6

Wind

Table (4-2): The water balance and hydrological data recorded by observation station of Dokan dam’s department.

Date

Elevation (m)

Capacity (109 m3)

Change in Storage (m3/s)

Generation (MW-H)

Outflow (m3/s)

Inflow (m3/s)

Ferry Gauge (m)

Irrigation Outflow (m3/s)

October, 24th,2014

483.5

1.775

0

39

70

70

409.8

0

April, 2nd, 2015

488.8

2.055

236

21

49

285

409.5

14

[ 68 ]

Chapter Four

Experiment Works and Processes

4.3 Collecting and Analyzing in Situ Data Before collecting the samples, first, twenty sampling stations were established on the map of Dokan Lake. Sampling stations were selected in the study area at slightly equal distance from each other. The number of stations were selected in order to sufficiently estimate the quality of a water body of this size. Second, a GPS receiver (Garmin 62S) was used in a boat to locate each sample station in the lake for collecting the water samples. Twenty samples were taken from predetermined stations at two different dates. These twenty stations consisted ground reference data of water quality parameters that were collected and measurement of Secchi disk in a day which was nearly coincident with the sensor overpass. The sampling done in Autumn of 2014 and in Spring of 2015 in time of this study. Water samples were collected on 24 October of 2014 as an Autumn season and 02 April 2015 as a Spring season. At each station, two bottles of 250 ml samples (one is dark and other is transparent) were taken for DO and BOD test (Winkler method) as shown in Fig. (4-3). A bottle of 200 ml was taken for determining of E-choli test and two plastic bottles which are approximately 500 ml of water sample was collected for other parameters tests at 20 to 25 cm below the water surface where the temperature and sunlight have no effect on them. To prevent the deterioration of algae and other organic matter, the sample bottles were kept in an ice cooler until analyzed in the laboratory for E-coli and BOD. As mentioned earlier, the sampling proceeded in two different seasons of the year as discussed in next sections. 4.3.1 Autumn Season Sampling Conditions The first set of water samples were scheduled to be collected at the end of October due to acquisition date of Landsat 8 image which overpass and capture image at the same time and date where the water samples were collected from all twenty stations between 7:00 and 10:12 AM on 24 October 2014 although the time of capturing of Landsat 8 is 7:39 AM. [ 69 ]

Chapter Four

Experiment Works and Processes

Fig. (4-3): Water sampling from the lake and type of bottles used for samples. The short difference between the sampling and image capturing time does not affect the accuracy of the analyses results because the concentration change of water quality parameters in several hours is very rare. The stations of sampling on October 24th, 2014 in Dokan lake are shown in Fig. (4-4). The sampling day was sunny with a temperature in the range 13-22 C0 and a slight wind with a speed of 2.1 m/s coming from 1590 N. On the sampling day, inflow discharge was 70 m3/s same as the rate of outflow discharge which was less than the average daily because there was no precipitation in the catchment area. Table (4-3) lists the sampling time and stations geographical information in the lake.

[ 70 ]

Chapter Four

Experiment Works and Processes

Fig. (4-4): Station points of sampling on October 24th, 2014 in Dokan Lake.

[ 71 ]

Chapter Four

Experiment Works and Processes

Table (4-3): Sampling time and station points information on October 24th, 2014 in Dokan Lake. ID

Acquisition Date

Time

Longitudinal

Latitude

1

Friday, October 24, 2014

6:59:13 AM

44.959211

35.958623

2

Friday, October 24, 2014

7:10:10 AM

44.964604

35.971851

3

Friday, October 24, 2014

7:23:31 AM

44.976265

36.008782

4

Friday, October 24, 2014

7:36:47 AM

44.999336

36.013172

5

Friday, October 24, 2014

7:48:58 AM

44.990867

36.021374

6

Friday, October 24, 2014

7:54:25 AM

44.981172

36.030556

7

Friday, October 24, 2014

8:01:33 AM

44.974166

36.052853

8

Friday, October 24, 2014

8:06:58 AM

44.966473

36.052248

9

Friday, October 24, 2014

8:10:59 AM

44.952596

36.051073

10

Friday, October 24, 2014

8:19:33 AM

44.928456

36.077796

11

Friday, October 24, 2014

8:25:42 AM

44.943032

36.08977

12

Friday, October 24, 2014

8:31:33 AM

44.949309

36.097161

13

Friday, October 24, 2014

8:46:08 AM

44.955313

36.133124

14

Friday, October 24, 2014

8:53:56 AM

44.917151

36.124877

15

Friday, October 24, 2014

9:01:33 AM

44.880882

36.113177

16

Friday, October 24, 2014

9:13:40 AM

44.845987

36.13443

17

Friday, October 24, 2014

9:30:12 AM

44.898501

36.128671

18

Friday, October 24, 2014

9:40:16 AM

44.924108

36.152825

19

Friday, October 24, 2014

10:02:12 AM

44.929764

36.103832

20

Friday, October 24, 2014

10:11:57 AM

44.953490

36.062801

[ 72 ]

Chapter Four

Experiment Works and Processes

4.3.2 Spring Season Sampling Conditions The Second time of water samples were scheduled to be collected in early April due to the Landsat 8 image overpass and image availability the water samples were collected in same time of acquisition date on April 2 nd, 2015. The late winter / early Spring time of year was chosen for sampling because it is before deciduous plant species have fully re-vegetated therefore increasing the rate of erosion into lake thus lead to increase the area of water bodies in Dokan lake. Water samples were collected from all twenty stations between 6:45 AM and 10:15 AM on 2 April 2015. It was partly overcast but there is no cloud at the same area of Dokan Lake with a temperature in the range 9-19 C0 and a slight wind coming with the speed of 1.6 m/s from the146 N0. The stations of sampling on April 2nd, 2015 in Dokan lake are shown in Fig. (4-5). On the sampling day, inflow discharge was 285 m3/s which is higher than rate of outflow discharge and less than the average daily because there is no precipitation in the catchment area. Table (4-4) lists the sampling time and stations geographical information in the lake. 4.4 Water Quality Tests and Laboratory Analyses The collected water samples from the 20 point stations within Dokan Lake, in both dates, were tested and analyzed in more than one laboratory except water temperature, pH which were measured in the field, to specify some of the physical, chemical and biological characteristics for the water of these samples. These tests were achieved by chemical process or by using recent portable instruments for water quality tests. The measured water quality parameters are: 1. Secchi Disk Transparency (SDT.): Water Clarity and transparency are measure by 20 cm in diameter of Secchi disk in field. 2. Water Temperature (T.): the water temperature was measured in field by using laboratory thermometers (Liquid-in-glass thermometers). 3. Acidity (pH): The pH value was measured in field by using Trans Instrument (Senz pH (0~14.0 pH)) after calibrate it with buffer solutions with (4.01, 7.01 and 10.01) pH values. [ 73 ]

Chapter Four

Experiment Works and Processes

Fig. (4-5): Station points of sampling on April 2nd, 2015 in Dokan Lake.

[ 74 ]

Chapter Four

Experiment Works and Processes

Table (4-4): Sampling time and station points information on April 2nd, 2015 in Dokan lake. ID

Acquisition Date

Time

Longitudinal

Latitude

1

Thursday, April 2, 2015

6:52:24 AM

44.963411

35.96378

2

Thursday, April 2, 2015

6:55:19 AM

44.966693

35.974187

3

Thursday, April 2, 2015

7:08:45 AM

44.97677

36.009105

4

Thursday, April 2, 2015

7:20:57 AM

44.992224

36.018916

5

Thursday, April 2, 2015

7:24:34 AM

44.984426

36.02681

6

Thursday, April 2, 2015

7:29:55 AM

44.965368

36.031234

7

Thursday, April 2, 2015

7:34:37 AM

44.96829

36.039283

8

Thursday, April 2, 2015

7:38:38 AM

44.977725

36.045466

9

Thursday, April 2, 2015

7:50:15 AM

44.973249

36.060926

10

Thursday, April 2, 2015

7:56:55 AM

44.957427

36.06445

11

Thursday, April 2, 2015

8:03:43 AM

44.937887

36.065932

12

Thursday, April 2, 2015

8:14:03 AM

44.946238

36.081515

13

Thursday, April 2, 2015

8:41:23 AM

44.95459

36.111685

14

Thursday, April 2, 2015

8:49:27 AM

44.927076

36.108605

15

Thursday, April 2, 2015

8:59:05 AM

44.897742

36.106698

16

Thursday, April 2, 2015

9:09:49 AM

44.874922

36.136583

17

Thursday, April 2, 2015

9:19:44 AM

44.898911

36.161144

18

Thursday, April 2, 2015

9:33:09 AM

44.927273

36.160093

19

Thursday, April 2, 2015

9:54:09 AM

44.921886

36.083402

20

Thursday, April 2, 2015

10:14:08 AM

44.964749

36.082256

[ 75 ]

Chapter Four

Experiment Works and Processes

4. Turbidity: The turbidity of the samples was measured in lab by using Loviband Model: TB 300 IR and described with (NTU) units. 5. Total Suspend solid (TSS): The TSS concentration was measured by using. Laboratory gravimetric procedure where the solids from the water sample are filtered through a 47mm glass fiber filter, dried and weighed to determine the total nonfilter-able residue (TNR) of the sample reported as mg/L. 6. Total Dissolved Solid (TDS): The TDS concentration was measured by using 3540 Bench Combined Conductivity/pH Meter after calibrating it with buffer solutions with (1382 ppm). 7. Electrical conductivity (EC): The EC was measured by using 3540 Bench Combined Conductivity/pH Meter after calibrating it with buffer solutions with (1413μS/cm). 8. Nitrate (NO3): The NO3 concentration was measured by using APEL PD303UV. 9. Phosphate (PO4): The PO4 concentration was measured by using HANNA HI93751. 10. Dissolved Oxygen (DO) and Biochemical Oxygen Demand (BOD): The concentration for both Do and BOD, which is calculated from (DO5-DO0), is calculated in Lab. by Winkler method. 11. Biological characteristics: the test indicated the presence of E. coli in the water samples. Tests of water quality parameters correspond to the stations numbers on October 24th, 2014 and on April 2nd, 2015 are shown in Table (4-5) and Table (4-6).

[ 76 ]

Chapter Four

Experimental Works and Processes

Table (4-5): Tests of water quality parameters correspond to the stations numbers on October 24th, 2014. ID

pH

NO3 (mg/Ɩ)

EC (μhos/cm)

TDS (mg/l)

Temperature (Co)

Turbidity (NTU)

DO (mg/Ɩ)

PO4 (mg/Ɩ)

TSS ( mg/Ɩ)

M.P.N of E. coli

Secchi Disk (m)

BOD (mg/Ɩ)

NO3-N (mg/Ɩ)

TP (mg/Ɩ)

1

8.6

5.2

523

261.5

22.0

1.49

8.8

0.22

1.788

Positive

3.30

2.8

1.17

0.07

2

8.0

4.8

350

175

21.5

1.42

8.0

0.2

1.704

Positive

2.50

1.2

1.09

0.07

3

8.0

4.7

353

176.5

20.0

1.42

7.6

0.17

1.704

Positive

3.30

0.8

1.06

0.06

4

8.0

4.9

346

173

21.0

1.7

7.4

0.17

2.04

Positive

2.00

2.4

1.11

0.06

5

8.0

5.1

335

167.5

21.0

1.6

6.8

0.18

1.92

Positive

1.55

1.2

1.15

0.06

6

8.1

5.2

346

173

20.5

2.2

6.4

0.18

2.64

Positive

1.40

1.2

1.17

0.06

7

8.1

4.5

350

175

21.0

3

6.8

0.20

3.6

Positive

1.20

2.4

1.02

0.07

8

8.1

4.7

112.5

56.25

21.0

2.9

7.0

0.22

3.48

Positive

1.50

3.6

1.06

0.07

9

8.0

5.3

343

171.5

20.5

3.01

6

0.20

3.612

Positive

1.25

2.0

1.20

0.07

10

8.2

5.1

346

173

20.5

2.8

6.8

0.20

3.36

Positive

1.10

1.6

1.15

0.07

11

8.1

3.5

319

159.5

20.5

3

5.6

0.20

3.6

Positive

1.10

2.8

0.79

0.07

12

8.0

5.2

353

176.5

20.5

3.3

5.4

0.22

3.96

Positive

0.90

0.4

1.17

0.07

13

8.2

3.8

344

172

21.0

2.8

6.0

0.22

3.36

Positive

1.15

2.0

0.86

0.07

14

8.2

4.5

333

166.5

20.5

3.1

5.6

0.26

3.72

Positive

1.30

0.8

1.02

0.08

15

8.3

5.4

351

175.5

20.5

5

7.0

0.32

6.00

Positive

0.90

0.4

1.22

0.10

16

8.2

1.3

455

227.5

19.0

58

7.6

0.30

69.6

Positive

0.30

1.2

0.29

0.10

17

8.4

0.5

347

173.5

20.5

8.7

6.8

0.24

10.44

Positive

0.40

0.4

0.11

0.08

18

8.2

1.6

353

176.5

21.0

3.37

6.6

0.29

4.04

Positive

0.90

2.0

0.36

0.09

19

8.1

1.7

332

166

20.5

3.8

5.4

0.27

4.56

Positive

0.40

1.2

0.38

0.09

20

8.1

1.0

335

167.5

21.0

3.8

7.4

0.22

4.56

Positive

1.55

1.2

0.23

0.07

[ 77 ]

Chapter Four

Experimental Works and Processes

Table (4-6): Tests of water quality parameters correspond to the stations numbers on April 2nd, 2015. ID

pH

NO3 (mg/Ɩ)

EC (μhos/cm)

TDS (mg/l)

Temperature (Co)

Turbidity (NTU)

DO (mg/Ɩ)

PO4 (mg/Ɩ)

TSS (mg/Ɩ)

M.P.N of E. coli

Secchi Disk (m)

BOD (mg/Ɩ)

NO3-N (mg/Ɩ)

TP (mg/Ɩ)

1

8.7

3.286

413

206.5

14.0

2.60

2.4

0.244

3.12

Positive

3.40

0.1

0.74

0.08

2

8.6

3.413

387

193.5

14.0

3.90

2.4

0.244

4.68

Positive

3.25

0.1

0.77

0.08

3

8.5

3.249

391

195.5

14.0

2.01

2.5

0.243

2.412

Positive

3.10

0.2

0.73

0.08

4

8.6

3.522

365

182.5

14.5

2.65

2.3

0.248

3.18

Positive

2.85

0.3

0.80

0.08

5

8.6

0.156

380

190.0

14.0

1.25

2.6

0.52

1.5

Positive

2.80

0.5

0.04

0.17

6

8.6

3.322

421

210.5

14.0

2.65

2.7

0.287

3.18

Positive

2.65

0.5

0.75

0.09

7

8.6

3.177

384

192.0

14.0

2.20

2.4

0.27

2.64

Positive

2.60

0.3

0.72

0.09

8

8.6

3.05

374

187.0

14.0

1.80

2.6

0.239

2.16

Positive

2.50

0.3

0.69

0.08

9

8.6

3.268

389

194.5

14.5

3.20

2.8

0.41

3.84

Positive

2.45

0.6

0.74

0.13

10

8.6

3.286

403

201.5

14.5

2.20

2.7

0.423

2.64

Positive

2.20

0.6

0.74

0.14

11

8.7

3.140

412

206.0

15.0

1.90

2.3

0.331

2.28

Positive

1.90

0.2

0.71

0.11

12

8.7

3.340

402

201.0

14.0

1.40

2.8

0.253

1.68

Positive

2.40

0.2

0.75

0.08

13

8.6

3.249

383

191.5

14.5

1.16

2.4

0.253

1.392

Positive

2.40

0.3

0.73

0.08

14

8.7

3.504

397

198.5

14.5

1.40

2.3

0.27

1.68

Positive

2.10

0.1

0.79

0.09

15

8.5

2.142

430

215.0

15.0

1.70

2.2

0.26

2.04

Positive

1.45

0.2

0.48

0.08

16

8.5

1.307

416

208.0

16.0

2.60

2.5

0.25

3.12

Positive

1.30

0.5

0.30

0.08

17

8.5

1.833

390

190.0

16.0

2.11

2.3

0.244

2.532

Positive

0.37

0.8

0.41

0.08

18

8.6

3.122

389

189.5

17.0

1.50

2.6

0.218

1.8

Positive

1.40

0.4

0.70

0.07

19

8.6

3.013

355

177.5

15.0

1.80

2.3

0.266

2.16

Positive

1.60

0.2

0.68

0.09

20

8.6

3.322

428

214.0

15.0

1.16

2.9

0.211

1.392

Positive

2.10

0.5

0.75

0.07

[ 78 ]

Chapter Four

Experiment Works and Processes

4.5 Water Quality Index Calculation As mentioned earlier, there are different methods to calculate the water quality index. Next sections describe the methods used in the present study and calculation of water quality index for Dokan lake water. 4.5.1 NSF WQI Calculation The NSF WQI was estimated from calculation of a numerical value (Q-value) of nine parameters to provide a standardized method for comparing the water quality of various bodies of water in Dokan Lake. Calculating Q-value of parameters are:

 Dissolved Oxygen (DO%) o If DO% = (0) %: 𝑄𝑉. 𝐷𝑂 = 0 o If DO% = (0-100) %: 𝑄𝑉. 𝐷𝑂 = 1.01307189533432𝐸 − 09 ∗ 𝐷𝑂6 − 2.51036953231865𝐸 − 07 ∗ 𝐷𝑂5 + 0.0000180332453547649 ∗ 𝐷𝑂4 − 0.000198705094646812 ∗ 𝐷𝑂3 − 0.00852221706372802 ∗ 𝐷𝑂2 + 0.758129540663504 ∗ 𝐷𝑂 + 1.05563561210511 o If DO% = (100-140) %: 𝑄𝑉. 𝐷𝑂 = −0.0000125000000000264 ∗ 𝐷𝑂^4 + 0.00591666666668278 ∗ 𝐷𝑂^3 − 1.04875000000354 ∗ 𝐷𝑂^2 + 82.0083333336592 ∗ 𝐷𝑂 − 2281.00000001055 o If DO%= (140) %: 𝑄𝑉. 𝐷𝑂 = −0.0000125000000000264 ∗ 𝐷𝑂4 + 0.00591666666668278 ∗ 𝐷𝑂3 − 1.04875000000354 ∗ 𝐷𝑂2 + 82.0083333336592 ∗ 𝐷𝑂 − 2281.00000001055 o If DO% > (140) %: 𝑄𝑉. 𝐷𝑂 = 50

 Fecal Coliform o If No. F Coliform < (100,000): 𝑄𝑉. 𝐹. 𝐶𝑜𝑙𝑖𝑓𝑜𝑟𝑚 = 2

 pH o If pH < (2.0): 𝑄𝑉. 𝑝𝐻 = 0 o If pH = (2.0 - 6.0):

[ 79 ]

Chapter Four

Experiment Works and Processes

𝑄𝑉. 𝑝𝐻 = 0.291666666667652 ∗ 𝑝𝐻 4 + 5.75000000000546 ∗ 𝑝𝐻 3 − 34.7083333333449 ∗ 𝑝𝐻 2 + 85.2500000001091 ∗ 𝑝𝐻 − 71.0000000003465 o If pH = (6.0-9.0): 𝑄𝑉. 𝑝𝐻 = −0.858313448072295 ∗ 𝑝𝐻 6 + 37.530249732954 ∗ 𝑝𝐻 5 − 680.545427632809 ∗ 𝑝𝐻 4 + 6551.40354862855 ∗ 𝑝𝐻 3 − 35335.6700858718 ∗ 𝑝𝐻 2 + 101370.53913819 ∗ 𝑝𝐻 − 120990.176842687 o If pH= (9.0-12.0): 𝑄𝑉. 𝑝𝐻 = −0.666666666666671 ∗ 𝑝𝐻 3 + 27.0000000000001 ∗ 𝑝𝐻 2 − 361.333333333334 ∗ 𝑝𝐻 + 1602 o If pH= (12.0): 𝑄𝑉. 𝑃𝐻 = 0.666666666666671 ∗ 𝑝𝐻 3 + 27.0000000000001 ∗ 𝑝𝐻 2 − 361.333333333334 ∗ 𝑝𝐻 + 1602 o If pH < (2.0): 𝑄𝑉. 𝑝𝐻 = 0

 Biochemical Oxygen Demand (BOD5) o If BOD < (30.0) mg/l: 𝑄𝑉. 𝐵𝑂𝐷 = −2.34099938989894𝐸 − 06 ∗ 𝐵𝑂𝐷6 + 0.00022214828759981 ∗ 𝐵𝑂𝐷5 − 0.00784791328915802 ∗ 𝐵𝑂𝐷4 + 0.120683917172016 ∗ 𝐵𝑂𝐷3 − 0.509524519829029 ∗ 𝐵𝑂𝐷2 − 7.58366372497312 ∗ 𝐵𝑂𝐷 + 97.6255511774702 o If BOD = (30.0) mg/l: 𝑄𝑉. 𝐵𝑂𝐷 = −2.34099938989894𝐸 − 06 ∗ 𝐵𝑂𝐷6 + 0.00022214828759981 ∗ 𝐵𝑂𝐷5 − 0.00784791328915802 ∗ 𝐵𝑂𝐷4 + 0.120683917172016 ∗ 𝐵𝑂𝐷3 − 0.509524519829029 ∗ 𝐵𝑂𝐷2 − 7.58366372497312 ∗ 𝐵𝑂𝐷 + 97.6255511774702 o IF BOD > (30) mg/l: 𝑄𝑉. 𝐵𝑂𝐷 = 2

 Temperature o IF Temperature < (0) C0: 𝑄. 𝑉 𝑇𝑒𝑚𝑝𝑒𝑟𝑎𝑡𝑢𝑟𝑒 = −0.0000476731862306679 ∗ 𝑇𝐸𝑀𝑃.6 − 0.0031647877750629 ∗ 𝑇𝐸𝑀𝑃.5 − 0.0548298963109346 ∗ 𝑇𝐸𝑀𝑃.4 − 0.385317925573872 ∗ 𝑇𝐸𝑀𝑃.3 ^3 − 1.38790009116792 ∗ 𝑇𝐸𝑀𝑃.2 + 0.905611458261774 ∗ 𝑇𝐸𝑀𝑃. + 92.9967373993877 o IF Temperature = (0-10) C0:

[ 80 ]

Chapter Four

Experiment Works and Processes

𝑄. 𝑉 𝑇𝑒𝑚𝑝𝑒𝑟𝑎𝑡𝑢𝑟𝑒 = 0.00103008118360748 ∗ 𝑇𝐸𝑀𝑃. ^6 − 0.029249784530748 ∗ 𝑇𝐸𝑀𝑃.5 + 0.310431904086727 ∗ 𝑇𝐸𝑀𝑃.4 − 1.50289190054173 ∗ 𝑇𝐸𝑀𝑃.3 + 2.91690296295565 ∗ 𝑇𝐸𝑀𝑃.2 − 4.72537887119688 ∗ 𝑇𝐸𝑀𝑃. + 93.0095969367721 o IF Temperature = (10-30) C0: 𝑄. 𝑉 𝑇𝑒𝑚𝑝𝑒𝑟𝑎𝑡𝑢𝑟𝑒 = − 0.00169696969696953 ∗ 𝑇𝐸𝑀𝑃.3 + 0.181645021645031 ∗ 𝑇𝐸𝑀𝑃.2 − 6.75000000000038 ∗ 𝑇𝐸𝑀𝑃. + 95.1233766233793 o IF Temperature = (30) C0: 𝑄. 𝑉 𝑇𝑒𝑚𝑝𝑒𝑟𝑎𝑡𝑢𝑟𝑒 = −0.00169696969696953 ∗ 𝑇𝐸𝑀𝑃. ^3 + 0.181645021645031 ∗ 𝑇𝐸𝑀𝑃. ^2 − 6.75000000000038 ∗ 𝑇𝐸𝑀𝑃. + 95.1233766233793

 Total Phosphate (TP) o IF Total Phosphate (TP) < (0.1) mg/l: 𝑄. 𝑉 𝑇𝑃 = −3818.50533807828 ∗ 𝑇𝑃2 + 34.19928825622 ∗ 𝑇𝑃 + 98.22953736655 o IF Total Phosphate (TP) = (0.1 - 0.4) mg/l: 𝑄. 𝑉 𝑇𝑃 = 250.000000000004 ∗ 𝑇𝑃2 − 215.000000000001 ∗ 𝑇𝑃 + 83 o IF Total Phosphate (TP) = (0.4 – 3.0) mg/l: 𝑄. 𝑉 𝑇𝑃 = 1.7769523572 ∗ 𝑇𝑃^4 − 14.93199539525 ∗ 𝑇𝑃^3 + 48.76939791778 ∗ 𝑇𝑃^2 − 77.86535904492 ∗ 𝑇𝑃 + 60.88864660144 o IF Total Phosphate (TP) = (3.0) mg/l: 𝑄. 𝑉 𝑇𝑃 = 1.7769523572 ∗ 𝑇𝑃^4 − 14.93199539525 ∗ 𝑇𝑃^3 + 48.76939791778 ∗ 𝑇𝑃^2 − 77.86535904492 ∗ 𝑇𝑃 + 60.88864660144 o IF Total Phosphate (TP) < (3.0): 𝑄. 𝑉 𝑇𝑃 = 5

 Nitrate-Nitrogen (NO3-N) o IF Nitrate (NO3) < (2.0) mg/l: 𝑄. 𝑉 𝑁𝑂3 − 𝑁 = −12.321069721329 ∗ 𝑇2^4 + 54.0470740880701 ∗ 𝑇2^3 − 73.47623222119 ∗ 𝑇2^2 + 6.93214909692324 ∗ 𝑇2 + 97.816818775812 o IF Nitrate (NO3) < (2.0-20.0) mg/l: 𝑄. 𝑉 𝑁𝑂3 − 𝑁 = −0.000175270428069196 ∗ 𝑇2^4 + 0.00307174014734368 ∗ 𝑇2^3 + 0.21617336676968 ∗ 𝑇2^2 − 7.34174058882788 ∗ 𝑇2 + 65.8577957058549 o IF Nitrate (NO3) = (20.0) mg/l: 𝑄. 𝑉 𝑁𝑂3 − 𝑁 = −0.000175270428069196 ∗ 𝑇2^4 + 0.00307174014734368 ∗ 𝑇2^3 + 0.21617336676968 ∗ 𝑇2^2 − 7.34174058882788 ∗ 𝑇2 + 65.8577957058549\ [ 81 ]

Chapter Four

Experiment Works and Processes

o IF Nitrate (NO3) > (20.0) mg/l: 𝑄. 𝑉 𝑁𝑂3 − 𝑁 = 1

 Turbidity o IF Turbidity < (100.0) NTU: 𝑄. 𝑉 𝑁𝑇𝑈 = −5.14705882393731𝐸 − 10 ∗ 𝑁𝑇𝑈^6 + 1.17552790352554𝐸 − 07 ∗ 𝑁𝑇𝑈^5 − 6.89196832492911𝐸 − 06 ∗ 𝑁𝑇𝑈^4 − 0.000260476655824959 ∗ 𝑁𝑇𝑈^3 + 0.0461429247295655 ∗ 𝑁𝑇𝑈^2 − 2.52553304584762 ∗ 𝑁𝑇𝑈 + 96.9874537419573 o IF Turbidity = (100.0) NTU: 𝑄. 𝑉 𝑁𝑇𝑈 = −5.14705882393731𝐸 − 10 ∗ 𝑁𝑇𝑈^6 + 1.17552790352554𝐸 − 07 ∗ 𝑁𝑇𝑈^5 − 6.89196832492911𝐸 − 06 ∗ 𝑁𝑇𝑈^4 − 0.000260476655824959 ∗ 𝑁𝑇𝑈^3 + 0.0461429247295655 ∗ 𝑁𝑇𝑈^2 − 2.52553304584762 ∗ 𝑁𝑇𝑈 + 96.9874537419573 o IF Turbidity = (100.0) NTU: 𝑄. 𝑉 𝑁𝑇𝑈 = 5

After the Q-value is obtained, it is multiplied by a weighting factor, based on that test’s importance in water quality which is shown Table (4-7) (Oram, 2014). The nine resulting values are then added to compute the overall water quality index (WQI) as shown in Table (4-8) and Table (4-9).

Table (4-7): Water quality parameters and their weight. Parameter

Units

Weighting Factor

pH

pH units

0.12

Change in temperature

Degrees Co

0.11

DO

% saturation

0.18

BOD

mg/l

0.12

Turbidity

NTU

0.09

Total Phosphorus

mg/l P

0.11

Nitrate Nitrogen

mg/l NO3-N

0.10

E. coli

CFU/100 mL

0.17

Fecal Coliforms

CFU/100 mL

0.17

[ 82 ]

Chapter Four

Experimental Works and Processes

Table (4-8): Sensitivity of WQPs that used to calculate NSF water quality index for October 24th 2014. ID

pH

Change in Temperature

DO%

DO

BOD

Turbidity

Total Phosphorus

Nitrate Nitrogen

E. coli*

Fecal Coliforms*

Subtotal

Water Quality Index

Water Quality Rating

1

70

16

108.7

96

75

93

81

69

2

NM

61

61

MEDIUM

2

85

17

97.91

98

88

93

84

71

2

NM

65

65

MEDIUM

3

85

19

90.30

95

91

93

88

71

2

NM

66

66

MEDIUM

4

85

18

89.69

94

78

93

88

70

2

NM

64

64

MEDIUM

5

85

18

82.41

89

88

93

87

69

2

NM

64

64

MEDIUM

6

83

18

76.80

84

88

92

87

69

2

NM

63

63

MEDIUM

7

83

18

82.41

89

78

90

84

73

2

NM

62

62

MEDIUM

8

83

18

84.84

91

68

90

81

71

2

NM

61

61

MEDIUM

9

85

18

72.00

77

81

90

84

68

2

NM

60

60

MEDIUM

10

81

18

81.60

89

85

90

84

69

2

NM

63

63

MEDIUM

11

83

18

67.20

70

75

90

84

79

2

NM

59

59

MEDIUM

12

85

18

64.80

66

95

89

81

69

2

NM

60

60

MEDIUM

13

81

18

72.72

78

81

90

81

77

2

NM

61

61

MEDIUM

14

81

18

67.20

70

91

90

74

73

2

NM

59

59

MEDIUM

15

79

18

84.00

91

95

85

63

68

2

NM

61

61

MEDIUM

16

81

21

88.49

94

88

34

65

95

2

NM

60

60

MEDIUM

17

76

18

81.60

89

95

78

77

98

2

NM

64

64

MEDIUM

18

81

18

79.99

87

81

89

67

93

2

NM

62

62

MEDIUM

19

83

18

64.80

66

88

88

72

92

2

NM

60

60

MEDIUM

20

83

18

89.69

94

88

88

81

96

2

NM

66

66

MEDIUM

WF

0.12

0.11

0.18

0.12

0.09

0.11

0.10

0.17

NM

*Only use one microorganism, not Fecal coliforms and E. coli (NM: Not Measured). [ 83 ]

Chapter Four

Experimental Works and Processes

Table (4-9): Sensitivity of WQPs that used to calculate NSF water quality index for April 2nd, 2015. ID

pH

Change in Temperature

DO%

DO

BOD

Turbidity

Total Phosphorus

Nitrate Nitrogen

E. coli*

Fecal Coliforms*

Subtotal

Water Quality Index

Water Quality Rating

1

66

32

25.11

16

97

91

77

81

2

NM

51

51

MEDIUM

2

70

32

25.11

16

97

88

77

80

2

NM

51

51

MEDIUM

3

73

32

26.16

17

96

92

77

81

2

NM

52

52

MEDIUM

4

70

30

24.33

16

95

91

76

79

2

NM

51

51

MEDIUM

5

70

32

27.20

18

94

94

54

98

2

NM

51

51

MEDIUM

6

70

32

28.25

19

94

91

68

81

2

NM

50

50

MEDIUM

7

70

32

25.11

16

95

92

72

82

2

NM

51

51

MEDIUM

8

70

32

27.20

18

95

93

78

83

2

NM

52

52

MEDIUM

9

70

30

29.62

20

93

89

59

81

2

NM

49

49

BAD

10

70

30

28.56

19

93

92

58

81

2

NM

49

49

BAD

11

66

29

24.60

16

96

92

63

82

2

NM

49

49

BAD

12

66

32

29.29

19

96

94

75

80

2

NM

51

51

MEDIUM

13

70

30

25.39

17

95

94

75

81

2

NM

51

51

MEDIUM

14

66

30

24.33

16

97

94

72

79

2

NM

50

50

MEDIUM

15

73

29

23.53

15

96

93

74

89

2

NM

52

52

MEDIUM

16

73

27

27.33

18

94

91

76

95

2

NM

52

52

MEDIUM

17

73

27

25.14

17

91

92

77

92

2

NM

52

52

MEDIUM

18

70

25

29.04

19

95

93

81

82

2

NM

52

52

MEDIUM

19

70

29

24.60

16

96

93

72

83

2

NM

51

51

MEDIUM

20

70

29

31.02

21

94

94

82

81

2

NM

52

52

MEDIUM

WF

0.12

0.11

0.18

0.12

0.09

0.11

0.10

0.17

NM

*Only use one microorganism, not Fecal coliforms and E. coli (NM: Not Measured). [ 84 ]

Chapter Four

Experiment Works and Processes

4.5.2 Arithmetic Weight WQI Calculation In this method, Water quality indices were estimated form water quality parameters multiplied by a weighting factor and are then aggregated using simple arithmetic mean. As the AW-WQI was considered for human consumption or uses and the maximum permissible WQI for the drinking water which was taken as 100 score. For assessing the quality of water, firstly, the quality rating scale (𝑄𝑖 ) for each parameter were calculated from Recommended WHO standard of the water quality parameter for each water quality parameters. 𝑉𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑 for selected parameters are listed in Table (4-10). Then, after calculating the quality rating scale (𝑄𝑖 ), the relative unit weight (𝑊𝑖 ) was calculated by a value inversely proportional to the recommended standard (𝑆𝑖 ) for the corresponding parameter using calculation listed in both Table (4-11) for calculation of Autumn season models and Table (4-12) for Spring season model according to the Eq. (3-4) that mentioned before. Table (4-10): Recommended WHO standard of the water quality parameter. Water quality parameters

Unit

𝐕𝐬𝐭𝐚𝐧𝐝𝐚𝐫𝐝 for Water Uses Drinking

Aquatic

Irrigation

𝐕𝐢𝐝𝐞𝐚𝐥

Turbidity

(JTU)

1

N/A

N/A

0

Dissolved Oxygen (DO)

(mg/L)

N/A

5.5

N/A

14.6 (0 C0)

Maximum

8.5

9

N/A

7

(mg/L)

0.1

0.1

N/A

0

(mg/L)

5 ON

N/A

N/A

0

(mg/L)

200

0.062

N/A

0

(mg/L)

10

0.062

10

0

(mg/L)

1.5

N/A

1

0

C

15

N/A

N/A

0

(No./Dl)

2.2

NA

1000

0

PH Phosphorous(P) Dissolved Organic Carbon (DOC) Sodium (Na) Total Nitrate (NO3) Sodium (Na) Total Nitrate (NO3) Color Fluoride (F) Temperature Coliform

0

[ 85 ]

Chapter Four

Experimental Works and Processes

Table (4-11): Sensitivity of WQPs that used to calculate AW water quality index for October 24th, 2014. ID

QI PH

QI TEMP

QI DO

QI TURB.

QI P

QI TSS

QI TDS

QI NO3

WIQI

Water Quality Index

Water Quality Rating

1

106.6667

146.6667

60.41667

149

71.77289

357.6

52.3

52

1266.54

66.8359

Poor Water

2

66.66667

143.3333

68.75

142

65.24808

340.8

35

48.4

1042.971

55.03807

Poor Water

3

66.66667

133.3333

72.91667

142

55.46087

340.8

35.3

47

964.1643

50.87938

Poor Water

4

66.66667

140

75

170

55.46087

408

34.6

49

987.019

52.08544

Poor Water

5

66.66667

140

81.25

160

58.72327

384

33.5

51

995.5978

52.53814

Poor Water

6

73.33333

136.6667

85.41667

220

58.72327

528

34.6

52

1067.585

56.33693

Poor Water

7

73.33333

140

81.25

300

65.24808

720

35

45

1204.656

63.57026

Poor Water

8

73.33333

140

79.16667

290

71.77289

696

11.25

47

1032.856

54.50427

Poor Water

9

66.66667

136.6667

89.58333

301

65.24808

722.4

34.3

53

1200.532

63.35259

Poor Water

10

80

136.6667

81.25

280

65.24808

672

34.6

51

1182.291

62.39003

Poor Water

11

73.33333

136.6667

93.75

300

65.24808

720

31.9

35

1177.744

62.15009

Poor Water

12

66.66667

136.6667

95.83333

330

71.77289

792

35.3

52

1294.262

68.29881

Poor Water

13

80

140

89.58333

280

71.77289

672

34.4

38

1235.425

65.19392

Poor Water

14

80

136.6667

93.75

310

84.8225

744

33.3

45

1365.573

72.0619

Poor Water

15

86.66667

136.6667

79.16667

500

104.3969

1200

35.1

54

1734.885

91.55068

Very Poor Water

16

80

126.6667

72.91667

5800

97.87212

13920

45.5

13

7093.46

374.3251

Unsuitable for Drinking

17

93.33333

136.6667

81.25

870

78.29769

2088

34.7

5

1882.135

99.32113

Very Poor Water

18

80

140

83.33333

337

94.60971

808.8

35.3

16

1487.66

78.50449

Very Poor Water

19

73.33333

136.6667

95.83333

380

88.0849

912

33.2

17

1458.793

76.98114

Very Poor Water

20

73.33333

140

75

380

71.77289

912

33.5

10

1321.007

69.71014

Poor Water

VS

6.5

15

5

1

0.12

500

0.11

10

SUM

WI

0.153846

0.066667

0.2

1

8.333333

0.002

9.090909

0.1

18.94676

[ 86 ]

Chapter Four

Experimental Works and Processes

Table (4-12): Sensitivity of WQPs that used to calculate AW water quality index for April 2nd, 2015. ID

QI PH

QI TEMP

QI DO

QI TURB.

QI P

QI TSS

QI TDS

QI NO3

WIQI

Water Quality Index

Water Quality Rating

1

113.3333

93.33333

127.0833

260

79.60265

624

41.3

32.86

1351.991

71.34518

Poor Water

2

106.6667

93.33333

127.0833

390

79.60265

936

38.7

34.13

1458.108

76.94502

Very Poor Water

3

100

93.33333

126.0417

201

79.27641

482.4

39.1

32.49

1267.747

66.89958

Poor Water

4

106.6667

96.66667

128.125

265

80.90762

636

36.5

35.22

1323.931

69.86444

Poor Water

5

106.6667

93.33333

125

125

169.645

300

38

1.56

1931.852

101.9447

Unsuitable for Drinking

6

106.6667

93.33333

123.9583

265

93.63099

636

42.1

33.22

1479.554

78.07674

Very Poor Water

7

106.6667

93.33333

127.0833

220

88.0849

528

38.4

31.77

1354.986

71.50323

Poor Water

8

106.6667

93.33333

125

180

77.97145

432

37.4

30.5

1220.916

64.42826

Poor Water

9

106.6667

96.66667

122.9167

320

133.7586

768

38.9

32.68

1839.964

97.09571

Very Poor Water

10

106.6667

96.66667

123.9583

220

137.9997

528

40.3

32.86

1787.765

94.34115

Very Poor Water

11

113.3333

100

128.125

190

107.9856

456

41.2

31.4

1517.705

80.08996

Very Poor Water

12

113.3333

93.33333

122.9167

140

82.53882

336

40.2

33.4

1245.095

65.70422

Poor Water

13

106.6667

96.66667

127.0833

116

82.53882

278.4

38.3

32.49

1203.684

63.51897

Poor Water

14

113.3333

96.66667

128.125

140

88.0849

336

39.7

35.04

1288.188

67.97825

Poor Water

15

100

100

129.1667

170

84.8225

408

43

21.42

1318.233

69.56373

Poor Water

16

100

106.6667

126.0417

260

81.5601

624

41.6

13.07

1367.77

72.17782

Poor Water

17

100

106.6667

128.125

211

79.60265

506.4

38

18.33

1270.448

67.04209

Poor Water

18

106.6667

113.3333

125

150

71.1204

360

37.9

31.22

1139.719

60.1435

Poor Water

19

106.6667

100

128.125

180

86.77994

432

35.5

30.13

1278.074

67.44453

Poor Water

20

106.6667

100

121.875

116

68.83672

278.4

42.8

33.22

1129.716

59.6156

Poor Water

VS

6.5

15

5

1

0.12

500

0.11

10

SUM

WI

0.153846

0.066667

0.2

1

8.333333

0.002

9.090909

0.1

18.94676

[ 87 ]

Chapter Four

Experiment Works and Processes

4.5.3 CCMEWQI Calculation In the CCME WQI, the values of the three measures of variance from selected objectives for water quality are combined to create a vector in an imaginary objective exceedance space. In the index, objectives refer to Canada wide water-quality guidelines or site-specific water-quality objectives. CCME’s Water Quality Indices provide a convenient means of summarizing complex data and facilitating its communication. The SeQI, SoQI and WQI calculators are Microsoft Excel workbooks which contain macros. As it is downloaded with their manuals from CCME.ca website and used to calculate water quality index for CCMEWQI method. Table (4-13) and Table (4-14) show the output of SeQI, SoQI and WQI calculators for the data used in this study. Table (4-13): Calculation of CCMEWQI for Autumn Season on October 24th, 2014. ID

Date

Index Period

F1

F2

F3

CCME WQI

Type of WQI

Water Quality Rating

1

Oct. 2014

Autumn

28.6

28.6

34.4

69.4

60-90

GOOD

2

Oct. 2014

Autumn

28.6

28.6

33.6

69.7

60-90

GOOD

3

Oct. 2014

Autumn

28.6

28.6

31.1

70.6

60-90

GOOD

4

Oct. 2014

Autumn

28.6

28.6

27.6

71.7

60-90

GOOD

5

Oct. 2014

Autumn

28.6

28.6

29.7

71.1

60-90

GOOD

6

Oct. 2014

Autumn

28.6

28.6

24.3

72.8

60-90

GOOD

7

Oct. 2014

Autumn

28.6

28.6

22.2

73.4

60-90

GOOD

8

Oct. 2014

Autumn

28.6

28.6

24.8

72.6

60-90

GOOD

9

Oct. 2014

Autumn

28.6

28.6

22.2

73.4

60-90

GOOD

10

Oct. 2014

Autumn

28.6

28.6

23.1

73.1

60-90

GOOD

11

Oct. 2014

Autumn

28.6

28.6

22.2

73.4

60-90

GOOD

12

Oct. 2014

Autumn

42.9

42.9

23.5

62.5

60-90

GOOD

13

Oct. 2014

Autumn

28.6

28.6

25.2

72.5

60-90

GOOD

14

Oct. 2014

Autumn

28.6

28.6

28.1

71.6

60-90

GOOD

15

Oct. 2014

Autumn

14.3

14.3

31.2

78.5

60-90

GOOD

16

Oct. 2014

Autumn

14.3

14.3

29.4

79.4

60-90

GOOD

17

Oct. 2014

Autumn

14.3

14.3

23.3

82.2

60-90

GOOD

18

Oct. 2014

Autumn

28.6

28.6

30.1

70.9

60-90

GOOD

19

Oct. 2014

Autumn

42.9

42.9

27.4

61.6

60-90

GOOD

20

Oct. 2014

Autumn

28.6

28.6

21.9

73.5

60-90

GOOD

[ 88 ]

Chapter Four

Experiment Works and Processes

Table (4-14): Calculation of CCMEWQI for Spring Season on April 2nd, 2015. ID

Date

Index Period

F1

F2

F3

CCME WQI

Type of WQI

Water Quality Rating

1

Apr. 2015

Spring

42.9

42.9

36.8

59.1

60-90

FAIR

2

Apr. 2015

Spring

42.9

42.9

33.6

60.0

60-90

FAIR

3

Apr. 2015

Spring

42.9

42.9

38.8

58.4

60-90

MARGINAL

4

Apr. 2015

Spring

42.9

42.9

37.5

58.8

60-90

MARGINAL

5

Apr. 2015

Spring

42.9

42.9

56.9

52.0

60-90

MARGINAL

6

Apr. 2015

Spring

42.9

42.9

38.4

58.6

60-90

MARGINAL

7

Apr. 2015

Spring

42.9

42.9

40.2

58.0

60-90

MARGINAL

8

Apr. 2015

Spring

42.9

42.9

39.4

58.3

60-90

MARGINAL

9

Apr. 2015

Spring

42.9

42.9

44.5

56.6

60-90

MARGINAL

10

Apr. 2015

Spring

42.9

42.9

48.0

55.4

60-90

MARGINAL

11

Apr. 2015

Spring

42.9

42.9

45.8

56.2

60-90

MARGINAL

12

Apr. 2015

Spring

42.9

42.9

42.8

57.2

60-90

MARGINAL

13

Apr. 2015

Spring

42.9

42.9

46.9

55.7

60-90

MARGINAL

14

Apr. 2015

Spring

42.9

42.9

45.7

56.2

60-90

MARGINAL

15

Apr. 2015

Spring

42.9

42.9

43.3

57.0

60-90

MARGINAL

16

Apr. 2015

Spring

42.9

42.9

36.7

59.1

60-90

FAIR

17

Apr. 2015

Spring

42.9

42.9

39.4

58.3

60-90

MARGINAL

18

Apr. 2015

Spring

42.9

42.9

40.4

58.0

60-90

MARGINAL

19

Apr. 2015

Spring

42.9

42.9

42.5

57.3

60-90

MARGINAL

20

Apr. 2015

Spring

42.9

42.9

42.8

57.2

60-90

MARGINAL

4.5.4 Oregon Water Quality Index The Oregon Water Quality Index (OWQI) is a single number which expresses water quality by integrating measurements of eight carefully selected water quality parameters (temperature, dissolved oxygen, biochemical oxygen demand, pH, ammonia + nitrate nitrogen, total phosphates, total solids, Fecal coliform). To provide a simple and concise method for expressing the ambient water quality, the Q-values and sensitivity of WQI are calculated by using temperature, concentration of DO, BOD, and pH. This is calculated by these equations:  Temperature: o

If ( 𝑇 ≤ 11): 𝑄𝑇 = 100

[ 89 ]

Chapter Four o

Experiment Works and Processes

If (11 < 𝑇 ≤ 29): 𝑄𝑇 = (76.54 + 4.172 ∗ 𝑇 − 0.1623 ∗ (𝑇 2 ) − 2.0557 ∗ 0.001 ∗ (𝑇 3 ))

o

If (𝑇 > 29): 𝑄𝑇 = 10

 Dissolved Oxygen: o

If(𝐷𝑂 ≤ 3.3): QDO=10

o

If (3.3 < 𝐷𝑂 < 10.5): 𝑄𝐷𝑂 = −80.29 + 31.88 ∗ 𝐷𝑂 − 1.401 ∗ (𝐷𝑂2 ))

o

If (10.5 ≤ 𝐷𝑂): 𝑄𝐷𝑂 = 100

o

If (100 < DO < 275): 𝑄𝐷𝑂 = (100 ∗ 𝐸𝑋𝑃((𝐷𝑂 − 100) ∗ −1.19 ∗ 0.01))

o

If (𝐶2 > 275): 𝑄𝐷𝑂 = 10

 Biochemical Oxygen demand: o

If (𝐵𝑂𝐷 ≤ 8): 𝑄𝐵𝑂𝐷 = (100 ∗ 𝐸𝑋𝑃(𝐵𝑂𝐷 ∗ −0.1993))

o

If(8 < 𝐵𝑂𝐷), 𝑄𝐵𝑂𝐷 = 10

 Total Phosphate: o

If (𝑇𝑃 < 0.25): 𝑄𝑇𝑃 = (100 − 299.5 ∗ TP − 0.1384 ∗ TP 2 )

o

If (0.25 ≤ 𝑇𝑃): 𝑄𝑇𝑃 = (10)

 pH: 𝑄𝑇𝑃 = 100 ∗ 𝐸𝑋𝑃 ((𝑝𝐻 − 8) ∗ −0.5188)

Table (4-15) represents O-value, sensitivity and the calculation of OWQI for Autumn season model which modelled from water quality parameters of October 24th, 2014 sampling dates and Table (4-16) represents O-value, sensitivity and the calculation of OWQI for Spring season model which modelled from water quality parameters of April 2nd, 2015 sampling dates.

[ 90 ]

Chapter Four

Experiment Works and Processes

Table (4-15): OWQI calculation for Autumn Season on October 24th, 2014. ID

Temp

DO

BOD

TP

pH

Sensitivity

OQWI

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

67.88171 70.78451 78.6144 73.53986 73.53986 76.14931 73.53986 73.53986 76.14931 76.14931 76.14931 76.14931 73.53986 76.14931 76.14931 83.11765 76.14931 73.53986 76.14931 73.53986

91.76056 85.086 81.07624 78.90324 71.71176 66.35704 71.71176 74.221 60.554 71.71176 54.30264 51.00884 60.554 54.30264 74.221 81.07624 71.71176 69.09044 51.00884 78.90324

57.23297 78.72889 85.26211 61.98238 78.72889 78.72889 61.98238 48.79804 67.12592 72.69628 57.23297 92.33749 67.12592 85.26211 92.33749 78.72889 92.33749 67.12592 78.72889 78.72889

78.50331 80.45761 83.38904 83.38904 82.4119 82.4119 80.45761 78.50331 80.45761 80.45761 80.45761 78.50331 78.50331 74.59467 68.73161 70.68598 76.54899 71.66315 73.6175 78.50331

73.25087 100 100 100 100 94.94428 94.94428 94.94428 100 90.14416 94.94428 100 90.14416 90.14416 85.58672 90.14416 81.2597 90.14416 94.94428 94.94428

71.07758 81.45843 84.80047 76.71308 79.65982 78.13148 74.32666 68.69381 73.65777 77.444 67.96989 73.07504 71.98588 72.47475 78.11227 79.99151 78.74791 73.16204 70.23562 80.06071

GOOD GOOD GOOD GOOD GOOD GOOD GOOD FAIR GOOD GOOD FAIR GOOD GOOD GOOD GOOD GOOD GOOD GOOD GOOD GOOD

Table (4-16): OWQI calculation for Spring Season on April 2nd, 2015. ID

Temp

DO

BOD

TP

pH

Sensitivity

OQWI

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

97.49636 97.49636 97.49636 96.64337 97.49636 97.49636 97.49636 97.49636 96.64337 96.64337 95.66451 97.49636 96.64337 96.64337 95.66451 93.32305 93.32305 90.45965 95.66451 95.66451

10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10

98.02673 98.02673 96.0924 94.19623 90.51542 90.51542 94.19623 94.19623 88.7293 88.7293 96.0924 96.0924 94.19623 98.02673 96.0924 90.51542 85.26211 92.33749 96.0924 90.51542

76.15813 76.15813 76.25584 75.76726 49.18734 71.9563 73.6175 76.64671 59.93684 58.66646 67.65671 75.27868 75.27868 73.6175 74.59467 75.57183 76.15813 78.69874 74.00836 79.38275

69.54751 73.25087 77.15144 73.25087 73.25087 73.25087 73.25087 73.25087 73.25087 73.25087 69.54751 69.54751 73.25087 69.54751 77.15144 77.15144 77.15144 73.25087 73.25087 73.25087

21.73048 21.75142 21.76644 21.73879 21.48978 21.71169 21.7301 21.74479 21.61808 21.60578 21.67492 21.72198 21.73646 21.71616 21.75427 21.73916 21.726 21.73161 21.73244 21.74285

VERY POOR

[ 91 ]

VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR VERY POOR

Chapter Four

Experiment Works and Processes

4.6 Water Clarity and TSI Calculation Water clarity, or transparency, is commonly measured with a Secchi disk as shown in Fig. (4-6). The clarity of lake water is reduced by the presence of suspended sediment and Marl (CaCO3), bits of organic matter, free-floating algae, and zooplankton. As water-quality characteristics that are associated with water clarity can be computed for a Lake by relating sampled measurements of Secchidisk transparency (SDT) to satellite imagery such as Landsat 8 OLI.

Fig. (4-6): Measurement of Secchi disk in Dokan Lake.

Trophic state index (TSI) is an indicator of the biological productivity which can be calculated based on SDT measurements, Chl-a concentrations, and total phosphorus (TP) concentrations measured near the lake’s surface. Through this process, un-sampled Dokan lakes within the two Landsat satellite scenes and 20 point stations of sampling encompassing Dokan Lake can be translated into estimated TSI from computed SDT. TSI can be calculated by using Eq. (3-12) which has been discussed in chapter three and calculation shown in Table (4-17) and Table (4-18).

[ 92 ]

Chapter Four

Experimental Works and Processes

Table (4-17): TSI Calculation for samples of Autumn Season on October 24th, 2014. ID

SD (m)

TSI (SD)

TSI (SD)

SD (m)

1 2

3.3 2.5

42.8 46.8

40-50 40-50

2–4 2-4

3

3.3

42.8

40-50

2-4

4 5 6 7 8 9 10 11

2 1.55 1.4 1.2 1.5 1.25 1.1 1.1

50.01 53.68 55.15 57.37 54.16 56.78 58.63 58.63

50-60 50-60 50-60 50-60 50-60 50-60 50-60 50-60

1-2 1-2 1-2 1-2 1-2 1-2 1-2 1-2

12

0.9

61.52

60-70

0.5 - 1

13 14

1.15 1.3

57.99 56.22

50-60 50-60

1-2 1-2

15

0.9

61.52

60-70

0.5 - 1

16 17

0.3 0.4

77.35 73.2

70-80 70-80

0.25 -0.5 0.25 -0.5

18

0.9

61.52

60-70

0.5 - 1

19

0.4

73.2

70-80

0.25 -0.5

20

1.55

53.68

50-60

1-2

Attributes Mesotrophy: Water moderately clear; increasing probability of hypolimnetic anoxia during summer

Water Supply

Fisheries and Recreation

Iron, manganese, taste, and odor problems worsen. Raw water turbidity requires filtration.

Hypolimnetic anoxia results in loss of salmonids. Walleye may predominate

Eutrophy: Anoxic hypolimnia, macrophyte problems possible

Warm-water fisheries only. Bass may dominate.

Nuisance macrophyte, algal scums, and low transparency may discourage swimming and boating. Eutrophy: Anoxic hypolimnia, Warm-water fisheries only. Bass may macrophyte problems possible dominate. Nuisance macrophyte, algal scums, Blue-green algae dominate, algal scums Episodes of severe taste and low transparency may discourage and macrophyte problems and odor possible. swimming and boating. Hypereutrophy: (light limited productivity). Dense algae and macrophyte Blue-green algae dominate, algal scums and macrophyte problems

Episodes of severe taste and odor possible.

Blue-green algae dominate, algal scums and macrophyte problems

Episodes of severe taste and odor possible.

Nuisance macrophyte, algal scums, and low transparency may discourage swimming and boating.

Hypereutrophy: (light limited productivity). Dense algae and macrophyte Eutrophy: Anoxic hypolimnia, Warm-water fisheries only. Bass macrophyte problems possible may dominate.

[ 93 ]

Chapter Four

Experimental Works and Processes

Table (4-18): TSI calculation for samples of Spring Season on April 2nd, 2015. ID

SD (m)

TSI (SD)

TSI (SD)

SD (m)

1 2 3 4 5 6 7 8 9 10

3.4 3.25 3.1 2.85 2.8 2.65 2.6 2.5 2.45 2.2

42.37 43.02 43.70 44.91 45.16 45.96 46.23 46.80 47.09 48.64

40-50 40-50 40-50 40-50 40-50 40-50 40-50 40-50 40-50 40-50

2-4 2-4 2-4 2-4 2-4 2-4 2-4 2-4 2-4 2-4

Mesotrophy: Water moderately clear; increasing probability of hypolimnetic anoxia during summer

11

1.9

50.75

50-60

1-2

Eutrophy: Anoxic hypolimnia, macrophyte problems possible

12 13

2.4 2.4

47.38 47.38

40-50 40-50

14

2.1

49.31

40-50

2-4 2-4 2-4

Mesotrophy: Water moderately clear; increasing probability of hypolimnetic anoxia during summer

15 16

1.45 1.3

54.65 56.22

50-60 50-60

1-2 1-2

Eutrophy: Anoxic hypolimnia, macrophyte problems possible

Warm-water fisheries only. Bass may dominate.

17

0.37

74.33

70-80

0.25 -0.5

18

1.4

55.15

50-60

1-2

Warm-water fisheries only. Bass may dominate.

19

1.6

53.23

70-80

0.25 -0.5

Hypereutrophy: (light limited productivity). Dense algae and macrophyte Eutrophy: Anoxic hypolimnia, macrophyte problems possible Hypereutrophy: (light limited productivity). Dense algae and macrophyte

20

2.1

49.31

40-50

2-4

Attributes

Mesotrophy: Water moderately clear; increasing probability of hypolimnetic anoxia during summer

[ 94 ]

Water Supply

Fisheries and Recreation

Iron, manganese, taste, and odor problems worsen. Raw water turbidity requires filtration.

Hypolimnetic anoxia results in loss of salmonids. Walleye may predominate

Warm-water fisheries only. Bass may dominate. Iron, manganese, taste, and odor problems worsen. Raw water turbidity requires filtration.

Iron, manganese, taste, and odor problems worsen. Raw water turbidity requires filtration.

Hypolimnetic anoxia results in loss of salmonids. Walleye may predominate

Hypolimnetic anoxia results in loss of salmonids. Walleye may predominate

Chapter Four

Experimental Works and Processes

4.7 Remote Sensing Processing The remote sensed data image which used in this study include two Landsat-8 OLI images (path: 169 and row: 35) acquired on 24 October 2014 as an (ID: LC81690352014297LGN00)

and

04

April

2015

as

an

(ID:

LC81690352015092LGN00). The images were downloaded from the USGS, Earth Explorer (EE) website. Both images in October of 2014 and April of 2015 contain small pockets of clouds, which cover about 0.23% and 7.29% of the image respectively and they are less than the minimum removing percentage (9%) of full scene of satellite image (USGS, 2015). The date of acquisition has been selected in such way that no sample locations in the lake were impacted. So there is no need of haze of cloud removal. ENVI software has been used for geospatial analysis and spectral image processing of Landsat 8 OLI data imagery. Fig. (4-7) displays both the original Landsat 8 OLI October and April full scenes downloaded from USGS website.

Fig. (4-7): Landsat 8 OLI image full scene RGB true captured on October 24 th, 2014 (left) and April 2nd, 2015 (right). 4.7.1 Pre-Processing (Atmospheric Correction) After selection of date of acquisition and downloading images, the images should be imported to geometrically corrected and prepared for atmospheric [ 95 ]

Chapter Four

Experimental Works and Processes

correction which, involve removing the effects of clouds and aerosols from a radiance image. The result is an apparent surface reflectance image, which can be used to extract accurate spectral information from features on the Earth's surface. Atmospheric correction is carried out to minimize these atmospheric effects through correcting DN values to radiance. Radiance is the amount of radiation coming from an area. To derive a radiance image from an uncalibrated image, a gain and offset must be applied to the pixel values. These gain and offset values are typically retrieved from the image's metadata or received from the data provider. ENVI software provides a tool called radiometric calibration undertakes this process for many data products that are distributed with calibration gain and offset values in the metadata. Finally, radiance is converted to Surface Reflectance (SR) which represents the reflectance of the surface of the Earth as shown in Fig. (4-8) and Fig. (4-9). Therefore, clouds and other atmospheric components do not affect the surface reflectance spectra. The SR reflectance is unlike the Top-ofAtmosphere (TOA) reflectance which is the reflectance measured by a spacebased sensor flying higher than the earth's atmosphere. These TOA reflectance values includes contributions from clouds and atmospheric aerosols and gases.

Fig. (4-8): Landsat 8 OLI image on October 24th, 2014, atmospheic and radiometric correction, (left) before pre-procccesing, (right) after pre-processing. [ 96 ]

Chapter Four

Experimental Works and Processes

Fig. (4-9): Landsat 8 OLI image on April 2nd, 2015, atmospheic and radiometric correction, (left) before pre-procccesing, (right) after pre-processing. 4.7.2 Image Subset and Classification Following atmospheric corrections, seven bands of Landsat 8 OLI for each image were extracted as a layer stacking of multispectral bands image. Each corrected image was subset to much small study area to be easy for the selection of ground control points (GCP) that used in correction and study area subset images. As images undergo unsupervised classification process, pixels were grouped based on the reflectance properties which are known as “clusters”. The ENVI software was used for generation many identified clusters to define the bands used by ISODATA image clustering algorithms. Each cluster with different land cover classes were identified and multiple clusters which represent a single land cover class merged into a land cover type in order to subset and mask out the wetted bodies from satellite data as illustrated in Fig. (4-10) and Fig. (4-11). To ensure the accuracy of the classification, an accuracy assessment of the land use / land cover map was carried out to compare certain pixels in classified subsets with reference pixels through using supervised classification which

[ 97 ]

Chapter Four

Experimental Works and Processes

determined by integrated approach based on technique which is used to improve the accuracy of predictive models by ENVI software.

Fig. (4-10): Landsat 8 OLI image on October 24th, 2014, after subset (left), unsupervised classification (middle), after extraction of wetted area (right).

Fig. (4-11 ): Landsat 8 OLI image on April 2nd, 2015, after subset (left), unsupervised classification (middle), after extraction of wetted area (right). 4.7.3 Image Transformation Finally, a new process to the satellite image (extracted studied area) is provided, which typically involve the manipulation of multiple bands of data. It [ 98 ]

Chapter Four

Experimental Works and Processes

is called image transformations whether it is from a single multispectral image or from two or more images of the same area acquired at different times. Before transformation, image enhancement methods are applied separately to each band of a multi-spectral image such as contrast enhancement, linear contrast stretch, and spatial filtering. Image transformation in this study creates an output dataset where each output band is a linear combination of all the input bands. The first transform is Independent Component Analysis method (ICA) which works well with hyperspectral data because it is more likely to treat sparse targets as important features, compared with the Principal Component Analysis (PCA) or Minimum Noise Fraction (MNF) methods. However, ICA method can take a significantly longer time to process. The next method is PCA which creates a number of PC bands, which are linear combinations of the original spectral bands that are uncorrelated. It can be calculated the same number of output PC bands as input spectral bands. The first PC band contains the largest percentage of data variance and the second PC band contains the second largest data variance, and so on. Final transformation method is MNF. The MNF transform is a linear transformation that uses separate PCA rotations to segregate noise in the data and to reduce the dimensionality of the original dataset. 4.7.4 Band Ratios, Combination and Spectral Indices Band Ratios which enhance the spectral differences between bands were used in the present study to increase and provide more independent variables for the models and to reduce the effects of topography. Dividing one spectral band by another produces an image that provides relative band intensities. The image enhances the spectral differences between bands. One of the band that have been used is Coastal Blues which is Landsat 8 OLI’s new band that not available in Landsat 7. It was used with other spectral bands whether as a combination or band ratios such as (C/B), (C/G), (C/R), (C/NIR), (B/C), (B/G), (B/R), (B/NIR), (G/C), (G/B), (G/R), (G/NIR), (R/C), (R/G), (R,B), (R/NIR), (NIR/C), (NIR/B), (NIR/G), (NIR/R), (B+C), (B+G), (B+R), (B+NIR), (G+C), (G+R), (G+NIR), [ 99 ]

Chapter Four

Experimental Works and Processes

(R+C), (R+NIR), (C+NIR), (B+G+C), (B+G+R), (B+G+NIR), (B+R+NIR), (B+R+C), (B+C+NIR), (G+C+R), (G+C+NIR), (R+C+NIR), (G+R+NIR), (B+G+R+NIR), (C+B+G+R), (C+B+G+NIR), (C+G+R+NIR), (C+B+R+NIR), (C+B+G+R+NIR),

(B/NIR+C),

(B/NIR+B),

(B/NIR+G),

(B/NIR+R),

(B/NIR+NIR), (B/R+C), (B/R+B), (B/R+G), (B/R+R), (B/R+NIR), (NIR/B+C), (NIR/B+B), (NIR/B+G), (NIR/B+R) and (NIR/B+NIR). Then the spectral indices are combinations of surface reflectance at two or more wavelengths that indicate relative abundance of features of interest. Vegetation indices (NDVI) are the most popular type, but other indices such as (LSWI), (MNDWI), (MSI), (NBR), (DVI), (IPVI), (NDMI), and (RVI) are available for water features which are used in this study. 4.8 Correlation and Regression Models The atmospherically corrected water masked Landsat 8 OLI images were used for water quality spatial analysis in this study. Twenty sampling points were located on the image based on the UTM coordinates determined with a GPS during water sampling and extracted the spectral band of used image that mentioned before as independent variables, in ArcGIS Desktop software, to be used in correlation and regression modeling. Only the single pixel value for each sampling station was extracted. A Pearson correlation matrix was first developed to determine the strengths of correlation between water quality concentrations and spectral bands. Multiple linear regression methods were subsequently used to further explore the correlation between water quality parameters and spectral bands. The spectral band values at each sample station were extracted from images for use as independent variables in bivariate regression models. Only the first seven bands that known as OLI were used for the analysis based on examination of the data. Independent variables include 14 surface reflectance bands, 7 PCA bands, 7 ICA bands,7 MNF bands, 20 band ratios, 26 band combinations and finally 9 indices. Dependent variables that were analyzed include WQPs such as (𝑁𝑂3− , PO4, TSS, TDS, EC, Turbidity, DO and BOD), indices of WQ (OWQI, CCME WQI, AW [ 100 ]

Chapter Four

Experimental Works and Processes

WQI and NSF WQI), and clarity of water by indication of SDT, and TSI. All statistical analysis in this study has been done using SPSS computer software. Regression models were chosen to estimate both WQPs, SDT, TSI and WQI in Dokan Lake based on correlation coefficients, coefficients of determinations. Maps have been produced for measured versus computed water quality values and visual interpretation of concentration of parameters. The regression equations would succeed for mapping the water quality parameters concentrations in Dokan Lake.

[ 101 ]

CHAPTER FIVE RESULT AND DISCUSSION

Chapter Five

Result and Discussion

Chapter Five Result and Discussion 5.1 Introduction Satellite image data undergoes several processing and transformation, therefore, by using statistical correlation and developing a model equation of water quality whether it’s an index or a concentration are estimated. The processes depend on the spectral reflectance characteristics of the water in OLI bands of the Landsat 8 spectral bands. In the present study, the values of water quality indices and some parameters obtained from the developed models were studied and compared to the measured field and laboratory data as explained in the next sections. 5.2 Estimating Water Quality Parameters Water quality parameter testing is an important part of environmental monitoring and evaluations. Most important parameters that affect the quality of water in the environment whether it is physical, chemical or biological factors not only affects aquatic life but also the surrounding ecosystem. Physical properties of water quality include pH, temperature, EC, TSS, TDS and Turbidity. Chemical characteristics involve parameters such as NO3, PO4, DO, and BOD. Biological indicators of water quality include E-coli and coliform. These parameters are relevant to surface water studies of the ocean, lakes and rivers. Water quality monitoring can help researchers to predict and learn from natural processes in the environment and determine human impacts on an ecosystem. It also assists in restoration projects or ensure environmental standards. The remote sensing techniques can be used to monitor water quality parameters using satellite image data. This can be provided by both spatial and temporal information that needed to monitor changes in water quality parameters for developing management practices to improve water quality. Therefore, in the present study, twelve parameters were selected as dependent variables to generate cartographic maps and estimation water quality for two different seasons for the [ 102 ]

Chapter Five

Result and Discussion

entire area of Dokan Lake. The first estimation models were generated in Autumn season and the second models were generated in Spring season which is before plant have been revegetated. The reflectance in water surface depends upon the sky reflection, the color of the chemical contents dissolved in the water or that of plants and animals at bottom of lake. In Autumn season, the apparent blue color of lake gives high reflectance in water surface. The blue color is pure water which results from blue light scattering by water molecules. This certain amount of light that be reflected back into air depending upon the angle at which the light strikes the water surface. The surface reflectance of Landsat 8 OLI bands for selected stations in Dokan Lake on October 24th, 2014 are shown in Fig. (5-1). After the image processing, the multiple linear regressions were used to establish the relationship between the water quality parameters as dependent variables and Landsat 8 OLI spectral data as independent variables. Spectral data of Autumn images and in situ measurement data on October 24th, 2014 of water quality parameters correlated and multiple linear regression models were developed for estimating the pH, EC, TDS, T, NO3, NO3-N, PO4, TP, TSS, NTU and DO water quality parameters as shown in Table (5-1). The Pearson correlation between WQPs and spectral bands of Landsat 8 OLI are shown in appendix A. Among these models, the most appropriate models with highest R 2 value were selected. The R2 value and coefficients values of regression equations for each water quality parameter of Autumn season shown in Table (5-1). As temperatures increase and melt winter ice, lakes in Spring experience warming at the surface that leads to stratification, or temperature layering which is known as Spring season. At this season whole lake mixing usually occurs just after ice melts, but the difference in temperature between the surface and the bottom soon prevents mixing of the two layers. Frequent Spring storms bring not only water to the lake, but an influx of nutrients from the landscape. This often leads to a series of Spring algae and zooplankton blooms that allow nutrients in the lake to be cycled up through the food chain. [ 103 ]

Chapter Five

Result and Discussion

COASTAL BLUE

BLUE

GREEN

RED

NIR

SWIR1

SWIR2

1600 1400

Reflectance

1200 1000 800 600 400 200 0 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

Station Points

Fig. (5-1): Surface reflectance of Landsat 8 OLI bands for selected stations in Dokan Lake on October 24th, 2014. Table (5-1): WQPs models for Autumn Season on October 24th, 2014. Parameter

Autumn Season Models*

No.

R2

1

𝑝𝐻 = 8.268 + 0.111 (𝐼𝐶𝐴 𝑆𝑊𝐼𝑅1)

0.464

2

𝑝𝐻 = 8.790 + 0.141 (𝐼𝐶𝐴 𝑆𝑊𝐼𝑅1) − 0.228 (𝐺/𝑅)

0.707

3

𝑝𝐻 = 10.142 − 0.635 (𝐺/𝑅) − 0.115 (𝐺: 𝐹𝑙𝑜𝑎𝑡)

0.798

4

𝑝𝐻 = 10.491 − 0.051 (𝐼𝐶𝐴 𝑆𝑊𝐼𝑅1) − 0.742 (𝐺/𝑅) − 0.147 (𝐺: 𝐹𝑙𝑜𝑎𝑡)

0.803

EC

1

𝐸𝐶 = 241.50 + 529.504 (𝑁𝐼𝑅/𝐶)

0.701

TDS

1

𝑇𝐷𝑆 = 120.75 + 264.752 (𝑁𝐼𝑅/𝐶)

0.701

T

1

𝑇 = 21.765 − 0.001 (𝐶 + 𝑅)

0.726

1

𝑁𝑂3 = 6.001 − 8.329 (𝑁𝐵𝑅)

0.735

2

𝑁𝑂3 = 4.678 − 10.400 (𝑁𝐵𝑅) + 0.004 (𝐵/𝑁𝐼𝑅 + 𝐵)

0.808

1

𝑁𝑂3 − 𝑁 = 1.159 − 1.880 (𝐿𝑆𝑊𝐼)

0.736

2

𝑁𝑂3 − 𝑁 = 0.772 − 2.445 (𝐿𝑆𝑊𝐼) + 0.001 (𝐵/𝑁𝐼𝑅 + 𝐵)

0.823

PO4

1

𝑃𝑂4 = 0.401 − 0.083 (𝐺/𝑅)

0.702

TP

1

𝑇𝑃 = 0.127 − 0.025 (𝐺/𝑅)

0.746

TSS

1

𝑇𝑆𝑆 = −16.641 + 0.046 (𝑅 + 𝑁𝐼𝑅)

0.967

2

𝑇𝑆𝑆 = −50.077 + 0.063 (𝑅 + 𝑁𝐼𝑅) + 16.708 (𝐶/𝑅)

0.998

1

𝑁𝑇𝑈 = −13.867 + 0.039 (𝑅 + 𝑁𝐼𝑅)

0.967

2

𝑁𝑇𝑈 = −42.564 + 0.052 (𝑅 + 𝑁𝐼𝑅) + 13.923 (𝐶/𝑅)

0.998

1

𝐷𝑂 = 10.160 − 0.709 (𝐶/𝑁𝐼𝑅)

0.738

2

𝐷𝑂 = 10.841 − 0.682 (𝐶/𝑁𝐼𝑅) − 0.002 (𝐵/𝑁𝐼𝑅 + 𝐵)

0.832

pH

NO3

NO3-N

NTU

DO

* Selected models (highest R2). [ 104 ]

Chapter Five

Result and Discussion

Thus in this study, for this season with a same process, a new multiple linear regression models were developed after the spectral data of Spring season’s images and in situ measurement data in April 2nd, 2015 of water quality parameters correlated and most appropriate models with highest R 2 value have been selected. The multiple linear regression equations were used for estimating the 𝑝𝐻, 𝐸𝐶, 𝑇𝐷𝑆, 𝑇, 𝑁𝑂3 , 𝑁𝑂3 − 𝑁, 𝑃𝑂4 , 𝑇𝑃, 𝑇𝑆𝑆, 𝑇𝑢𝑟𝑏𝑖𝑑𝑖𝑡𝑦, 𝐷𝑂 and 𝐵𝑂𝐷 water quality parameters. The R2 value for some water quality parameters models were decreased due to decreasing the reflectance of Landsat 8 OLI that captured on April 2 nd, 2015 and seasonal changes of water quality parameters of Dokan Lake as shown in Fig. (5-2). The reflectance reduction is due to the light penetration from water surface, intensity of incipient of light, angle of ray incidence and scattering and absorption light within the water. The penetration of light anywhere in the water is reduced either when the sun is away from zenith, that means sun elevation for October 24th, 2014 is 52.9582630 is more than the sun elevation of April 2 nd, 2015 which is 39.774230270. A certain amount of light will be reflected back into air depending upon the angle at which the light strikes the water surface. If the striking angle from the horizontal water surface is high, a low amount of light is reflected because it is penetrated to water. In this region, the sun may be directly overhead only during summer months but may be far from this position at other times of the year and as a result, the amount of photosynthesis also changes, being reasonably high in the summer and much lower in the winter. In the study area all the water quality parameters showed a wide variation in space and time. Temporal variations were due to seasonal influences mainly, the effect of rainfall. The data analyses revealed that most of the parameters showed a substantial decrease after winter season. The models of 𝑝𝐻, 𝐸𝐶, 𝑇, 𝑁𝑂3 − 𝑁, 𝑃𝑂4 , 𝑇𝑃, 𝑇𝑆𝑆 and 𝑇𝑢𝑟𝑏𝑖𝑑𝑖𝑡𝑦 are developed for both seasons, Autumn and Spring season while for 𝐵𝑂𝐷 and 𝑁𝑂3 Spring season models are developed only as shown in Table (5-2).

[ 105 ]

Chapter Five

Result and Discussion

COASTAL BLUE

BLUE

GREEN

RED

NIR

SWIR1

SWIR2

1200

Reflectance

1000

800 600 400 200 0 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

Station Points

Fig. (5-2): Surface reflectance of Landsat 8 OLI bands for selected station in Dokan Lake on April 2nd, 2015. Table (5-2): WQP, WQI & TSI models for Spring Season on April 2nd, 2015. Parameter

Spring Season Models*

No.

R2

1

𝑝𝐻 = 8.485 + 0.045 (𝐼𝐶𝐴 𝐶)

0.278

2

𝑝𝐻 = 8.575 − 0.002 (𝑃𝐶𝐴 𝑅)

0.246

1

𝐸𝐶 = 422.034 − 1080.365 (𝑆𝑊𝐼𝑅1: 𝐹𝑙𝑜𝑎𝑡)

0.214

1

𝑇 = 20.829 − 7.317 (𝐶/𝐺)

0.779

2

𝑇 = 20.881 − 7.755 (𝐶/𝐺) − 0.008 (𝑃𝐶𝐴 𝐵)

0.850

3

𝑇 = 14.630 + 0.508 (𝐼𝐶𝐴 𝐺)

0.786

4

𝑇 = 10.516 + 0.386 (𝐼𝐶𝐴 𝐺) + 2.933 (𝐺/𝐵)

0.862

1

𝑁𝑇𝑈 = 0.727 + 0.694 (𝐵/𝑅)

0.272

2

𝑁𝑇𝑈 = −4.223 + 1.412 (𝐵/𝑅) + 2.957 (𝐺/𝐶)

0.463

1

𝑇𝑆𝑆 = 0.872 + 0.833 (𝐵/𝑅)

0.272

2

𝑇𝑆𝑆 = −5.068 + 1.695 (𝐵/𝑅) + 3.548 (𝐺/𝐶)

0.463

NO3

1

𝑁𝑂3 = 4.043 − 2.068 (𝑅/𝐶)

0.553

NO3-N

1

𝑁𝑂3 − 𝑁 = 0.913 − 0.469 (𝑅/𝐶)

0.559

PO4

1

𝑃𝑂4 = 0.146 + 0.552 (𝑁𝐼𝑅/𝐶)

0.213

1

𝑇𝑃 = 0.089 + 0.016 (𝐼𝐶𝐴 𝑁𝐼𝑅)

0.323

2

𝑇𝑃 = 0.050 + 0.142 (𝑁𝐼𝑅/𝐵)

0.229

1

𝐵𝑂𝐷 = −1.463 + 1.286 (𝐺/𝐵)

0.419

2

𝐵𝑂𝐷 = −1.463 + 1.286 (𝐺/𝐵)

0.419

3

𝐵𝑂𝐷 = −1.037 + 1.062 (𝐺/𝐵) + 0.003 (𝑃𝐶𝐴 𝐵)

0.570

pH EC

T

NTU

TSS

TP

BOD

* Selected models (highest R2).

[ 106 ]

Chapter Five

Result and Discussion

5.2.1 pH, Acidity and Alkalinity In Autumn season, four multiple linear regression models were selected for estimation of pH water quality parameter as shown in Table (5-1). As it is clear from the results, the model with (ICA SWIR1), (G/R) and (G: Float) as independent variables have the highest R2 of 0.803. On the other hand, the Spring season pH models no longer correlated with the spectral band due to higher variability of the values. The best two models that selected have low R2 values. The model with (ICA C) as independent variable has R2 value of 0.278 which is greater than R2 value for the second model with (PCA C) as independent variable. Fig. (5-3) shows the measured and computed pH values for Autumn and Spring season s data. Furthermore, Fig. (5-4) shows the mapping values of the selected pH model with the highest R2 value. In Autumn season, the computed pH values were of high values ranged between 9.5 and 10 at inlet of the lake due to the effect of pollution that results in higher algal and plant. But these values reduced where it is away from inlet, toward Dokan Dam ranged from 8.20 to 8.70. In Spring season, the pH values decreased to the range from 8.45 to8.52 at the inlet to highest in center to be from 8.60 to 9.0 due to high rains with a low pH (acid rain) of winter season. 5.2.2 Electrical Conductivity (EC) For estimation of electrical conductivity in Dokan Lake, only one multiple linear regression model was selected with highest R2 value of 0.701. The correlation shows that EC values for Autumn season is only correlated with bandratios of (NIR/C) band of Landsat 8 OLI which is a new band that is not available in Landsat 7 ETM. However, for Spring season model the R2 value is 0.214 with the band-ratio of (NIR/C) as independent variable and this may be because of seasonal change of 𝐸𝐶 values and change in solar and reflectance data of satellite image. The EC values in Autumn season include an outlier value in station point 8, therefore, the value of this point was assumed as a missed value in developing the models in order to increase the R2 value as illustrated in Fig. (5-5) which shows the measured and computed values of 𝐸𝐶 for Autumn and Spring season s data. [ 107 ]

Chapter Five

Result and Discussion

The 𝐸𝐶 values from the selected models were generated and mapped in Dokan Lake as shown in Fig. (5-6). In Autumn Season, the computed EC values were of high values ranged between 400 to 650 𝜇ℎ𝑜𝑠/𝑐𝑚 at inlet of the lake and at the end of lake (upstream of Dokan Lake) due to temperature’s direct effect and sedimentation, but for the centre of the lake it ranged from 315 to 345 𝜇ℎ𝑜𝑠/𝑐𝑚. For this reason, decreasing in temperature increases water density which leads to increase in salinity. In fact, the change in water density due to the high salinity leads to increase the EC value. Therefore, the EC in Spring season are of lower values which ranged from 360380 𝜇ℎ𝑜𝑠/𝑐𝑚 near the rocks at bank of the Lake and it is increase to 400420 𝜇ℎ𝑜𝑠/𝑐𝑚 due to increase in depth toward the centre and decrease due to sedimentation in Spring season. It can be concluded that this variety of values effected by the runoff after the rainfall season and snow melting from the mountain (winter season) that surrounding Dokan Lake. 5.2.3 Total Dissolved Solids (TDS) For estimating the concentration of total dissolved solids in Dokan Lake one multiple linear regression model have been selected only with highest R2 value of 0.701. As in EC model, the TDS values for Autumn season is highly correlated with band-ratios of NIR to coastal blue band (NIR/C) of Landsat 8 OLI. Just like EC, The TDS value in Autumn season include an outlier value in station point 8 as illustrated in Fig. (5-7) which shows the measured and computed values of TDS for Autumn season data. The selected model was used for generating and mapping TDS concentration values in Dokan Lake as shown in Fig. (5-8). On the other hand, for Spring season there is no model obtained because of the change in seasonal variations and decreasing in conductivity, while they affected by average temperatures, they also affected by water flow which increases in this season. In Autumn season, the computed values of TDS were of high concentration ranged between 215 and 250 mg/l at the inlet of the lake and decrease toward the centre of the lake to be ranged from 155 to 170 mg/l due same reason of EC. Because [ 108 ]

Chapter Five

Result and Discussion

TDS has been calculated by multiplying a conductivity value by an empirical factor. Deriving TDS from conductivity is quicker and suited for both field measurements and continuous monitoring. Measured V. Autumn Season

Computed V. Autumn Season

pH Value

9.00 8.50 8.00 7.50 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

17

18

19

20

Station Points

Measured V. Spring Season

Computed V. Spring Season

pH Value

9.00 8.50 8.00 7.50 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Station Points

Fig. (5-3): Measured and computed pH values in Dokan Lake for Autumn and Spring Seasons.

Fig. (5-4): Computed pH values in Dokan Lake for Autumn Season (left) and Spring Season (right). [ 109 ]

Chapter Five

Result and Discussion Measured V. Autumn Season

Computed V. Autumn Season

EC (µhos/cm)

600.00 500.00

400.00 300.00 200.00 100.00 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

17

18

19

20

Station Points

Measured V. Spring Season

Computed V. Spring Season

EC (µhos/cm)

500.00 400.00 300.00 200.00 100.00 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Station Points

Fig. (5-5): Measured and computed EC values in Dokan Lake for Autumn and Spring Seasons.

Fig. (5-6): Computed EC values in Dokan Lake for Autumn Season (left) and Spring Season (right). [ 110 ]

Chapter Five

Result and Discussion Measured V. Autumn Season

Computed V. Autumn Season

300.000

TDS (mg/l)

250.000 200.000 150.000 100.000 50.000

0.000 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

Station Points

Fig. (5-7): Measured and computed TDS values in Dokan Lake for Autumn Season.

Fig. (5-8): Computed TDS values in Dokan Lake for Autumn Season on October 24th, 2014. [ 111 ]

Chapter Five

Result and Discussion

5.2.4 Temperature In previous studies, it was shown that temperature have high correlated with radiance of satellite image and many model have been developed for its estimation and predictions. In this study, temperature in Autumn season has been correlated with combination of coastal blue and red reflectance band of Landsat 8 OLI. For Autumn season, only one multiple linear regression model has been selected with R2 value of 0.726 for estimating water temperature values in Dokan Lake. For Spring season, four multiple linear regression models have developed. The best model with highest R2 of 0.862 is obtained with (ICA G) and (PCA B) as independent variables. Fig. (5-9) shows the measured and computed values of temperature for Autumn and Spring season s of the selected models. The model was used for generating and mapping the temperature values in Dokan Lake for Spring season as shown in Fig. (5-10). The most obvious reason for temperature change in lakes is the change in seasonal air temperature. In Autumn season, the computed T values were of low values, lower than 19𝐶 0 , at the inlet of the lake and increase toward Dokan Dam to be ranged from 20 to 21𝐶 0 due to the same reasons of EC. Because light deceases exponentially with depth in the water column, the sun can heat a greater proportion of the water in a shallow lake than in a deep lake and so a shallow lake can warm up faster and to a higher temperature. Other reason is that increasing of turbidity will also increase water temperature. The suspended particles absorb heat from solar radiation more efficiently than water then the heat is transferred from the particles to water molecules, increasing the temperature of the surrounding water while in Spring season is occurred vice versa from inlet of the lake to be higher than 17𝐶 0 toward Dokan Dam with a lower value of 13𝐶 0 . 5.2.5 Nitrate In this study, nitrogen compounds of water are represented in two form such as nitrate (𝑁𝑂3 ) and nitrate-nitrogen (𝑁𝑂3 − 𝑁). Two multiple linear regression models were developed for predicting each water quality parameter. For nitrate [ 112 ]

Chapter Five

Result and Discussion

the best model with highest R2 value of 0.808 was obtained when 𝑁𝑂3 values for Autumn season are correlated with (NBR) index and (B/NIR + B) of Landsat 8 OLI. In addition, for Autumn season the best nitrate-nitrogen model with highest R2 value of 0.823 was obtained when 𝑁𝑂3 − 𝑁 values are correlated with (LSWI) index and (B/NIR + B) of Landsat 8 OLI. On the other hand, for Spring season, each one of 𝑁𝑂3 and 𝑁𝑂3 − 𝑁 values are correlated with (R/C) with R2 values of 0.553 and 0.559 respectively. The predict values of both 𝑁𝑂3 and 𝑁𝑂3 − 𝑁 from the best models and measured values for Autumn and Spring season s are plotted together as shown in Fig. (5-11) and Fig. (5-12) respectively. Also, the best models have been used for generating and mapping 𝑁𝑂3 and 𝑁𝑂3 − 𝑁 values for Autumn and Spring season s as shown in Fig. (5-13) and Fig. (5-14) respectively. In Autumn season, the computed 𝑁𝑂3 and 𝑁𝑂3 − 𝑁 were of high values, 6.5 mg/l and 1.3 mg/l, at the inlet of the lake and decrease toward Dokan Dam to be 2.5 mg/l and 0.4 mg/l respectively. In Spring season, the values decreased due to seasonal factors such as dilution of rainfall, aeration, and others. The values of 𝑁𝑂3 and 𝑁𝑂3 − 𝑁 are ranged 3.5-1.0 mg/l and 0.8-0.2 mg/l from the inlet of the lake toward the dam respectively.

[ 113 ]

Chapter Five

Result and Discussion Measured V. Autumn Season

Computed V. Autumn Season

T (Co)

25.000 20.000 15.000 10.000 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

17

18

19

20

Station Points

Measured V. Spring Season

Computed V. Spring Season

18.00

T (Co)

16.00 14.00

12.00 10.00 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Station Points

Fig. (5-9): Measured and computed Temperature in Dokan Lake for Autumn and Spring Seasons.

Fig. (5-10): Computed Temperature in Dokan Lake for Autumn Season (left) and Spring Season (right). [ 114 ]

Chapter Five

Result and Discussion Measured V. Autumn Season

Computed V. Autumn Season

NO3 (mg/l)

8 6 4 2 0 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

17

18

19

20

Station Points

Measured V. Spring Season

Computed V. Spring Season

NO3 (mg/l)

4.00 3.00 2.00 1.00 0.00 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Station Points

Fig. (5-11): Measured and computed values of 𝑁𝑂3 for Autumn and Spring Seasons. Measured V. Autumn Season

Computed V. Autumn Season

NO3-N (mg/l)

1.500 1.000 0.500 0.000 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

17

18

19

20

Station Points

Measured V. Spring Season

Computed V. Spring Season

NO3-N (mg/l)

1.00 .80 .60 .40 .20 .00 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Station Points

Fig. (5-12): Measured and computed values of 𝑁𝑂3 − 𝑁 in Dokan Lake for Autumn and Spring Seasons. [ 115 ]

Chapter Five

Result and Discussion

Fig. (5-13): Computed 𝑁𝑂3 values in Dokan Lake for Autumn Season (left) and Spring Season (right).

Fig. (5-14): Computed 𝑁𝑂3 − 𝑁 values in Dokan Lake for Autumn Season (left) and Spring Season (right).

[ 116 ]

Chapter Five

Result and Discussion

5.2.6 Phosphate In this study, phosphate represented in two form as phosphate (𝑃𝑂4 ) and total phosphate (𝑇𝑃) and for Autumn season one model were developed for each one. The best models for 𝑃𝑂4 and 𝑇𝑃 with highest R2 values of 0.702 and 0.746 respectively were obtained when 𝑃𝑂4 and 𝑇𝑃 values are correlated with bandratio (G/R) of Landsat 8 OLI. However, for Spring season, one model was developed for 𝑃𝑂4 and two models for 𝑇𝑃. The model of 𝑃𝑂4 is weakly correlated with (NIR/C) with R2 value of 0.213 and the two models of 𝑇𝑃 also are weakly correlated with (NIR/B) and (ICA NIR) with R2 values of 0.323 and 0.213 respectively. Measured and computed values of 𝑃𝑂4 and 𝑇𝑃 for Autumn and Spring season s are shown in Fig. (5-15) and Fig. (5-16). The developed models were used for generating and mapping 𝑃𝑂4 and 𝑇𝑃 in the Lake as shown in Fig. (5-17) and Fig. (5-18). In Autumn season, the computed 𝑃𝑂4 and 𝑇𝑃 were of high values, 0.4 mg/l and 0.1 mg/l, at the inlet of the lake and decrease toward Dokan Dam to be 0.15 mg/l and 0.055 mg/l respectively. In Spring season, the values decreased due to seasonal factors such as dilution of rainfall, aeration, and others. The values of 𝑃𝑂4 and 𝑇𝑃 are ranged 0.37-.15 mg/l and 0.18-0.02 mg/l from the inlet of the lake toward the dam respectively. 5.2.7 Dissolved Oxygen and Biochemical Oxygen Demand In the present study, 𝐷𝑂 is represented as concentration (mg/l). For Autumn season data, two multiple linear regression models were developed for 𝐷𝑂 concentration. The first 𝐷𝑂 model has R2 value of 0.738 when 𝐷𝑂 values are correlated with band-ratio (C/NIR) while the second model has R2 value of 0.832 when 𝐷𝑂 values is correlated with band-ratios (C/NIR) and (B/NIR + B). On the other hand, for Spring season no 𝐷𝑂 models were obtained because of the changes such as aeration and plant growth in the lake. Furthermore, 𝐵𝑂𝐷 models only developed for Spring season data and the best model obtained by using the band ratio (G/B) and (PCA B) with R 2 of 0.57. Measured and computed values of 𝐷𝑂 and 𝐵𝑂𝐷 in Dokan Lake for Autumn and [ 117 ]

Chapter Five

Result and Discussion

Spring season s are shown in Fig. (5-19). The developed models were used for generating and mapping 𝐷𝑂 and 𝐵𝑂𝐷 in the Lake as shown in Fig. (5-20). In Autumn season, DO concentration values that estimated from the model ranged from 9.45-10.84 mg/l at the inlet of the lake and decreased at centre of the lake to less than 3.6 mg/l due to the diffusion and aeration, photosynthesis, respiration and decomposition. In Lakes, dissolved oxygen concentrations will vary by season, location and water depth. In Spring season, there is no model developed for DO. In Autumn season, the values of BOD are ranged 0.4-3.6 mg/l while for Spring season are ranged 0.1-0.8 mg/l.

Measured V. Autumn Season

Computed V. Autumn Season

PO4 (mg/l)

.35 .30 .25 .20 .15 .10 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

17

18

19

20

Station Points

Measured V. Spring Season

Computed V. Spring Season

PO4 (mg/l)

.60 .50 .40 .30 .20 .10

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Station Points

Fig. (5-15): Measured and computed values of 𝑃𝑂4 in Dokan Lake for Autumn and Spring Seasons.

[ 118 ]

Chapter Five

Result and Discussion

TP (mg/l)

Measured V. Autumn Season

Computed V. Autumn Season

.12 .10 .08 .06 .04 .02 .00 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

17

18

19

20

Station Points

Measured V. Spring Season

Computed V. Spring Season

TP (mg/l)

.20 .15 .10

.05 .00 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Station Points

Fig. (5-16): Measured and computed values of 𝑇𝑃 in Dokan Lake for Autumn and Spring Seasons.

Fig. (5-17): Computed 𝑃𝑂4 values in Dokan Lake for Autumn Season (left) and Spring Season (right).

[ 119 ]

Chapter Five

Result and Discussion

Fig. (5-18): Computed 𝑇𝑃 values in Dokan Lake for Autumn Season (left) and Spring Season (right).

DO Measured V. Autumn Season

DO Computed V. Autumn Season

DO, BOD (mg/l)

10.00 8.00 6.00 4.00 2.00 0.00 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

18

19

20

Station Points

BOD Measured V. Spring Season

BOD Computed V. Spring Season

DO, BOD (mg/l)

1.00 .80 .60

.40 .20 .00 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

Station Points

Fig. (5-19): Measured and computed values of 𝐷𝑂 and 𝐵𝑂𝐷 in Dokan Lake for Autumn and Spring Seasons respectively.

[ 120 ]

Chapter Five

Result and Discussion

Fig. (5-20): Computed 𝐷𝑂 values for Autumn Season (left) and 𝐵𝑂𝐷 values for Spring Season (right) in Dokan Lake.

5.2.8 Total Suspended Solid and Turbidity For Autumn season data, two multiple linear regression models have been developed for predicting the total suspended solid (𝑇𝑆𝑆) and turbidity of Dokan lake water. The best model of them was obtained by correlating total suspended solid and turbidity values with (R+NIR) and band-ratio of (C/R) with a highest R2 value of 0.998. On the other hand, for Spring season, two models have been developed for predicting the total suspended solid and turbidity in the Lake. The model that have a highest R2 value (0.463) was obtained by correlating total suspended solid and turbidity values with band-ratios of (B/R) and (G/C). The performance reduction of the model in this season is due to the outlying value of 𝑇𝑆𝑆 (high value) at station point 16 which is close to Lake margin. The 𝑇𝑆𝑆 and Turbidity values in Autumn season include an outlier value in station point 16, therefore, the value of this point was assumed as a missed value in developing the [ 121 ]

Chapter Five

Result and Discussion

models in order to increase the R2 value as illustrated Fig. (5-21) and Fig. (5-22) which show the measured and computed values of 𝑇𝑆𝑆 and turbidity for Autumn and Spring season in Dokan Lake. The models of highest R2 values were selected to generate and map the 𝑇𝑆𝑆 and turbidity values in the Lake as shown in Fig. (523) and Fig. (5-24). In Autumn season, turbidity values estimated from the developed models are ranged 69.05-85.20 𝑁𝑇𝑈 at the inlet of the lake and decreased toward the Dokan dam which are less than 2.65 𝑁𝑇𝑈. However, the TSS concentration that estimated from the best model is greater than 50 mg/l at the inlet of the lake and decreased toward the Dokan dam which is less than 1.0 mg/l. the higher values at the inlet are due to soil erosion, runoff, discharges and stirred bottom sediments or algal blooms. In Spring season, the values reversed that the value of Turbidity is less than 1 𝑁𝑇𝑈 at the inlet of the lake and increased toward the Dokan dam to be 5.6 𝑁𝑇𝑈. Also, for TSS the values ranged from 1.0 mg/l at the inlet to 4.5 mg/l at the dam. Measured V. Autumn Season

Computed V. Autumn Season

TSS (mg/l)

80.00 60.00 40.00 20.00 0.00 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

17

18

19

20

Station Points

Measured V. Spring Season

Computed V. Spring Season

TSS (mg/l)

5.00 4.00 3.00 2.00 1.00 0.00 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Station Points

Fig. (5-21): Measured and computed values of 𝑇𝑆𝑆 in Dokan Lake for Autumn and Spring Seasons. [ 122 ]

Chapter Five

Result and Discussion Actual V. Autumn Season

Predicted V. Autumn Season

Turbidity (NTU)

80.00 60.00 40.00 20.00 0.00 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

16

17

18

19

20

Station Points

Actual V. Spring Season

Predicted V. Spring Season

Turbidity (NTU)

5.00 4.00 3.00 2.00 1.00 0.00 1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Station Points

Fig. (5-22): Measured and computed values of Turbidity in Dokan Lake for Autumn and Spring Seasons.

Fig. (5-23): Computed 𝑇𝑆𝑆 values in Dokan Lake for Autumn Season (left) and Spring Season (right).

[ 123 ]

Chapter Five

Result and Discussion

Fig. (5-24): Computed Turbidity values in Dokan Lake for Autumn Season (left) and Spring Season (right). 5.3 Estimating Water Quality Index Water quality index from the OLI bands of Landsat 8 spectral bands using statistical correlation and developing a model equation have been estimated in this study. Four methods of water quality index, OWQI, CCMEWQI, AWWQI and NSFWQI were used and based on the results of field and laboratory data of water quality parameters for two different season of the year. The selecting of the model parameters for developing the water quality index models is based on the Pearson correlation between water quality index and the spectral reflectance characteristics of the water in the OLI spectral bands as shown in appendix A. 5.3.1 Autumn Season Models Thirteen water quality indices models were developed for Autumn season data on October 24th, 2014 based on the highest determination coefficient (R2) as shown in Table (5-3). The models developed for the OWQI, CCMEWQI, AWWQI and NSFWQI water quality index calculating methods. [ 124 ]

Chapter Five

Result and Discussion

OWQI model has a low correlation with band-ratio of (C/NIR) with R2 value of 0.265. The low performance of the model is because the method depends on 𝐵𝑂𝐷 parameter in calculation which is not correlated with spectral bands in Autumn season as shown in previous sections. The AWWQI models show a high correlation with blue, green and red bands and NIR of Landsat 8 OLI. The best model has R2 value of 0.978 which is correlated with blue, green and red bands. On the other hand, the CCMEWQI model have correlation with many bands but mostly with band-ratios of (G/B) and (ICA SWIR 2) with R2 value of 0.637. The last model is NFSWQI which have a highest correlation with spectral reflectance of (B/NIR+B) and band ratio (C/NIR) with R2 value of 0.583. Finally, the best model has been chosen based on the highest R2 value which is AWWQI model with R2 value of 0.978 to generate and map the water quality index in Dokan Lake as shown in Fig. (5-25).

Table (5-3): Water quality index models for Autumn Season data on October 24th, 2014. Model Type

No.

Water Quality Index Model*

R2

1

𝑂𝑊𝑄𝐼 = 67.984 + 0.090 (𝑆𝑊𝐼𝑅 1)

0.214

2

𝑂𝑊𝑄𝐼 = 85.853 − 2.135 (𝐶/𝑁𝐼𝑅)

0.265

1

𝐴𝑊𝑊𝑄𝐼 = −14.656 + 0.244 (𝑅)

0.917

2

𝐴𝑊𝑊𝑄𝐼 = 70.006 + 0.392 (𝑅) − 0.189 (𝐺)

0.968

3

𝐴𝑊𝑊𝑄𝐼 = 40.395 + 0.446 (𝑅) − 0.470 (𝐺) + 0.394 (𝐵)

0.978

4

𝐴𝑊𝑊𝑄𝐼 = −87.589 (𝑁𝐼𝑅: 𝐹𝑙𝑜𝑎𝑡)

0.967

1

𝐶𝐶𝑀𝐸𝑊𝑄𝐼 = 69.146 + 0.008 (𝑅)

0.201

2

𝐶𝐶𝑀𝐸𝑊𝑄𝐼 = 31.452 + 29.877 (𝐺/𝐵)

0.395

3

𝐶𝐶𝑀𝐸𝑊𝑄𝐼 = −36.460 + 95.307 (𝐺/𝐵) + 3.739 (𝐼𝐶𝐴 𝑆𝑊𝐼𝑅 2)

0.637

1

𝑁𝑆𝐹𝑊𝑄𝐼 = 66.032 − 0.008 (𝐵/𝑁𝐼𝑅 + 𝐵)

0.384

2

𝑁𝑆𝐹𝑊𝑄𝐼 = 69.48 − 0.007 (𝐵/𝑁𝐼𝑅 + 𝐵) − 0.0773 (𝐶/𝑁𝐼𝑅)

0.583

OWQI

AWWQI

CCMEWQI

NSFWQI

* Selected models have the highest R2. [ 125 ]

Chapter Five

Result and Discussion

Fig. (5-25): Computed AWWQI values in Dokan Lake for Autumn Season (left) and Spring Season (right).

5.3.2 Spring Season Models In Spring season models, as it mentioned before, estimating the water quality index form depends on the spectral reflectance characteristics of the water in many spectral bands which is shown in appendix A. four models have been developed for water quality index for data on April 2nd, 2015 as shown in Table (5-4). The CCMEWQI and NSFWQI models show low performance and low correlation with band-ratios (NIR/C) and (ICA NIR) with R2 values of 0.38 and 0.29 respectively. On the other hand, AWWQI and OWQI models show moderate performance and correlation with band ratios (NIR/C) and (C/NIR) with R2 value of 0.612 and 0.61 respectively. As discussed before, OWQI calculations is based on 𝑇, 𝐷𝑂, 𝐵𝑂𝐷, 𝑇𝑃 and 𝑝𝐻, and the most important variables are 𝐷𝑂 and 𝐵𝑂𝐷. In this season, weak correlation one or more of these parameters with spectral bands led to modeling failure. The AWWQI model with highest R2 value of 0.612 [ 126 ]

Chapter Five

Result and Discussion

have been chosen for generating and mapping the water quality index values in Dokan Lake for Spring season as shown in Fig. (5-25) (see Sec. 5.3.1). The results in the present study show that the more eligible and more acceptable WQI method is AWWQI which gives the highest R2 values for both of Autumn and Spring seasons. Table (5-4): Water quality index models for Spring Season data on April 2nd, 2015. Model Type

No.

Water Quality Index Model*

R2

1

𝑂𝑊𝑄𝐼 = 62.927 − 0.961 (𝑁𝐼𝑅/𝐶)

0.456

2

𝑂𝑊𝑄𝐼 = 23.075 − 3.482 (𝑁𝐼𝑅/𝐶) − 0.126 (𝐶/𝑁𝐼𝑅

0.610

1

𝐶𝐶𝑀𝐸𝑊𝑄𝐼 = 62.706 − 23.450 (𝑁𝐼𝑅/𝐶)

0.380

1

𝐴𝑊𝑊𝑄𝐼 = 41.887 + 137.984 (𝑁𝐼𝑅/𝐶)

0.290

2

𝐴𝑊𝑊𝑄𝐼 = −248.386 + 775.499 (𝑁𝐼𝑅/𝐶) + 31.741 (𝐶/𝑁𝐼𝑅)

0.612

1

𝑁𝑆𝐹𝑊𝑄𝐼 = 63.459 − 1.723 (𝐼𝐶𝐴 𝑁𝐼𝑅)

0.223

OWQI CCMEWQI AWWQI NSFWQI

* Selected models have the highest R2.

5.4 Water Clarity Estimating Models Water clarity is an indirect measure of a lake’s trophic state and its status in terms of nutrient concentrations and biological productivity. Many studies have shown that Trophic State Index (TSI) and Secchi Disk Transparency SDT have correlation with satellite image. Many studies have shown that they can be estimated by remote sensing. One of the most measured properties of water in lakes is SDT which is the most consistently collected trophic state indicator. SDT can be used for water clarity measurements, along with various transformations such as TSI. The estimated SDT can be converted to Carlson’s trophic state index based on calculation and correlation of TSI with spectral responses. In the present study the SDT and TSI are tested in Dokan lake for two different seasons of the year, Autumn and Spring seasons. One SDT model has been developed for [ 127 ]

Chapter Five

Result and Discussion

Autumn season as shown in Table (5-5). The model shows that the SDT is strongly correlated with the responses in spectral indices (NDMI) of Landsat 8 OLI data with R2 of 0.963. Furthermore, two models of TSI were developed and the result shows that the TSI is highly correlated with band-ratios (R/NIR) and (LSWI) which gives R2 value of 0.935. The best developed models have been used to generate and map the SDT and TSI values in Dokan Lake for Autumn season as shown in Fig. (526) and Fig. (5-27). In Autumn season, SDT values in the lake vary from less than 0.25 m at the inlet of the lake to greater than 4 m at Dokan dam. In this season, the actual values of measured SDT ranged 0.3-3.3 m. In Spring season, generally, the values of SDT and TSI are remained as in Autumn season without a significant change. Table (5-5): Water clarity models for Autumn Season on October 24th, 2014. Autumn Season Models*

R2

Model Type

No.

SDT

1

𝑆𝐷𝑇 = 3.973 − 5.134 (𝑁𝐷𝑀𝐼)

0.963

1

𝑇𝑆𝐼 = 32.227 + 7.662 (𝑅/𝑁𝐼𝑅)

0.915

2

𝑇𝑆𝐼 = 35.993 + 5.692 (𝑅/𝑁𝐼𝑅) + 18.724 (𝐿𝑆𝑊𝐼)

0.935

TSI

* Selected models have highest R2.

On the other hand, for Spring season, six models have been developed for SDT and TSI, three for each one as shown in Table (5-6). The results show that the SDT is strongly correlated with the responses in spectral indices such as (MNDWI) and with band ratios (C/G) and (G/R) of Landsat 8 OLI data, and having highest R2 value of 0.951. In addition, through the correlation process, three models have been developed for TSI and the results show a high correlation with band ratios (R/C) and (MNDWI) which give R2 value of 0.973. As the models developed, the values of SDT and TSI are computed and mapped on Landsat 8 OLI image of Dokan Lake for Autumn and Spring season s as shown in Fig (5-26) and Fig. (5-27) respectively. [ 128 ]

Chapter Five

Result and Discussion

Table (5-6): Water clarity models for Spring Season on April 2nd, 2015. Model Type

SDT

TSI

Spring Seasons Model*

No

R2

1

𝑆𝐷𝑇 = −3.541 + 6.875 (𝐶/𝐺)

0.859

2

𝑆𝐷𝑇 = −3.035 + 4.897 (𝐶/𝐺) + 0.436 (𝐺/𝑅 )

0.899

3

𝑆𝐷𝑇 = 1.677 + 2.083 (𝐶/𝐺) + 0.675 (𝐺/𝑅) − 3.684 (𝑀𝑁𝐷𝑊𝐼)

0.951

1

𝑇𝑆𝐼 = 33.429 + 32.934 (𝑅/𝐶)

0.946

2

𝑇𝑆𝐼 = 18.826 + 29.115 (𝑅/𝐶) + 20.364 (𝑀𝑁𝐷𝑊𝐼)

0.973

3

𝑇𝑆𝐼 = 37.175 + 31.632 (𝑅/𝐶) − 134.590 (𝑆𝑊𝐼𝑅2: 𝐹𝑙𝑜𝑎𝑡)

0.973

* Selected Models have highest R2.

Fig. (5-26): Computed SDT values in Dokan Lake for Autumn Season (left) and Spring Season (right).

[ 129 ]

Chapter Five

Result and Discussion

Fig. (5-27): Predicted TSI values in Dokan Lake for Autumn Season (left) and Spring Season (right).

5.5 Generality of Models The performance of any model is measured by how well they can predict unseen data (an unseen data set is one that was not used during developing the model) and this is known as generality of model. In this study, to check the generality of the developed models, the models that developed for Autumn season data have been tested by using the Spring season data and vice versa. Eighteen models have been developed for data on October 24 th, 2014 (Autumn Season data). These models were used to predict the water quality indices and other parameters for data of Dokan lake water on April 2 nd, 2015 (Spring season data). A new coefficients of determination (R2) have been calculated for each model as shown in Table (5-7). The results show that the values of R2 are decreased below 0.5 except for Secchi disk the value decreased to 0.754.

[ 130 ]

Chapter Five

Result and Discussion

On the other hand, sixteen models have been developed for data on April 2 nd, 2015 (Spring season data). These models were used to predict the water quality indices and other parameters for data of Dokan lake water on October 24th, 2014 (Autumn season data). A new coefficients of determination (R2) have been calculated for each model as shown in Table (5-8). The results show that the values of R2 are decreased below 0.5. The main factor which causes to change the performance of models is high variation of all data when the season is changed. As clear from the results, no model can be used as general model for all seasons of the year.

Table (5-7): Generality of Autumn Season models on October 24th, 2014. WQPs

Autumn Season Models

R2 (Autumn season)

R2 (Spring season)

pH

𝑝𝐻 = 10.491 − 0.051 ( 𝐼𝐶𝐴 𝑆𝑊𝐼𝑅1 ) − 0.742 ( 𝐺 / 𝑅 ) − 0.147 ( 𝐺 ∶ 𝐹𝑙𝑜𝑎𝑡 )

0.645

0.470

EC

𝐸𝐶 = 241.500 + 529.504 ( 𝑁𝐼𝑅 / 𝐶 )

0.492

0.416

TDS

𝑇𝐷𝑆 = 120.750 + 264.752 ( 𝑁𝐼𝑅 / 𝐶 )

0.492

0.415

𝑇. = 21.765 − 0.001 ( 𝐶 + 𝑅 )

0.527

0.495

𝑁𝑂3 = 4.678 − 10.400 ( 𝑁𝐵𝑅 ) + 0.004 ( 𝐵 / 𝑁𝐼𝑅 + 𝐵 )

0.652

0.461

𝑁𝑂3 − 𝑁 = 0.772 − 2.445 ( 𝐿𝑆𝑊𝐼 ) + 0.001 ( 𝐵 / 𝑁𝐼𝑅 + 𝐵 )

0.678

0.454

PO4

𝑃𝑂4 = 0.401 − 0.083 (𝐺 / 𝑅)

0.492

0.414

TP

𝑇𝑃 = 0.127 − 0.025 ( 𝐺 / 𝑅 )

0.556

0.405

TSS

𝑇𝑆𝑆 = − 50.077 + 0.063 ( 𝑅 + 𝑁𝐼𝑅 ) + 16.708 ( 𝐶 / 𝑅 )

0.982

0.499

NTU

𝑁𝑇𝑈 = − 42.564 + 0.052 ( 𝑅 + 𝑁𝐼𝑅 ) + 13.923 ( 𝐶 / 𝑅)

0.982

0.499

DO

𝐷𝑂 = 10.841 − 0.682 ( 𝐶 / 𝑁𝐼𝑅 ) − 0.002 ( 𝐵 / 𝑁𝐼𝑅 + 𝐵 )

0.832

0.500

-

-

T NO3 NO3-N

BOD

No Model

SDT

𝑆𝐷𝑇 = 3.973 − 5.134 ( 𝑁𝐷𝑀𝐼 )

0.928

0.754

TSI

𝑇𝑆𝐼 = 35.993 + 5.692 ( 𝑅 / 𝑁𝐼𝑅 ) + 18.724 ( 𝐿𝑆𝑊𝐼 )

0.873

0.748

𝑁𝑆𝐹𝑊𝑄𝐼 = 69.48 − 0.007 (𝐵/𝑁𝐼𝑅 + 𝐵) − 0.0773 (𝐶/𝑁𝐼𝑅)

0.583

0.364

𝑂𝑊𝑄𝐼 = 85.853 − 2.135 ( 𝐶 / 𝑁𝐼𝑅 )

0.265

0.127

𝐴𝑊𝑊𝑄𝐼 = 40.395 + 0.446 (𝑅) − 0.470 (𝐺) + 0.394 (𝐵)

0.978

0.475

𝐶𝐶𝑀𝐸𝑊𝑄𝐼 = −36.460 + 95.307 (𝐺/𝐵) + 3.739 (𝐼𝐶𝐴 𝑆𝑊𝐼𝑅 2)

0.637

0.499

NSFWQI OWQI AWWQI CCMEWQI

[ 131 ]

Chapter Five

Result and Discussion

Table (5-8): Generality of Spring Season models on April 2nd, 2015. WQPs

Spring Season Models

R2 (Spring season)

R2 (Autumn season)

pH

𝑝𝐻 = 8.485 + 0.045 ( 𝐼𝐶𝐴 𝐶 )

0.278

0.460

EC

𝐸𝐶 = 422.034 − 1080.365 ( 𝑆𝑊𝐼𝑅1 ∶ 𝐹𝑙𝑜𝑎𝑡 )

0.214

0.421

-

-

𝑇. = 10.516 + 0.386 ( 𝐼𝐶𝐴 𝐵3 ) + 2.933 ( 𝐺 / 𝐵 )

0.862

0.497

𝑁𝑂3 = 4.043 − 2.068 ( 𝑅 / 𝐶 )

0.553

0.356

𝑁𝑂3 − 𝑁 = 0.913 − 0.469 ( 𝑅 / 𝐶 )

0.559

0.357

PO4

𝑃𝑂4 = 0.146 + 0.552 ( 𝑁𝐼𝑅 / 𝐶 )

0.213

0.371

TP

𝑇𝑃 = 0.089 + 0.016 ( 𝐼𝐶𝐴 𝑁𝐼𝑅)

0.323

0.420

TSS

𝑇𝑆𝑆 = − 5.068 + 1.695 ( 𝐵 / 𝑅 ) + 3.548 ( 𝐺 / 𝐶 )

0.463

0.076

NTU

𝑁𝑇𝑈 = −4.223 + 1.412 ( 𝐵 / 𝑅 ) + 2.957 ( 𝐺 / 𝐶 )

0.463

0.076

-

-

TDS T. NO3 NO3-N

DO

No Model

No Model

BOD

𝐵𝑂𝐷 = − 1.037 + 1.062 ( 𝐺 / 𝐵 ) + 0.003 ( 𝑃𝐶𝐴 𝐵 )

0.570

0.389

SDT

𝑆𝐷𝑇 = 1.677 + 2.083 ( 𝐶 / 𝐺 ) + 0.675 ( 𝐺 / 𝑅 ) − 3.684 ( 𝑀𝑁𝐷𝑊𝐼 )

0.951

0.435

TSI

𝑇𝑆𝐼 = 37.175 + 31.632 ( 𝑅 / 𝐶 ) − 134.590 ( 𝑆𝑊𝐼𝑅2 ∶ 𝐹𝑙𝑜𝑎𝑡 )

0.973

0.676

𝑁𝑆𝐹𝑊𝑄𝐼 = 63.459 − 1.723 ( 𝐼𝐶𝐴 𝑁𝐼𝑅 )

0.223

0.490

𝑂𝑊𝑄𝐼 = 23.075 − 3.482 (𝑁𝐼𝑅/𝐶) − 0.126 (𝐶 ⁄𝑁𝐼𝑅 )

0.610

0.120

𝐴𝑊𝑊𝑄𝐼 = −248.386 + 775.499 ( 𝑁𝐼𝑅 / 𝐶 ) + 31.741 ( 𝐶 / 𝑁𝐼𝑅 )

0.612

0.066

𝐶𝐶𝑀𝐸𝑊𝑄𝐼 = 62.706 − 23.450 ( 𝑁𝐼𝑅 / 𝐶)

0.380

0.475

NSFWQI OWQI AWWQI CCMEWQI

[ 132 ]

CHAPTER SIX CONCLUTIONS AND RECOMENDATIONS

Chapter Six

Conclusions and Recommendations

Chapter Six Conclusions and Recommendations 6.1 Conclusions From the results analyses of experimental works, remote sensing processing works and WQPs, WQI and TSI models in the present study, the following facts were observed and concluded: 1. Landsat 8 OLI images can be used to predict the water quality indices and parameters in lakes. 2. The results show that all bands of Landsat 8 OLI can be correlated with water quality indices and parameters especially the new band (Coastal Blue) with its combinations and band-ratios, independent component analysis (ICA) and minimum noise fraction (MNF) which are not used before. 3. The arithmetic weighted water quality index (AWWQI) is highly correlated with reflectance bands of Landsat 8 OLI. It can be used to estimate the quality class of Dokan Lake water for both Autumn and Spring seasons. 4. For Canadian method of water quality index (CCMEWQI), it has been noticed that the effect of the coliform bacteria is very strong in reduction the water quality class of some stations to Poor and Very Poor classes. 5. Also it has been noticed there is a weak correlation between Oregon water quality index (OWQI) and spectral bands for both Autumn and Spring seasons. 6. The national sanitation foundation water quality index (NSFWQI) model show that it can be used for Autumn Season data (small variance between station points data), and it is not suitable for Spring season data (high variance between station points data). 7. The obtained results of Secchi disk transparency parameter prove that it can modeled to be used for a whole year. 8. Trophic State Index (𝑇𝑆𝐼) results show a high correlation with reflectance bands of Landsat 8 OLI as shown for other satellites in previous studies. [ 133 ]

Chapter Six

Conclusions and Recommendations

9. The correlation of water quality whether its parameters and indices with reflectance are decrease in Spring season due to seasonal changes (high variance between station points data). 10. All water quality parameters that used in this study can be computed by using satellite images except 𝐵𝑂𝐷 parameter for Autumn season while for Spring season only Temperature, 𝑁𝑂3 , 𝑁𝑂3 − 𝑁 and 𝐵𝑂𝐷 can be computed using satellite images. 11. The results represent that temperature not only correlated with radiance of satellite image date but also correlated indirectly with reflectance of satellite data and can be accurately computed and mapped in Dokan Lake because it has a correlation with other WQPs. Because change in Temperature cause the change in other WQPs. 6.2 Recommendations The following recommendations were found to provide a guidance for further studies: 1. It is important to continue studying remote sensing of water quality index and parameters at different depths in Dokan Lake using thermal and Microwave Remote sensing. 2. In future studies, it is recommended to correlate the water quality index with spectral bands using nonlinear multiple regression, different spectral bands ratios and combinations and different satellite images data. 3. It is recommended to study other WQI and WQPs for Dokan Lake such as heavy metals or any hazards martials. 4. To develop the general model represents a whole year, it needs to get samples corresponding the acquisition date of satellite data at interval times of specific location for at least a year.

[ 134 ]

REFERENCES

References

References 1. Abbasi, T. & Abbasi, S., 2012. Water Quality Indices. Chinnakalapet, Puducherry, India: ELSEVIER. 2. Akbar, T. A., Hassan, Q. K. & Achari, G., 2014. Development of Remote Sensing Based Models for Surface Water Quality. Clean – Soil, Air, Water, 42(8), p. 1044–1051. 3. Al-Bahrani, H. S. M., Abdulrazzaq, D. K. A. & Saleh, D. S. A. A.-H., 2012. A satellite image model for prediction water quality index of Euphrates river in Iraq. Baghdad: University of Baghdad. 4. Ali, S. S. & Salley, S. K. R., 2003. Numerical Groundwater Flow Modeling for The Intwrgranular Aquifer in Sarsian Sub-Basin, Dokan Lake, Iraqi Kurdistan Region. Journal of Zankoy Sulaimani-Part A, 15(1), pp. 125-141. 5. Almdny, A. H., Belhaj, O. & Afan, A. M., 2010. Impact Of Water Quality On The Spatial Distribution Of Vegetation Cover In The Coastal Regions With Remote Sensing And Gis Methods. Cairo, Egypt, Fourteenth International Water Technology Conference, IWTC 14. 6. Almdny, A. H., Ekhmaj, A. I., Abdul Aziz, A. M. & AL Jaml, H., 2008. Possibility Of Using Gis And Remote Sensing Approach In Monitoring Of Change In Plant Cover Due To Change In Water Quality In Costal Regions. Alexandria, Egypt, Twelfth International Water Technology Conference, IWTC12. 7. Alparslan, E., Aydöner, C., Tufekci, V. & Tüfekci, H., 2007. Water quality assessment at Ömerli Dam using remote sensing techniques. Environ Monit Assess, Volume 135, pp. 391-398. 8. Alparslan, E., Gonca Coşkun, H. & Alg, U., 2010. An Investigation on Water Quality of Darlık Dam Drinking Water using Satellite Images. 10(1293– 1306).

[ 135 ]

References

9. Ararat, K., Hassan , N. A. & Abdul Rahman, S., 2009. Key Biodiversity Survey of Kurdistan, Northern Iraq, Sulaimani, Kurdistan, Iraq: Nature Iraq Report. 10. Baboo, C. D. S. S. & Thirunavukkarasu, S., 2014. Geometric Correction in High Resolution Satellite Imagery using Mathematical Methods: A Case Study in Kiliyar Sub Basin. Global Journal of Computer Science and Technology: Graphics & Vision, 25(1). 11. Bartram, J. & Ballance, R., 1996 . Water Quality Monitoring - A Practical Guide to the Design and Implementation of Freshwater Quality Studies and Monitoring Programmes. s.l.:UNEP/WHO. 12. Berthouex, P. M. & Brown, L. C., 2002. Statistics for Enviromental Engineers. 2 ed. Boca Raton, London, New York, Washington, D.C.: Lewis Publishers. 13. Bilbas, A. H. A., 2014. Ecosystem Health Assessment of Dukan Lake, Sulaimani, Kurdistan Region of Iraq. Erbil: Salahaddin University. 14. Bilgehan, N. et al., 2010. An Application of Landsat-5TM Image Data for Water Quality Mapping in Lake Beysehir, Turkey. Water Air Soil Pollut, Volume 212, p. 183–197. 15. Campbell, D. & Campbell, S., 2008. Introduction to regression and Data Analysis. s.l., StatLab Workshop Series 2008. 16. Campbell, J. B. & Wynne, R. H., 2011. Introduction of Remote sensing. Fifth Edition ed. New York: A Division of Guilfod Publication Inc.. 17. Carlson, E. & Ecker, M. D., 2002. A Statistical Examination of Water Quality in Two Iowa Lakes. American Journal Of Undergraduate Research, 1(2). 18. Carlson, R. E., 1977. A trophic state index for lakes. Limnology and Oceonography , pp. 361-369. 19. CCRS, C. C. f. R. S., 2014. Tutorial: Fundamentals of Remote Sensing. [Online] Available at: http://www.nrcan.gc.ca/earth-sciences/geomatics/satellite-imagery-airphotos/satellite-imagery-products/educational-resources/9309 [Accessed 03 07 2015]. [ 136 ]

References

20. Chapman, D., 1996. Water Quality Assessments - A Guide to Use of Biota,Sediments and Water in Environmental Monitoring. Second ed. Great Britain at the University Press, Cambridge: UNESCO/WHO/UNEP. 21. Chatterjee, S. & Hadi, A. S., 2006. Regression Analysis by Example. Fourth Edition ed. New Jersey: John Wiley & Sons, Inc., Publication. 22. Chen, P. Y. & Popovich, P. M., 2002. Correlation: Parametic and Nonparametric measure. 07 ed. USA: Sage Publications Inc.. 23. Cude, C. G., 2001. Oregon water qualitty index: A tool for evaluating water quality mangement effectiveness. JAWRA : Journal of the American water resources association, 37(1), pp. 125-137. 24. Danbara, T. T., 2014. Deriving Water Quality Indicators of lake Tana, Ethopia, from Landsat-8. Enschede, Netherlands: s.n. 25. Dowdy, S., Weardon, S. & Chilko, D., 2004. Statistics for reasearch. Third Edition ed. Hoboken, New Jersey.: John Wiley & Sons, Inc. 26. Doxaran, D., Cherkuru, R. C. N. & Lavender, S. J., 2005. Use of reflectance band ratios to estimate suspended and dissolved matter concentrations in estuarine waters. International Journal of Remote Sensing, 26(8), pp. 17631769. 27. EPA, E. P. A., 2001. Parameters of Water Quality Interpretation and Standards. Ireland: Environmental Protection Agency,. 28. Fan, C., 2014. Spectral Analysis of Water Reflectance for Hyperspectral Remote Sensing of Water Quailty in Estuarine Water. Journal of Geoscience and Environment Protection, Volume 2, pp. 19-27. 29. Fewtrell, L. & Bartram, J., 2001. Water Quality Guidelines, Standards and Health: Assessment of risk and risk management for water-related infectious disease. Padstow, Cornwall, UK: World Health Organization. 30. Gardino, C. et al., 2014. Optical remote sensing of lakes: an overview on Lake Maggiore. J. Limnol, Volume 73, pp. 201-214.

[ 137 ]

References

31. Ghazal, N. K. & Hassoon, K. I., 2012. Temperature Calculation Using Thermal Bands Of (ETM+) Sensor. Iraqi Journal of Science , 53(2), pp. 435443. 32. Giardino, C. et al., 2014. Evaluation of Multi-Resolution Satellite Sensors for Assessing Water Quality and Bottom Depth of Lake Garda. Sensors, Volume 14, pp. 24116-24131. 33. Gibson, P. J., 2000. Introdutory Remote Sensing : Principle and Concepts. First ed. New York NY: Routledge. 34. Goslee, S. C., 2011. Analyzing Remote Sensing Data in R: The landsat Package. Journal of Statistical Software, 43(4), pp. 1-25. 35. Hedger, R. D., Atkinson, P. M. & Malthus, T. J., 2001. Optimizing sampling strategies for estimating mean water quality in lakes using geostatistical techniques with remote sensing. Lakes & Reservoirs: Research and Management, Volume 6, pp. 279-288. 36. Hellwegera, F., Schlosser, P., Lall, U. & Weissel, J., 2004. Use of satellite imagery for water quality studies in New York Harbor. Estuarine, Coastal and Shelf Science, Volume 61, pp. 437-448. 37. HE, W., CHEN, S., LIU, X. & CHEN, J., 2008. Water quality monitoring in slightly-polluted inland water body through remote sensing —A case study in Guanting Reservoir, Beijing, China. Environ. Sci. Engin. China, Volume 1. 38. Hua, C. et al., 2004. Assessment of estuarine water-quality indicators using MODIS medium-resolution bands: Initial results from Tampa Bay, FL. Remote Sensing of Environment 93 (2004) 423–441, Volume 93, pp. 423-441. 39. Johnson, R. A. & Bhattacharyya, G. K., 2010. Statistics : Principles and Methods. Six Edition ed. Hoboken, USA: John Wiley & Sons, Inc.. 40. Joseph, G., 2005. Fundamental of Remote Sensing. Second Edition ed. Delhi: Universities Press (India) Private Limited. 41. Kabbaraa, N., J. B., M. A. & V. B., 2008. Monitoring water quality in the coastal area of Tripoli (Lebanon) using high-resolution satellite data. ISPRS Journal of Photogrammetry & Remote Sensing, Volume 63, pp. 488-495. [ 138 ]

References

42. Kemker, C., 2014. Fundamentals of Environmental Measurements. Fondriest Environmental, Inc.. [Online] Available at: http://www.fondriest.com/environmental-measurements/parameters/waterquality/water-temperature/. [Accessed 01 07 2015]. 43. Kottegoda, N. T. & Rosso, R., 2008. Applied statistics for civil and enviromental engineers. Second Edition ed. New York: Blackwell Publishing Ltd. 44. Kutser, T. et al., 2012. Remote Sensing Of Water Quality In Optically Complex Lakes. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B8, pp. 165-169. 45. Levin, N., 1999. 1st Hydrographic Data Management course: Fundamentals of Remote Sensing. Trieste, Italy: IMO - International Maritime Academy, Remote Sensing Laboratory, Geography Department, Tel Aviv University, Israel, GIS unit, the Society for the Protection of Nature in Israel. 46. Lillesand, T. M., Kiefeer, R. W. & Chipman, J. W., 2007. Remote sensing and image interpretation. Six Edition ed. USA: John Wiley & Sons Inc.. 47. Mahdi, M. S., Ziboon, D. A. R. T. & Al Zubaidy, D. R. Z., 2009. Remote sensing model for monitering water depth and clarity of Al Huweizah marsh. Baghdad: University of Technology. 48. Mancino, G. et al., 2009. Assessing water quality by remote sensing in small lakes: the case study of Monticchio lakes in southern Italy. iForest – Biogeosciences and Forestry, pp. 154-161. 49. Meer, F. D. v. d. & Jong, S. M. d., 2006. Image Spectrometry ; Basic Principle and Prospective applications. Third Edition ed. Donrecht, The Netherlands: Springer. 50. Michaud, J. P., 1991. A Citizens' Guide to Understanding and Monitoring Lakes and -Streams. s.l.:Washington state Department of Ecology. 51. Ming, S. Y., Carolyn , J. M. & Robert, M. S., 1996. Adaptive Short-Term Water Quality Forecasts Using Remote Sensing And Gis. Awra Symposium On Gis And Water Resources, Issue Sept 22-26. [ 139 ]

References

52. Nas, B., Karabork, H., Ekercin, S. & Berktay, A., 2008. Assessing Water Quality in the Beysehir Lake ( Turkey ) by the Application of GIS, Geostatistics and Remote Sensing. Konya, Turkey, The 12th World lake Conference: 639-646. 53. Navulur, K., 2007. Multispectral image analysis using the object-oriented paradigm. First Edition ed. Boca Raton, FL: CRCpress, Taylor & Francis Group. LLC. 54. Olet, E., 2010. Water Quality Monitering of Roxo resevoir Landsat images and In-Situ Measurements. Enschede, The Netherlands: s.n. 55. Olmanson, L. G., Bauer, M. E. & Brezonik, P. L., 2002. Use of Landsat Imagery to develop a water quality ATLAS of Minnesota's 10,000 lakes. USA, Pecora 15/Land Satellite Information IV/ISPRS Commission I/FIEOS. 56. Omer, T. M. A., 2011. Country Pasture/Forage Resource Profiles. Rome, Italy.: FAO. 57. Oram, B., 2014. Monitoring the Quality of Surfacewaters : Calculating NSF Water Quality Index. [Online] Available at: http://www.water-research.net/index.php/water-treatment/watermonitoring/monitoring-the-quality-of-surfacewaters [Accessed 7 September 2015]. 58. Peng Li, L. J. a. Z. F., 2014. Cross-Comparison of Vegetation Indices Derived from Landsat-7 Enhanced Thematic Mapper Plus (ETM+) and Landsat-8 Operational Land Imager (OLI) Sensors. remote sensing, 6(ISSN 2072-4292), pp. 310-329. 59. Poonam, T., Tanushree, B. & Sukalyan, C., 2013. Water quality indices important tolls for water qaulity assessment : a review. International Journal of Advances in Chemistry (IJAC), 1(1), pp. 15-28. 60. Rawlings, J. O., Pantula, S. G. & Dickey, D. A., 1998. Applied Regression Analysis: A Research Tool,. Second Edition ed. Newyork: Springer.

[ 140 ]

References

61. Ray, T. W., 1994. A FAQ on Vegetation in Remote Sensing. [Online] Available

at:

http://www.yale.edu/ceo/Documentation/rsvegfaq.html

[Accessed 25 12 2014]. 62. Ritchie, J. C., Zimba, P. V. & Ever, J. H., 2003. Remote Sensing Techniques to Assess Water Quality. Photogrammetric Engineering & Remote Sensing, Volume 69, pp. 695-704. 63. Rokni, K., Ahmad, A., Selamat, A. & Hazini, S., 2014. Water Feature Extraction and Change Detection Using Multitemporal Landsat Imagery Multitemporal Landsat Imagery. remote sensing, Volume 6, pp. 4173-4189. 64. Ryan, T. P., 2007. Modern Engineering Statistics. Acworth, Georgia: John Wiley & Sons, Inc. 65. Sadek, A., 2009. Iraq Agriculture…Today Prospects for the national centre for organic agriculture. Iraq: Istituto Agronomico Mediterraneo di Bari. 66. Shayesh, A. K., 2006. Baghdad: University of Technology Baghdad. 67. Somvanshi, S., Kunwar, P., Singh, N. & Kachhwaha, T., 2011. Water Turbidity Assessment in Part of Gomti River Using High Resolution Google Earth’s Quickbird satellite data. Dimensions and Directions of Geospatial Industry. 68. Somvanshi, S. et al., 2012. Integrated remote sensing and GIS approach for water quality analysis of Gomti river, Uttar Pradesh. International Journal Of Environmental Sciences, Volume 3, pp. 62-74. 69. Stevanovic, ,. & Markovic, M., 2003. Hydrogeology of Northern Iraq. Vol. 2. General hydrogeology and aquifer systems. FAO. Vol.2 ed. Rome, Italy: Emergency Operations and Rehabilitation Div.. 70. Sudheer, K., I. C. & V. G., 2006. Lake Water Quality Assessment From Landsat Thematic Mapper Data Using Neural Network: An Approach To Optimal Band Combination Selection1. Journal Of The American Water Resources Association , Volume December, pp. 1683-1695. 71. Tempfli, K., Kerle, N., Huurneman, G. C. & Janssen, L. L. F., 2009. Principle of Remote Sensing ; An introductory textbook. Fourth Edition ed. Enschede, [ 141 ]

References

The Netherlands: The international institude for Geo-Information Science and Eaeth Observation (ITC). 72. Tyagi, S., Sharma, B. & Singh, P., 2013. Water qulaity assessment in term of water quality index. American Journal of Water Resources, 1(3), pp. 34-38. 73. U.-E. & B., 2013. Inventory of shared water resources in western asia. Beriut: United Nations Economic and Social Commission for Western; für, Bundesanstalt. 74. USGS, 2015. Landsat—A Global Land-Imaging Mission. [Online] Available at: http://remotesensing.usgs.gov. [Accessed 9 6 2015]. 75. Weisberg, S., 2005. Applied Linear Regression. Third Edition ed. Hoboken, New Jersey.: John Wiley & Sons, Inc. 76. WMO, W. M. O., 2013. Planning of water-quality monitering systems, Geneva, Switzerland: World Meteorological Organization Publication. 77. Yan, X. & Su, X. G., 2009. Linear Regression Analysis : Theory and Computing. First Edition ed. Singapore: World Scientific Publishing Co. Pte. Ltd..

[ 142 ]

APPENDICES

Appendix A

Appendix A The following tables are the output of IBM SPSS computer software for Pearson correlation of WQPs, WQI and TSI with spectral bands of Landsat 8 OLI. Table (A-1): Pearson Correlation between Water Quality (WQI, WQPs, SDT and TSI) and Spectral Bands for Autumn Season. Spectral Bands C

B

G

R

NIR

SWIR1

SWIR2

PCA C

PCA B

PCA G

PH

EC

TDS

T

NO3

NO3-N

PO4

TP

TSS

NTU

DO

BOD

OWQI

AWWQI

CQWI

NFSWQI

SDT

TSI

Cor.

-.022

-.006

-.006

-.665**

-.501*

-.502*

.561*

.592**

.724**

.724**

-.429

-.249

-.122

.729**

.216

-.471*

-.854**

.866**

Sig.

.926

.980

.980

.001

.024

.024

.010

.006

.000

.000

.059

.289

.608

.000

.360

.036

.000

.000

*

*

Cor.

-.064

.031

.031

**

-.661

*

-.477

*

-.478

.511

.549

**

.756

**

.756

-.377

-.224

-.092

**

.756

.207

-.444

**

-.812

.832**

Sig.

.790

.901

.901

.002

.034

.033

.021

.012

.000

.000

.101

.343

.699

.000

.382

.050

.000

.000

Cor.

.033

.085

.085

-.688**

-.521*

-.522*

.539*

.586**

.810**

.810**

-.299

-.246

-.027

.812**

.337

-.403

-.810**

.859**

Sig.

.890

.730

.730

.001

.019

.018

.014

.007

.000

.000

.200

.295

.909

.000

.146

.078

.000

.000

**

*

*

*

**

Cor.

.167

.321

.321

-.722

-.532

-.534

.552

.623

Sig.

.480

.180

.180

.000

.016

.015

.012

.003

.158

*

.564

*

.564

-.574

.366

*

Cor.

**

-.399

-.401

.445

**

.955

.955

-.012

-.244

.138

.958

.448

-.299

-.608

.766**

.000

.000

.960

.301

.561

.000

.048

.200

.004

.000

.394

-.073

-.204

.436

**

.950

**

**

.950

.362

-.108

.319

**

**

.944

*

**

Sig.

.507

.012

.012

.008

.081

.079

.113

.049

.000

.000

.117

.649

.171

.000

.086

.759

.389

.054

Cor.

-.197

.343

.343

.246

.497*

.496*

-.534*

-.517*

-.100

-.100

.618**

.092

.463*

-.125

-.149

.486*

.763**

-.667**

Sig.

.406

.151

.151

.295

.026

.026

.015

.020

.674

.674

.004

.701

.040

.598

.532

.030

.000

.001

*

*

*

*

Cor.

-.226

.359

.359

.197

.503

.502

-.540

-.513

-.064

-.064

.598

.084

.453

-.092

-.183

.454

.744

-.636**

Sig.

.339

.132

.132

.406

.024

.024

.014

.021

.788

.788

.005

.723

.045

.700

.440

.044

.000

.003

.706

*

.523

*

.524

Cor.

-.049

-.143

-.143

**

*

*

-.550

*

**

-.602

**

-.855

**

-.855

**

.242

.247

-.004

**

-.857

**

**

-.340

.392

.767

-.842**

Sig.

.838

.560

.560

.001

.018

.018

.012

.005

.000

.000

.303

.295

.987

.000

.143

.087

.000

.000

Cor.

-.273

-.555*

-.555*

.619**

.429

.431

-.421

-.509*

-.957**

-.957**

-.361

.184

-.343

-.956**

-.515*

.082

.242

-.495*

Sig.

.245

.014

.014

.004

.059

.058

.064

.022

.000

.000

.117

.437

.138

.000

.020

.730

.304

.027

**

**

**

**

**

**

Cor.

.228

.226

.226

-.713

-.600

-.602

.624

.687

.896

.896

-.124

-.252

.043

.904

.482

-.364

-.715

.845**

Sig.

.334

.352

.352

.000

.005

.005

.003

.001

.000

.000

.603

.284

.857

.000

.032

.114

.000

.000

[ A1 ]

**

**

*

**

Appendix A

Spectral Bands PCA R

PCA NIR

PCA SWIR1

PCA SWIR2

IPVI

LSWI

MSI

NBR

NDMI

NDVI

RVI

DVI

MNDWI

MNF C

PH

EC

TDS

T

NO3

NO3-N

PO4

TP

TSS

NTU

DO

BOD

OWQI

AWWQI

CQWI

NFSWQI

SDT

TSI

Cor.

-.167

-.397

-.397

.676**

.361

.362

-.369

-.454*

-.907**

-.907**

-.174

.291

-.333

-.903**

-.525*

.122

.412

-.600**

Sig.

.482

.093

.093

.001

.118

.117

.109

.044

.000

.000

.464

.213

.151

.000

.017

.607

.071

.005

Cor.

.212

*

*

.551

-.346

*

*

.217

*

.561

-.017

-.080

.259

.016

.551

-.132

-.134

.329

.379

.549

.549

.406

-.111

Sig.

.369

.015

.015

.135

.580

.572

.157

.100

.012

.012

.076

.642

.359

.010

.943

.739

.271

.948

Cor.

.442

-.249

-.249

.240

-.045

-.043

.114

.068

-.487*

-.487*

-.106

-.093

-.075

-.452*

.257

.089

-.007

-.038

Sig.

.051

.304

.304

.309

.851

.858

.633

.030

.030

.656

.697

.754

.045

.274

.710

.976

.874

*

*

.775 *

*

**

**

**

*

Cor.

-.186

-.060

-.060

.400

.514

.515

-.553

-.554

-.622

-.622

.117

.176

-.013

-.636

-.450

.240

Sig.

.432

.807

.807

.081

.020

.020

.011

.011

.003

.003

.624

.459

.958

.003

.047

.307

Cor.

-.017

-.337

-.337

-.168

-.287

-.285

.393

.383

.197

.197

**

-.577

-.028

-.388

.207

.193

**

.583

.007 *

-.456

**

-.880

-.602** .005 .683**

Sig.

.942

.158

.158

.478

.221

.223

.087

.096

.405

.405

.008

.908

.091

.381

.415

.043

.000

.001

Cor.

.316

.117

.117

-.557*

-.735**

-.736**

.680**

.718**

.695**

.695**

-.226

-.164

-.144

.713**

.388

-.392

-.724**

.831**

Sig.

.174

.634

.634

.011

.000

.000

.001

.000

.001

.001

.338

.490

.544

.000

.091

.088

.000

.000

Cor.

-.173

-.586

-.586

-.199

-.193

-.191

.265

.238

-.054

-.054

-.794

-.114

-.418

-.047

.031

-.431

-.833

.593**

Sig.

.465

.008

.008

.400

.414

.419

.259

.313

.821

.821

.000

.632

.067

.844

.897

.058

.000

.006

Cor.

.349

**

.070

**

.070

*

-.494

**

-.735

**

-.736

**

.687

**

**

**

.710

.627

.627

**

-.239

-.150

**

-.165

.648

.421

-.377

**

**

-.719

.798**

Sig.

.132

.776

.776

.027

.000

.000

.001

.000

.003

.003

.309

.527

.487

.002

.065

.101

.000

.000

Cor.

.041

-.321

-.321

-.464*

-.428

-.427

.525*

.527*

.352

.352

-.654**

-.243

-.291

.367

.262

-.495*

-.963**

.848**

Sig.

.862

.180

.180

.039

.060

.060

.018

.017

.128

.128

.002

.302

.213

.111

.265

.026

.000

.000

*

*

Cor.

-.041

.321

.321

.464

.428

.427

-.525

-.527

-.352

-.352

.654

.243

.291

-.367

-.262

.495

.963

-.848**

Sig.

.862

.180

.180

.039

.060

.060

.018

.017

.128

.128

.002

.302

.213

.111

.265

.026

.000

.000

-.245

*

.487

Cor.

-.007

.354

.354

.404

.393

.392

*

*

-.487

*

*

-.485

-.304

**

**

-.304

.661

.190

.319

-.317

**

**

.958

-.813**

Sig.

.978

.137

.137

.077

.087

.087

.029

.030

.193

.193

.002

.423

.171

.173

.298

.030

.000

.000

Cor.

-.319

-.044

-.044

.485*

.703**

.704**

-.681**

-.705**

-.581**

-.581**

.306

.144

.227

-.603**

-.338

.430

.743**

-.809**

Sig.

.170

.857

.857

.030

.001

.001

.001

.001

.007

.007

.189

.545

.336

.005

.144

.059

.000

.000

*

*

**

Cor.

-.043

-.323

-.323

-.374

-.390

-.389

.447

.456

.309

.309

-.646

Sig.

.858

.177

.177

.104

.089

.090

.048

.043

.185

.185

.002

*

*

**

-.100

-.378

.316

.161

-.511

-.933

.793**

.676

.100

.175

.497

.021

.000

.000

Cor.

-.245

-.211

-.211

-.433

-.326

-.326

.257

.274

.450

.450

-.517

-.117

-.200

.438

.108

-.380

-.808

.698**

Sig.

.299

.386

.386

.057

.161

.161

.275

.243

.047

.047

.020

.623

.399

.053

.650

.099

.000

.001

[ A2 ]

*

*

**

Appendix A

Spectral Bands MNF B

MNF G

MNF R

MNF NIR

MNF SWIR1

MNF SWIR2

ICA C

ICA B

ICA G

ICA R

ICA NIR

ICAB SWIR1

ICAB SWIR2

C (float)

PH

EC

TDS

T

NO3

NO3-N

PO4

TP

TSS

NTU

DO

BOD

OWQI

AWWQI

CQWI

NFSWQI

SDT

TSI

Cor.

.001

-.027

-.027

-.652**

-.521*

-.521*

.563**

.594**

.708**

.708**

-.440

-.243

-.134

.714**

.253

-.468*

-.878**

.882**

Sig.

.998

.911

.911

.002

.019

.018

.010

.006

.000

.000

.052

.302

.574

.000

.282

.037

.000

.000

-.059

*

.444

.357

.187

.386

-.071

Cor.

.379

**

.669

**

.669

-.259

-.130

-.132

.149

.229

**

.564

**

**

.564

.714

**

.566

Sig.

.099

.002

.002

.270

.586

.580

.529

.331

.010

.010

.000

.804

.050

.009

.122

.430

.092

.766

Cor.

.247

.471*

.471*

-.673**

-.490*

-.492*

.491*

.575**

.980**

.980**

.220

-.209

.257

.981**

.507*

-.180

-.396

.618**

Sig.

.295

.042

.042

.001

.028

.028

.028

.008

.000

.000

.352

.378

.274

.000

.023

.447

.084

.004

Cor.

-.188

.005

.005

-.474

-.361

-.361

.209

.255

.637

.637

-.254

-.064

-.040

.617

.248

-.264

-.688

.656**

Sig.

.428

.984

.984

.035

.118

.117

.377

.278

.003

.003

.279

.788

.868

.004

.292

.261

.001

.002

Cor.

.351

.434

.434

*

.240

.174

.173

-.148

-.125

**

-.100

**

-.100

**

.709

.090

.336

**

-.092

.110

**

**

.381

.766

-.557*

Sig.

.129

.063

.063

.309

.463

.465

.534

.599

.676

.676

.000

.707

.148

.699

.643

.097

.000

.011

Cor.

.215

.007

.007

-.647**

-.575**

-.575**

.602**

.639**

.697**

.697**

-.332

-.301

-.028

.710**

.506*

-.372

-.868**

.908**

Sig.

.362

.977

.977

.002

.008

.008

.005

.002

.001

.001

.153

.197

.905

.000

.023

.106

.000

.000

Cor.

-.159

-.060

-.060

-.550

-.408

-.408

.348

.384

.641

.641

-.399

-.150

-.112

.631

.206

-.380

-.804

.765**

Sig.

.503

.806

.806

.012

.074

.074

.132

.095

.002

.002

.081

.527

.637

.003

.384

.099

.000

.000

Cor.

-.060

.042

*

**

*

*

*

**

.042

-.654

-.492

-.493

.525

.562

**

**

.762

**

**

.762

-.371

-.203

-.113

**

**

.763

.187

*

**

**

-.457

-.801

.825**

Sig.

.800

.863

.863

.002

.028

.027

.018

.010

.000

.000

.107

.391

.636

.000

.429

.043

.000

.000

Cor.

.174

-.035

-.035

.127

.100

.100

.235

.176

-.268

-.268

-.034

-.131

-.075

-.232

-.263

-.068

.266

-.235

Sig.

.464

.887

.887

.593

.675

.676

.318

.459

.253

.253

.888

.581

.752

.324

.263

.774

.257

.318

Cor.

*

.445

.651

.651

-.326

-.245

-.248

.263

.341

.648

.648

.661

-.094

.428

.654

*

.469

.139

.241

.069

Sig.

.049

.003

.003

.160

.297

.293

.262

.141

.002

.002

.001

.693

.060

.002

.037

.559

.306

.772

*

*

Cor.

**

**

**

*

-.079

-.092

-.092

.700

.532

.533

-.560

**

-.609

**

**

-.816

**

**

**

-.816

.284

.272

.006

**

**

-.821

-.386

.395

**

.813

-.872**

Sig.

.739

.709

.709

.001

.016

.016

.010

.004

.000

.000

.225

.246

.978

.000

.093

.085

.000

.000

Cor.

.464*

.074

.074

.292

.101

.101

.001

-.022

-.403

-.403

.303

-.026

.123

-.374

.155

.251

.452*

-.363

Sig.

.039

.764

.764

.211

.673

.670

.997

.928

.078

.078

.195

.913

.604

.104

.515

.285

.046

.115

*

*

.541

-.549

-.581

-.669

-.669

.401

.014

.014

.012

.007

.001

.001

.080

*

*

**

Cor.

-.107

.049

.049

.628

Sig.

.652

.842

.842

.003 **

.541

*

*

*

**

**

**

**

*

.272

.074

-.678

-.426

.395

.246

.757

.001

.061

.085

**

**

.900

.000 *

**

-.903** .000

Cor.

-.046

-.041

-.041

-.638

-.466

-.467

.535

.557

.682

.682

-.463

-.233

-.145

.686

.193

-.471

-.870

.858**

Sig.

.847

.868

.868

.002

.038

.038

.015

.011

.001

.001

.040

.323

.542

.001

.415

.036

.000

.000

[ A3 ]

**

**

Appendix A

Spectral Bands B (float)

G (float)

R (float)

NIR (float)

SWIR1 (float)

SWIR2 (float)

C/B

C/G

C/R

C / NIR

B/C

B/G

B/R

B / NIR

PH

EC

TDS

T

NO3

NO3-N

PO4

TP

TSS

NTU

DO

BOD

OWQI

AWWQI

CQWI

NFSWQI

SDT

TSI

Cor.

-.078

.002

.002

-.657**

-.493*

-.494*

.528*

.574**

.751**

.751**

-.369

-.252

-.051

.754**

.230

-.409

-.836**

.854**

Sig.

.744

.992

.992

.002

.027

.027

.017

.008

.000

.000

.109

.283

.830

.000

.329

.074

.000

.000

Cor.

-.005

.073

.073

**

-.692

*

-.523

*

-.524

*

.540

**

.592

**

.814

**

.814

-.295

-.245

-.018

**

.815

.316

-.394

**

-.809

.860**

Sig.

.983

.768

.768

.001

.018

.018

.014

.006

.000

.000

.207

.298

.941

.000

.174

.086

.000

.000

Cor.

.187

.331

.331

-.699**

-.539*

-.541*

.544*

.612**

.953**

.953**

.012

-.237

.146

.956**

.468*

-.279

-.600**

.757**

Sig.

.429

.166

.166

.001

.014

.014

.013

.004

.000

.000

.961

.315

.538

.000

.037

.233

.005

.000

Cor.

.084

*

.485

*

.485

-.674

-.366

-.368

.413

*

Sig.

.724

.035

.035

.001

.112

.110

.070

Cor.

**

-.300

.377

.377

0.000

.443

.441

.497

.991

.991

.220

-.100

.227

.983

.358

-.218

-.312

.507*

.026

.000

.000

.352

.675

.337

.000

.121

.355

.181

.022

.209

*

-.362

*

-.417

-.474

**

.157

**

.157

.372

-.023

.381

**

.125

-.172

.457

Sig.

.198

.111

.111

1.000

.050

.051

.035

.068

.509

.509

.106

.923

.097

.599

.467

.377

.043

.117

Cor.

-.096

.290

.290

.399

.371

.369

-.492*

-.481*

-.205

-.205

.631**

.066

.408

-.222

-.212

.522*

.806**

-.676**

Sig.

.686

.228

.228

.081

.108

.109

.028

.032

.385

.385

.003

.781

.074

.346

.369

.018

.000

.001

*

Cor.

.342

.034

.034

.317

.223

.223

-.084

-.131

-.528

-.528

.194

-.019

.047

-.501

-.086

.202

.538

-.477*

Sig.

.139

.891

.891

.173

.345

.344

.725

.583

.017

.017

.412

.935

.845

.024

.720

.393

.014

.033

.077

*

*

*

Cor.

-.054

.077

.480

.453

.453

-.408

*

-.444

*

**

-.591

*

**

-.591

.326

.157

.073

*

**

-.590

*

-.466

.314

**

.854

-.807**

Sig.

.822

.754

.754

.032

.045

.045

.074

.050

.006

.006

.161

.507

.759

.006

.038

.177

.000

.000

Cor.

-.297

-.088

-.088

.575**

.572**

.573**

-.637**

-.673**

-.710**

-.710**

.253

.292

.036

-.730**

-.522*

.373

.845**

-.893**

Sig.

.204

.721

.721

.008

.008

.008

.003

.001

.000

.000

.283

.212

.881

.000

.018

.105

.000

.000

**

**

**

*

*

**

Cor.

-.211

-.638

-.638

-.057

-.030

-.028

.164

.114

-.330

-.330

-.859

-.113

-.515

-.316

-.264

-.448

-.609

Sig.

.371

.003

.003

.811

.900

.906

.491

.631

.156

.156

.000

.634

.020

.174

.260

.047

.004

*

Cor.

-.340

-.005

-.005

-.351

-.226

-.226

.089

.138

**

.562

**

.562

-.176

.005

-.025

.535

.088

-.202

.368 .111

*

-.524

.478*

Sig.

.142

.982

.982

.130

.339

.337

.710

.563

.010

.010

.457

.982

.917

.015

.711

.394

.018

.033

Cor.

-.250

.056

.056

.509*

.500*

.499*

-.507*

-.535*

-.556*

-.556*

.312

.226

.051

-.569**

-.588**

.312

.852**

-.836**

Sig.

.287

.821

.821

.022

.025

.025

.023

.015

.011

.011

.180

.339

.831

.009

.006

.181

.000

.000

**

**

**

**

**

**

**

Cor.

-.379

-.134

-.134

.597

.585

.586

-.678

-.713

-.717

-.717

.215

.322

.010

Sig.

.100

.585

.585

.005

.007

.007

.001

.000

.000

.000

.362

.167

.968

**

**

**

**

*

*

-.742

-.553

.367

.000

.011

.111

**

.806

.000 *

**

-.884** .000

Cor.

-.259

-.621

-.621

-.088

-.047

-.045

.148

.105

-.267

-.267

-.857

-.108

-.504

-.258

-.265

-.452

-.640

.399

Sig.

.270

.005

.005

.711

.845

.850

.533

.660

.255

.255

.000

.651

.024

.272

.259

.045

.002

.081

[ A4 ]

Appendix A

Spectral Bands G/C

G/B

G/R

G / NIR

R/C

R/B

R/G

R / NIR

NIR / C

NIR / B

NIR / G

NIR / R

B + C + NIR

B+G+C

PH

EC

TDS

T

NO3

NO3-N

PO4

TP

TSS

NTU

DO

BOD

OWQI

AWWQI

CQWI

NFSWQI

SDT

TSI

Cor.

.082

.020

.020

-.564**

-.494*

-.494*

.426

.477*

.706**

.706**

-.232

-.197

.018

.704**

.513*

-.280

-.807**

.818**

Sig.

.730

.934

.934

.010

.027

.027

.061

.034

.001

.001

.324

.406

.940

.001

.021

.233

.000

.000

*

*

Cor.

.278

-.004

-.004

*

-.541

*

-.534

*

-.533

.514

.552

**

.612

**

.612

-.245

-.260

**

.014

.624

**

.628

-.271

**

-.820

.842**

Sig.

.236

.987

.987

.014

.015

.016

.020

.012

.004

.004

.297

.268

.952

.003

.003

.247

.000

.000

Cor.

-.415

-.250

-.250

.623**

.602**

.603**

-.702**

-.746**

-.788**

-.788**

.100

.354

-.061

-.814**

-.527*

.333

.698**

-.840**

Sig.

.069

.302

.302

.003

.005

.005

.001

.000

.000

.000

.675

.125

.797

.000

.017

.152

.001

.000

**

**

Cor.

-.163

-.584

-.584

-.196

-.159

-.157

.253

.220

-.130

-.130

-.851

-.175

-.456

-.118

-.096

-.471

-.771

.554*

Sig.

.492

.009

.009

.407

.504

.508

.282

.352

.585

.585

.000

.460

.043

.622

.686

.036

.000

.011

Cor.

.263

.325

.325

**

-.689

**

-.563

**

-.565

**

.566

**

.638

**

.928

**

**

.928

.029

*

-.269

.168

**

.934

*

.547

*

-.254

**

**

-.611

.778**

Sig.

.262

.175

.175

.001

.010

.009

.009

.002

.000

.000

.902

.251

.479

.000

.013

.280

.004

.000

Cor.

.320

.320

.320

-.681**

-.581**

-.582**

.594**

.662**

.903**

.903**

.028

-.293

.169

.914**

.577**

-.249

-.615**

.788**

Sig.

.169

.182

.182

.001

.007

.007

.006

.001

.000

.000

.905

.210

.477

.000

.008

.290

.004

.000

**

**

Cor.

.325

.364

.364

-.684

-.571

Sig.

.162

.126

.126

.001

.009

Cor.

.149

-.171

**

**

-.573

.008 *

*

-.171

-.594

-.514

-.514

**

**

**

**

.609

.678

.922

.004

.001

.000

.000

.627

*

.510

*

.510

**

.605

**

.922

.058 .808 *

**

*

.539

-.258

-.574

.763**

.014

.273

.008

.000

-.294

.173

.934

.209

.465

.000

-.185

*

.530

.320

-.485

-.552

-.365

*

**

**

-.928

.915**

Sig.

.532

.483

.483

.006

.020

.020

.005

.003

.022

.022

.012

.113

.436

.016

.170

.030

.000

.000

Cor.

.234

.704**

.704**

.012

.033

.031

-.127

-.075

.354

.354

.829**

.076

.470*

.346

.152

.366

.644**

-.362

Sig.

.320

.001

.001

.959

.889

.897

.593

.752

.126

.126

.000

.750

.037

.136

.521

.112

.002

.116

*

**

**

**

**

Cor.

.265

.672

.672

.073

.078

.076

-.158

-.117

.234

.234

.819

.072

.454

.230

.112

.390

.716

-.439

Sig.

.259

.002

.002

.759

.743

.750

.507

.622

.321

.321

.000

.763

.044

.330

.638

.089

.000

.053

Cor.

**

**

.175

.581

.581

.198

.199

.197

-.267

-.242

.038

.038

**

.780

.103

.410

.032

-.049

.424

**

.840

-.598**

Sig.

.460

.009

.009

.404

.401

.406

.254

.305

.873

.873

.000

.665

.072

.893

.838

.062

.000

.005

Cor.

-.007

.354

.354

.404

.393

.392

-.487*

-.485*

-.304

-.304

.661**

.190

.319

-.317

-.245

.487*

.958**

-.813**

Sig.

.978

.137

.137

.077

.087

.029

.030

.193

.193

.002

.423

.171

.173

.298

.030

.000

.000

**

.087

Cor.

-.017

.097

.097

-.685

-.501

-.502

.537

.580

.812

.812

-.308

-.229

-.048

.813

.250

-.423

-.782

.831**

Sig.

.944

.693

.693

.001

.025

.024

.015

.007

.000

.000

.186

.332

.841

.000

.289

.063

.000

.000

**

*

*

*

*

*

*

**

**

**

**

**

**

**

Cor.

-.007

.049

.049

-.678

-.506

-.507

.539

.579

.779

.779

-.353

-.242

-.068

.781

.274

-.432

-.825

.857**

Sig.

.977

.843

.843

.001

.023

.022

.014

.007

.000

.000

.127

.304

.777

.000

.243

.057

.000

.000

[ A5 ]

**

**

**

Appendix A

Spectral Bands B + G + NIR

B+G+R

B + G + R + NIR

B+R+C

B + R + NIR

C + B + G + NIR

C+B+G+R C+B+G+R+ NIR C + B + R + NIR

C + G + R + NIR

G + C + NIR

G+C+R

G + R + NIR

R + C + NIR

PH

EC

TDS

T

NO3

NO3-N

PO4

TP

TSS

NTU

DO

BOD

OWQI

AWWQI

CQWI

NFSWQI

SDT

TSI

Cor.

.014

.119

.119

-.691**

-.511*

-.513*

.530*

.579**

.834**

.834**

-.268

-.233

-.015

.835**

.311

-.397

-.776**

.836**

Sig.

.953

.629

.629

.001

.021

.021

.016

.007

.000

.000

.253

.323

.951

.000

.183

.083

.000

.000

Cor.

.072

.179

.179

**

-.713

*

-.529

*

-.530

*

.551

**

.608

**

.881

**

.881

-.196

-.246

.031

**

.883

.366

-.375

**

-.740

.832**

Sig.

.761

.463

.463

.000

.017

.016

.012

.004

.000

.000

.407

.296

.896

.000

.112

.103

.000

.000

Cor.

.079

.205

.205

-.714**

-.528*

-.529*

.547*

.605**

.897**

.897**

-.165

-.241

.049

.898**

.373

-.361

-.717**

.818**

Sig.

.742

.400

.400

.000

.017

.016

.013

.005

.000

.000

.488

.307

.837

.000

.106

.117

.000

.000

Cor.

.070

.183

.183

-.717

-.529

-.530

.561

.617

.883

.883

-.201

-.248

.025

.886

.348

-.385

-.736

.830**

Sig.

.769

.454

.454

.000

.017

.016

.010

.004

.000

.000

.396

.292

.918

.000

.133

.094

.000

.000

Cor.

.102

.267

.267

**

**

-.719

*

*

-.524

*

*

-.526

*

*

.544

**

**

.608

**

**

.932

**

**

.932

-.090

-.235

.090

**

**

.933

.387

-.334

**

**

-.658

.786**

Sig.

.668

.269

.269

.000

.018

.017

.013

.004

.000

.000

.707

.319

.706

.000

.092

.150

.002

.000

Cor.

.006

.091

.091

-.688**

-.511*

-.512*

.539*

.584**

.813**

.813**

-.305

-.237

-.038

.815**

.291

-.415

-.796**

.846**

Sig.

.979

.710

.710

.001

.021

.021

.014

.007

.000

.000

.191

.313

.872

.000

.214

.069

.000

.000

Cor.

.059

.152

.152

-.710

-.527

-.529

.556

.609

.862

.862

-.232

-.248

.008

.865

.345

-.392

-.762

.842**

Sig.

.806

.534

.534

.000

.017

.017

.011

.004

.000

.000

.324

.292

.973

.000

.136

.088

.000

.000

Cor.

.065

.176

.176

**

**

-.712

*

*

-.527

*

*

-.529

*

*

.553

**

**

.608

**

**

.878

**

**

.878

-.204

-.244

.025

**

**

.880

.352

-.380

**

**

-.742

.831**

Sig.

.787

.472

.472

.000

.017

.017

.012

.004

.000

.000

.388

.301

.918

.000

.128

.099

.000

.000

Cor.

.078

.214

.214

-.718**

-.527*

-.528*

.555*

.614**

.902**

.902**

-.161

-.241

.047

.904**

.357

-.368

-.708**

.814**

Sig.

.744

.378

.378

.000

.017

.017

.011

.004

.000

.000

.496

.306

.844

.000

.122

.111

.000

.000

Cor.

.091

.205

.205

-.717

-.534

-.535

.557

.615

.896

.896

-.166

-.246

.049

.899

.380

-.364

-.721

.824**

Sig.

.704

.401

.401

.000

.015

.015

.011

.004

.000

.000

.483

.296

.838

.000

.098

.115

.000

.000

Cor.

.031

.113

.113

**

**

-.694

*

*

-.521

*

*

-.522

*

*

.547

**

**

.594

**

**

.830

**

**

.830

-.278

-.241

-.019

**

**

.832

.320

-.403

**

**

-.788

.848**

Sig.

.896

.646

.646

.001

.018

.018

.013

.006

.000

.000

.236

.305

.936

.000

.169

.078

.000

.000

Cor.

.085

.178

.178

-.716**

-.535*

-.536*

.562**

.618**

.880**

.880**

-.199

-.251

.030

.883**

.374

-.378

-.745**

.838**

Sig.

.722

.467

.467

.000

.015

.015

.010

.004

.000

.000

.400

.285

.900

.000

.105

.101

.000

.000

**

Cor.

.113

.246

.246

-.718

-.533

-.535

.549

.612

.920

.920

-.109

-.242

.084

.922

.409

-.336

-.684

.804**

Sig.

.635

.311

.311

.000

.015

.015

.012

.004

.000

.000

.646

.304

.726

.000

.073

.147

.001

.000

**

*

*

*

*

*

*

**

**

**

**

**

**

**

Cor.

.122

.270

.270

-.723

-.533

-.535

.559

.623

.932

.932

-.089

-.242

.091

.935

.399

-.336

-.661

.793**

Sig.

.608

.264

.264

.000

.015

.015

.010

.003

.000

.000

.710

.304

.702

.000

.082

.147

.002

.000

[ A6 ]

**

**

**

Appendix A

Spectral Bands B+G

B + NIR

B+R

C+B

C+G

C + NIR

C+R

G + NIR

G+R

R + NR

B / NIR + B

B / NIR + C

B / NIR + G

B / NIR + NIR

PH

EC

TDS

T

NO3

NO3-N

PO4

TP

TSS

NTU

DO

BOD

OWQI

AWWQI

CQWI

NFSWQI

SDT

TSI

Cor.

-.002

.065

.065

-.680**

-.507*

-.508*

.531*

.574**

.793**

.793**

-.329

-.239

-.051

.795**

.291

-.419

-.814**

.852**

Sig.

.993

.791

.791

.001

.023

.022

.016

.008

.000

.000

.157

.310

.831

.000

.213

.066

.000

.000

Cor.

-.013

.166

.166

**

-.688

*

-.493

*

-.494

*

.513

**

.564

**

.861

**

.861

-.220

-.212

.004

**

.860

.269

-.384

**

-.720

.795**

Sig.

.957

.496

.496

.001

.027

.027

.021

.010

.000

.000

.350

.370

.987

.000

.251

.095

.000

.000

Cor.

.095

.233

.233

-.721**

-.528*

-.530*

.553*

.615**

.915**

.915**

-.134

-.244

.065

.916**

.380

-.355

-.693**

.808**

Sig.

.690

.338

.338

.000

.017

.016

.011

.004

.000

.000

.573

.301

.785

.000

.099

.124

.001

.000

Cor.

-.045

.014

.014

-.664

-.489

-.490

.535

.569

.743

.743

-.402

-.236

-.106

.745

.211

-.457

-.833

.849**

Sig.

.851

.955

.955

.001

.029

.028

.015

.009

.000

.000

.079

.317

.656

.000

.371

.043

.000

.000

Cor.

.015

.056

.056

**

**

-.683

*

*

-.517

*

*

-.518

*

*

.548

**

**

.590

**

**

.785

**

**

.785

-.343

-.248

-.058

**

**

.789

.299

*

-.427

**

**

-.828

.865**

Sig.

.949

.821

.821

.001

.020

.019

.012

.006

.000

.000

.139

.291

.808

.000

.200

.061

.000

.000

Cor.

.028

.159

.159

-.700**

-.518*

-.519*

.556*

.604**

.857**

.857**

-.239

-.231

-.005

.859**

.288

-.399

-.744**

.822**

Sig.

.907

.515

.515

.001

.019

.019

.011

.005

.000

.000

.311

.326

.983

.000

.218

.081

.000

.000

**

*

*

**

Cor.

.116

.233

.233

-.726

-.538

-.540

.570

Sig.

.626

.336

.336

.000

.014

.014

.009

Cor.

.053

.162

.162

**

-.699

*

-.524

*

-.525

*

.535

**

.631

.914

.914

-.135

-.252

.065

.917

.392

-.358

-.698

.817**

.003

.000

.000

.570

.283

.785

.000

.087

.121

.001

.000

**

.589

**

**

.865

**

**

.865

-.210

-.235

.025

**

**

.866

.360

-.369

**

**

-.751

.830**

Sig.

.823

.508

.508

.001

.018

.017

.015

.006

.000

.000

.374

.318

.917

.000

.119

.109

.000

.000

Cor.

.108

.216

.216

-.719**

-.536*

-.538*

.555*

.616**

.904**

.904**

-.146

-.249

.064

.906**

.404

-.352

-.713**

.822**

Sig.

.651

.373

.373

.000

.015

.015

.011

.004

.000

.000

.540

.290

.790

.000

.077

.127

.000

.000

**

*

*

*

**

Cor.

.168

.357

.357

-.713

-.522

-.524

.535

.608

Sig.

.478

.134

.134

.000

.018

.018

.015

.004

.388

*

Cor.

.000

.101

.101

**

-.565

-.061

-.064

.453

**

.967

.967

.036

-.229

.163

.969

.447

-.274

-.564

.733**

.000

.000

.880

.330

.492

.000

.048

.243

.010

.000

**

.575

**

**

.575

-.398

-.094

-.231

**

**

*

.577

.103

**

-.620

**

**

-.625

.679**

Sig.

.999

.680

.680

.009

.797

.787

.091

.045

.008

.008

.082

.695

.327

.008

.664

.004

.003

.001

Cor.

-.025

-.013

-.013

-.663**

-.500*

-.501*

.560*

.591**

.718**

.718**

-.437

-.249

-.127

.724**

.212

-.474*

-.857**

.867**

Sig.

.917

.959

.959

.001

.025

.025

.010

.006

.000

.000

.054

.289

.593

.000

.369

.035

.000

.000

**

*

*

*

**

**

**

.808

.808

-.303

-.246

-.030

.810

.335

-.405

-.812

.860**

.000

.000

.194

.295

.901

.000

.148

.077

.000

.000

.950

.392

-.088

-.225

.455*

.000

.088

.714

.339

.044

Cor.

.032

.081

.081

-.687

-.520

-.521

.539

.585

Sig.

.895

.740

.740

.001

.019

.018

.014

.007

Cor.

.152

*

.554

*

.554

-.585

-.406

-.409

.376

*

.455

.956

.956

.342

-.113

.309

Sig.

.522

.014

.014

.007

.075

.074

.103

.044

.000

.000

.140

.635

.186

**

**

[ A7 ]

**

**

**

**

Appendix A

Spectral Bands B / NIR + R

B/R+B

B/R+C

B/R+G

B / R + NIR

B/R+R

NIR / B + B

NIR / B + C

NIR / B + G

NIR / B + NIR

NIR / B + R

PH

EC

TDS

T

NO3

NO3-N

PO4

TP

TSS

NTU

DO

BOD

OWQI

AWWQI

CQWI

NFSWQI

SDT

TSI

Cor.

.166

.318

.318

-.723**

-.533*

-.534*

.553*

.623**

.954**

.954**

-.016

-.244

.136

.957**

.447*

-.301

-.611**

.767**

Sig.

.483

.184

.184

.000

.016

.015

.011

.003

.000

.000

.948

.300

.568

.000

.048

.197

.004

.000

*

*

Cor.

-.065

.030

.030

**

-.661

*

-.476

*

-.477

.510

.548

**

.756

**

.756

-.378

-.223

-.092

**

.756

.206

-.444

**

-.812

.831**

Sig.

.786

.901

.901

.002

.034

.033

.022

.012

.000

.000

.101

.344

.699

.000

.384

.050

.000

.000

Cor.

-.024

-.007

-.007

-.665**

-.501*

-.502*

.560*

.591**

.724**

.724**

-.430

-.249

-.123

.729**

.215

-.471*

-.854**

.866**

Sig.

.921

.979

.979

.001

.024

.024

.010

.006

.000

.000

.059

.290

.607

.000

.363

.036

.000

.000

**

Cor.

.032

.085

.085

-.688

-.521

-.522

.539

.585

.810

.810

-.299

-.246

-.027

.812

.337

-.403

-.810

.859**

Sig.

.892

.731

.731

.001

.019

.018

.014

.007

.000

.000

.200

.296

.909

.000

.147

.078

.000

.000

Cor.

.155

*

.566

*

.566

.391

-.070

-.198

.431

**

-.572

*

-.396

*

-.398

*

.362

**

.442

**

**

.949

**

**

.949

.366

-.106

.321

**

**

.942

**

Sig.

.514

.012

.012

.008

.084

.082

.117

.051

.000

.000

.113

.656

.168

.000

.088

.768

.404

.058

Cor.

.167

.321

.321

-.723**

-.532*

-.534*

.552*

.622**

.955**

.955**

-.012

-.243

.138

.958**

.448*

-.299

-.608**

.765**

Sig.

.481

.180

.180

.000

.016

.015

.012

.003

.000

.000

.961

.301

.561

.000

.048

.201

.004

.000

*

*

Cor.

-.064

.031

.031

-.661

-.477

-.478

.511

.549

.756

.756

-.377

-.224

-.092

.757

.207

-.444

-.812

.832**

Sig.

.790

.900

.900

.002

.033

.033

.021

.012

.000

.000

.101

.343

.700

.000

.381

.050

.000

.000

Cor.

-.022

-.006

-.006

**

**

-.665

*

*

-.501

*

*

-.502

*

.561

**

.592

**

**

.724

**

**

.724

-.429

-.249

-.122

**

**

.730

.216

*

-.471

**

**

-.854

.866**

Sig.

.926

.982

.982

.001

.024

.024

.010

.006

.000

.000

.059

.289

.608

.000

.360

.036

.000

.000

Cor.

.033

.085

.085

-.688**

-.521*

-.522*

.539*

.586**

.810**

.810**

-.299

-.246

-.027

.812**

.337

-.403

-.810**

.859**

Sig.

.890

.730

.730

.001

.019

.018

.014

.007

.000

.000

.200

.295

.910

.000

.146

.078

.000

.000

Cor.

.158

*

.565

*

.565

-.573

-.399

-.401

.365

*

.445

.950

.950

.363

-.108

.319

.944

.394

-.072

-.202

.435

Sig.

.506

.012

.012

.008

.082

.080

.113

.049

.000

.000

.116

.650

.170

.000

.086

.761

.392

.055

**

**

*

*

*

**

**

**

**

Cor.

.168

.321

.321

-.722

-.532

-.534

.552

.623

.955

.955

-.012

-.244

.138

.958

.448

-.299

-.608

.765**

Sig.

.480

.180

.180

.000

.016

.015

.012

.003

.000

.000

.961

.301

.561

.000

.048

.200

.004

.000

* Correlation is significant at 0.05 level (2-tailed) as highlighted with Yellow color. ** Correlation is significant at 0.01 level (2-tailed) as highlighted with Red color.

[ A8 ]

**

**

**

*

**

Appendix A

Table (A-2): Pearson Correlation between Water Quality (WQI, WQPs, SDT and TSI) and Spectral Bands for Spring Season. Spectral Bands C

B

G

R

NIR

SWIR1

SWIR2

C (float)

B (float)

G (float)

R (float)

NIR (float)

SWIR1 (float)

PH

EC

TDS

T

NTU

TSS

NO3

NO3-N

PO4

TP

DO

BOD

OWQI

CWQI

AWWQI

NFSWQI

SD

TSI

Cor.

-.378

-.129

-.242

.601**

.007

.007

-.422

-.428

-.172

-.165

-.352

.415

.095

.154

-.182

.198

-.707**

.879**

Sig.

.101

.587

.303

.005

.978

.978

.064

.060

.469

.486

.128

.069

.691

.516

.442

.403

.000

.000

Cor.

-.423

-.105

-.222

**

.652

.009

.009

-.415

-.421

-.179

-.176

-.343

.441

.107

.159

-.186

.198

**

-.747

.906**

Sig.

.063

.659

.347

.002

.970

.970

.069

.064

.451

.457

.139

.052

.653

.503

.432

.403

.000

.000

Cor.

-.439

-.055

-.175

.714**

-.048

-.048

-.420

-.426

-.165

-.165

-.297

.529*

.090

.123

-.187

.154

-.816**

.952**

Sig.

.053

.817

.461

.000

.842

.842

.065

.061

.487

.487

.203

.017

.707

.605

.429

.516

.000

.000

*

**

**

Cor.

-.450

-.046

-.158

.615

-.089

-.089

-.437

-.443

-.124

-.121

-.266

.582

Sig.

.046

.847

.506

.004

.708

.708

.054

.051

.603

.612

.258

.007

Cor.

-.218

-.341

-.393

.311

-.193

-.193

**

-.568

**

-.566

.423

.437

-.052

.620

**

.055 .818

**

.075

-.161

.149

-.797

.752

.498

.530

.000

.000

*

-.407

.304

-.109

-.526

-.487

.950**

*

.589**

Sig.

.355

.141

.086

.182

.414

.414

.009

.009

.063

.054

.826

.004

.030

.075

.193

.648

.017

.006

Cor.

.005

-.434

-.388

-.381

.023

.023

-.308

-.300

.351

.360

-.019

.005

-.360

-.202

.305

.163

.387

-.328

Sig.

.982

.056

.091

.097

.924

.924

.187

.199

.130

.119

.936

.983

.119

.393

.191

.493

.092

.158

Cor.

.039

-.416

-.370

-.407

.031

.031

-.244

-.236

.348

.357

.031

.017

-.366

-.204

.307

.179

.439

-.376

Sig.

.872

.068

.108

.075

.898

.898

.300

.316

.132

.123

.897

.942

.113

.387

.188

.451

.053

.103

Cor.

-.378

-.121

-.234

**

.605

.004

.004

-.422

-.429

-.180

-.173

-.358

.411

.103

.158

-.189

.201

**

-.713

.883**

Sig.

.100

.612

.322

.005

.987

.987

.064

.059

.449

.466

.122

.072

.667

.505

.424

.396

.000

.000

Cor.

-.423

-.107

-.224

.652**

.004

.004

-.412

-.419

-.183

-.180

-.343

.440

.111

.159

-.192

.197

-.749**

.907**

Sig.

.063

.654

.343

.002

.985

.985

.071

.066

.441

.447

.138

.052

.643

.502

.418

.406

.000

.000

*

Cor.

-.442

-.052

-.171

.714

-.048

-.048

-.422

-.429

-.161

-.161

-.297

.530

.087

.119

-.183

.154

-.818

.953**

Sig.

.051

.829

.470

.000

.839

.839

.064

.059

.498

.497

.204

.016

.716

.616

.439

.518

.000

.000

*

-.045

Sig.

.047

Cor.

-.232

Sig.

.325

.147

Cor.

-.449

**

**

-.156

.615

.852

.511

.004

.691

.691

-.336

-.386

.305

-.191

-.191

.093

.191

.421

.421

-.410

-.400

.003

.073

.081

.991

Cor.

.000

-.462

Sig.

1.000

.040

*

-.095

-.095

-.433

-.439

-.129

-.125

.056

.053

.589

.599

.255

.008

-.584**

-.582**

.430

.442

-.043

.628**

.007

.007

.058

.051

.856

.003

.003

-.264

-.256

.328

.342

-.024

.991

.261

.277

.158

.140

.921

[ A9 ]

-.267

.579

**

.060

.148

**

**

-.801

.952**

.076

-.167

.801

.750

.482

.534

.000

.000

-.491*

-.408

.311

-.100

-.523*

.584**

.028

.074

.181

.676

.018

.007

-.003

-.329

-.204

.275

.150

.389

-.336

.990

.157

.387

.240

.529

.090

.148

Appendix A

Spectral Bands SWIR2 (float)

ICA C

ICA B

ICA G

ICA R

ICA NIR

ICA SWIR1

ICA SWIR2

MNF C

MNF B

MNF G

MNF R

MNF NIR

MNF SWIR1

PH

EC

TDS

T

NTU

TSS

NO3

NO3-N

PO4

TP

DO

BOD

OWQI

CWQI

AWWQI

NFSWQI

SD

TSI

Cor.

.038

-.462*

-.415

-.406

.045

.045

-.222

-.214

.328

.336

.037

.012

-.342

-.173

.289

.142

.442

-.381

Sig.

.872

.040

.069

.075

.851

.851

.347

.366

.157

.147

.877

.960

.140

.467

.216

.549

.051

.097

Cor.

*

.527

-.007

**

.111

-.792

.032

.032

.359

.364

.151

.165

*

.227

-.558

-.097

-.113

.163

-.145

**

.834

-.911**

Sig.

.017

.976

.642

.000

.895

.895

.120

.115

.526

.488

.337

.011

.683

.634

.491

.541

.000

.000

Cor.

-.440

-.019

-.113

.375

-.042

-.042

-.380

-.386

-.227

-.218

-.318

.435

.179

.190

-.239

.260

-.621**

.835**

Sig.

.052

.936

.636

.103

.862

.862

.098

.093

.336

.356

.173

.055

.451

.422

.309

.268

.003

.000

Cor.

-.331

-.007

-.108

.887

.049

.049

-.198

-.203

-.118

-.141

-.166

.310

.069

.101

-.105

.008

-.690

.640**

Sig.

.155

.976

.649

.000

.836

.836

.402

.391

.621

.554

.483

.184

.771

.673

.660

.973

.001

.002

.083

.247

*

.486

-.122

-.212

-.009

-.210

-.288

.217

Cor.

-.009

.243

**

.235

.175

-.347

-.347

-.051

-.048

.087

**

Sig.

.971

.301

.319

.460

.133

.133

.832

.839

.715

.728

.295

.030

.609

.370

.969

.373

.218

.359

Cor.

.078

-.105

-.124

.289

-.203

-.203

-.134

-.133

.512*

.516*

.250

.429

-.536*

-.565**

.409

-.472*

-.400

.272

Sig.

.743

.660

.601

.217

.391

.391

.574

.576

.021

.020

.288

.059

.015

.009

.073

.035

.080

.246

Cor.

-.196

.050

.038

.221

-.001

-.001

-.199

-.202

-.171

-.170

-.279

-.120

.220

.162

-.160

-.045

-.322

.304

Sig.

.407

.834

.875

.348

.997

.997

.401

.393

.470

.473

.233

.614

.352

.495

.500

.850

.166

.192

Cor.

.185

.400

.373

.172

-.021

-.021

.435

.429

-.277

-.282

.155

-.035

.260

.135

-.235

-.146

-.120

.073

Sig.

.436

.081

.105

.469

.929

.929

.055

.059

.238

.228

.514

.883

.269

.570

.318

.539

.615

.760

Cor.

-.338

-.080

-.195

.673**

.059

.059

-.339

-.347

-.265

-.262

-.384

.284

.194

.242

-.249

.182

-.699**

.840**

Sig.

.145

.737

.411

.001

.806

.806

.144

.134

.260

.265

.095

.225

.412

.303

.291

.443

.001

.000

Cor.

*

.492

-.001

.116

-.701

.066

.066

.397

.404

.169

.172

.264

-.569

-.108

-.117

.192

-.151

.837

Sig.

.028

.996

.626

.001

.781

.781

.083

.078

.477

.467

.261

.009

.652

.624

.417

.525

.000

Cor.

-.198

.104

**

**

.036

.804

-.006

-.006

-.047

-.048

-.100

-.132

-.011

**

.228

.064

.068

**

-.965** .000

*

-.093

-.080

-.502

.354

Sig.

.402

.663

.880

.000

.981

.981

.845

.840

.674

.579

.963

.333

.788

.777

.696

.738

.024

.126

Cor.

-.106

-.224

-.251

.120

.345

.345

-.008

-.012

-.133

-.138

-.280

-.345

.147

.245

-.037

.224

.058

.000

Sig.

.655

.341

.287

.614

.136

.136

.973

.961

.575

.562

.231

.137

.535

.297

.878

.342

.808

.998

*

Cor.

.120

-.022

-.029

.402

-.065

-.065

.032

.032

.319

.307

.185

.085

-.313

-.346

.279

-.449

-.269

.044

Sig.

.613

.925

.903

.079

.784

.784

.894

.892

.171

.189

.436

.721

.180

.136

.233

.047

.252

.854

*

Cor.

.187

-.166

-.146

-.099

-.069

-.069

.056

.062

.422

.418

.341

.227

-.464

-.392

.366

-.132

.218

-.293

Sig.

.430

.485

.540

.679

.774

.774

.816

.796

.064

.067

.142

.337

.039

.088

.113

.580

.356

.211

[ A10 ]

Appendix A

Spectral Bands MNF SWIR2

PCA C

PCA B

PCA G

PCA R

PCA NIR

PCA SWIR1

PCA SWIR2

LSWI

MNDWI

MSI

NBR

NDVI

DVI

PH

EC

TDS

T

NTU

TSS

NO3

NO3-N

PO4

TP

DO

BOD

OWQI

CWQI

AWWQI

NFSWQI

SD

TSI

Cor.

.233

.152

.150

-.056

-.049

-.049

.328

.329

.043

.039

.315

.129

-.090

-.108

.040

-.073

.157

-.187

Sig.

.324

.523

.529

.814

.839

.839

.158

.157

.856

.872

.176

.588

.706

.651

.868

.761

.508

.430

Cor.

.434

.078

.195

**

-.668

.043

.043

.433

.439

.152

.150

.307

*

-.078

-.520

-.117

.175

-.167

**

.791

-.941**

Sig.

.056

.743

.409

.001

.858

.858

.057

.053

.522

.528

.189

.019

.744

.624

.460

.482

.000

.000

Cor.

-.207

-.270

-.273

-.088

-.222

-.222

-.466*

-.460*

.424

.436

.064

.554*

-.453*

-.389

.305

.002

-.177

.255

Sig.

.382

.249

.244

.713

.347

.347

.039

.041

.063

.055

.790

.011

.045

.090

.192

.992

.455

.278

*

**

**

Cor.

-.457

.055

-.055

.631

-.128

-.128

-.375

-.381

-.153

-.153

-.213

.613

Sig.

.043

.817

.818

.003

.592

.592

.103

.098

.520

.519

.368

.004

Cor.

*

-.496

.071

-.040

**

.758

.067

.067

-.242

-.247

-.405

-.428

-.300

.337

.093

.075

-.192

.106

-.833

.962**

.696

.755

.418

.658

.000

.000

.345

.392

-.367

.346

**

**

-.614

.705**

Sig.

.026

.766

.868

.000

.779

.779

.304

.294

.076

.060

.198

.146

.136

.088

.112

.135

.004

.001

Cor.

-.031

.141

.107

.623**

-.184

-.184

-.027

-.026

.146

.123

.192

.336

-.172

-.223

.089

-.342

-.456*

.241

Sig.

.896

.554

.652

.003

.436

.436

.909

.912

.539

.605

.417

.148

.470

.344

.710

.140

.043

.305

*

Cor.

.207

.193

.252

-.468

-.254

-.254

.113

.117

.072

.083

.251

.094

-.065

-.146

.007

-.105

.278

-.281

Sig.

.381

.416

.284

.037

.280

.280

.635

.622

.764

.727

.285

.695

.786

.539

.975

.660

.235

.229

Cor.

.237

.012

.024

-.080

-.059

-.059

.231

.235

.233

.227

.356

.167

-.278

-.255

.205

-.117

.211

-.275

Sig.

.314

.959

.921

.738

.806

.806

.327

.320

.323

.336

.123

.482

.236

.279

.387

.622

.373

.240

Cor.

-.111

.121

.044

.640**

-.271

-.271

-.151

-.156

.018

.021

.002

.444*

-.071

-.212

-.066

-.285

-.798**

.734**

Sig.

.640

.611

.854

.002

.247

.247

.526

.511

.939

.930

.995

.050

.765

.369

.783

.223

.000

.000

**

Cor.

-.176

.305

.219

.682

Sig.

.457

.191

.354

.001 *

Cor.

-.275

.288

.208

.526

-.078

-.078

.016

.007

-.324

-.333

-.142

.193

.295

.159

-.313

-.097

-.715

.684**

.742

.742

.947

.976

.163

.152

.551

.416

.206

.503

.179

.145

.145

.078

.070

**

-.617

**

-.634

-.326

-.088

**

.601

*

.545

**

.685

.000

.001

*

.276

-.402

.472*

-.520

Sig.

.241

.218

.380

.017

.541

.541

.743

.771

.004

.003

.160

.713

.005

.013

.019

.239

.079

.035

Cor.

-.137

.128

.057

.631**

-.249

-.249

-.178

-.184

-.015

-.014

-.040

.372

-.020

-.166

-.089

-.286

-.797**

.723**

Sig.

.564

.592

.811

.003

.290

.290

.452

.438

.949

.954

.866

.107

.932

.485

.708

.221

.000

.000

Cor.

*

.464

-.183

-.075

-.776

.188

.188

.312

.320

.406

.414

.332

-.295

-.355

-.223

.439

-.195

.848

Sig.

.039

.441

.753

.000

.428

.428

.180

.169

.075

.070

.153

.207

.125

.346

.053

.409

.000

*

**

**

**

.000

Cor.

-.413

.044

-.074

.674

-.103

-.103

-.334

-.341

-.206

-.205

-.245

.542

.136

.115

-.235

.118

-.829

.960**

Sig.

.070

.855

.756

.001

.665

.665

.150

.141

.384

.386

.298

.014

.569

.629

.318

.621

.000

.000

[ A11 ]

**

-.876**

Appendix A

Spectral Bands IPVI

NDMI

RVI

B+C

B+G

B+R

B + NIR

G+C

G+R

G + NIR

R+C

R + NIR

C + NIR

B+G+C

PH

EC

TDS

T

NTU

TSS

NO3

NO3-N

PO4

TP

DO

BOD

OWQI

CWQI

AWWQI

NFSWQI

SD

TSI

Cor.

.464*

-.183

-.075

-.776**

.188

.188

.312

.320

.406

.414

.332

-.295

-.355

-.223

.439

-.195

.848**

-.876**

Sig.

.039

.441

.753

.000

.428

.428

.180

.169

.075

.070

.153

.207

.125

.346

.053

.409

.000

.000

*

Cor.

-.464

.183

.075

**

.776

-.188

-.188

-.312

-.320

-.406

-.414

-.332

.295

.355

.223

-.439

.195

**

-.848

.876**

Sig.

.039

.441

.753

.000

.428

.428

.180

.169

.075

.070

.153

.207

.125

.346

.053

.409

.000

.000

Cor.

.440

-.208

-.107

-.763**

.222

.222

.271

.280

.440

.447*

.335

-.229

-.394

-.225

.479*

-.181

.831**

-.836**

Sig.

.052

.378

.654

.000

.348

.348

.247

.232

.052

.048

.149

.331

.085

.340

.033

.446

.000

.000

Cor.

-.404

-.116

-.231

.631

.008

.008

-.418

-.425

-.176

-.172

-.347

.431

.102

.157

-.185

.198

-.731

.896**

Sig.

.077

.627

.327

.003

.973

.973

.066

.062

.458

.469

.133

.058

.668

.508

.435

.402

.000

.000

-.315

*

.499

.096

.137

Cor.

-.435

-.074

-.193

**

**

.694

-.027

-.027

-.420

-.426

-.170

-.170

-.188

.171

**

**

-.794

.938**

Sig.

.055

.758

.416

.001

.909

.909

.065

.061

.472

.474

.176

.025

.686

.566

.429

.472

.000

.000

Cor.

-.444*

-.069

-.184

.636**

-.052

-.052

-.433

-.439

-.147

-.144

-.298

.533*

.076

.109

-.172

.170

-.786**

.943**

Sig.

.050

.771

.436

.003

.828

.828

.057

.053

.538

.546

.201

.015

.751

.648

.468

.475

.000

.000

*

Cor.

-.412

-.158

-.270

.631

-.029

-.029

-.472

-.478

-.075

-.070

-.310

.505

.000

.061

-.105

.152

-.755

.906**

Sig.

.071

.507

.250

.003

.903

.903

.036

.033

.753

.768

.183

.023

.999

.798

.661

.521

.000

.000

-.316

*

.498

.092

.133

Cor.

-.424

-.078

-.196

**

**

.685

-.032

-.032

*

-.423

*

-.430

-.168

-.166

-.187

.168

**

**

-.789

.936**

Sig.

.063

.745

.407

.001

.894

.894

.063

.059

.479

.484

.175

.025

.700

.575

.430

.478

.000

.000

Cor.

-.447*

-.051

-.168

.670**

-.068

-.068

-.430

-.436

-.146

-.145

-.284

.557*

.073

.101

-.175

.153

-.811**

.955**

Sig.

.048

.831

.480

.001

.777

.777

.058

.055

.539

.543

.226

.011

.758

.672

.459

.521

.000

.000

*

Cor.

-.432

-.090

-.207

.696

-.066

-.066

-.454

-.460

-.104

-.102

-.281

.561

.027

.067

-.138

.130

-.816

.949**

Sig.

.057

.705

.381

.001

.781

.781

.044

.041

.663

.667

.229

.010

.910

.778

.561

.584

.000

.000

-.296

*

Cor.

-.432

-.073

-.187

**

**

.618

-.060

-.060

*

-.437

*

-.443

-.141

-.136

.536

.068

.101

-.169

.166

**

**

-.778

.939**

Sig.

.057

.759

.431

.004

.803

.803

.054

.050

.554

.566

.204

.015

.775

.670

.475

.483

.000

.000

Cor.

-.438

-.085

-.193

.600**

-.106

-.106

-.470*

-.475*

-.060

-.056

-.249

.609**

-.011

.018

-.109

.123

-.794**

.941**

Sig.

.053

.723

.414

.005

.658

.658

.036

.034

.801

.816

.290

.004

.964

.940

.648

.607

.000

.000

*

**

Cor.

-.368

-.192

-.299

.577

-.042

-.042

-.491

-.497

-.038

-.029

-.306

.499

-.041

.028

-.077

.138

-.719

.878**

Sig.

.110

.417

.200

.008

.859

.859

.028

.026

.872

.902

.189

.025

.863

.907

.749

.562

.000

.000

*

**

*

*

**

Cor.

-.424

-.086

-.204

.676

-.020

-.020

-.421

-.428

-.171

-.169

-.324

.483

.096

.141

-.187

.177

-.778

.929**

Sig.

.062

.720

.389

.001

.933

.933

.064

.060

.470

.475

.164

.031

.686

.554

.430

.455

.000

.000

[ A12 ]

**

Appendix A

Spectral Bands B+G+R

B + G + NIR

B + R + NIR

B+R+C

B + C + NIR

G+C+R

G + C + NIR

R + C + NIR

G + R + NIR

B + G + R + NIR

C+B+G+R

C + B + G + NIR

C + G + R + NIR

C + B + R + NIR

PH

EC

TDS

T

NTU

TSS

NO3

NO3-N

PO4

TP

DO

BOD

OWQI

CWQI

AWWQI

NFSWQI

SD

TSI

Cor.

-.443

-.064

-.181

.669**

-.050

-.050

-.429

-.435

-.154

-.153

-.298

.532*

.082

.115

-.179

.164

-.800**

.948**

Sig.

.050

.789

.445

.001

.833

.833

.059

.055

.516

.521

.201

.016

.732

.630

.451

.490

.000

.000

-.304

*

.522

.055

.099

Cor.

-.431

-.096

-.213

**

.685

-.041

-.041

-.443

*

-.130

-.449

-.129

-.156

.154

**

-.796

.938**

Sig.

.058

.688

.367

.001

.865

.865

.051

.047

.584

.589

.193

.018

.818

.677

.513

.516

.000

.000

Cor.

-.438

-.093

-.206

.627**

-.065

-.065

-.456*

-.461*

-.105

-.101

-.287

.555*

.032

.070

-.139

.152

-.786**

.940**

Sig.

.053

.696

.383

.003

.786

.786

.044

.041

.661

.673

.221

.011

.892

.769

.560

.523

.000

.000

*

Cor.

-.431

-.083

-.198

.631

-.039

-.039

-.432

-.439

-.153

-.149

-.312

.509

.080

.119

-.175

.177

-.772

.933**

Sig.

.058

.727

.402

.003

.870

.870

.057

.053

.520

.531

.181

.022

.736

.616

.460

.456

.000

.000

-.328

*

.472

.038

.098

Cor.

-.400

-.147

-.260

**

**

.621

-.015

-.015

*

*

-.460

-.114

-.454

-.108

-.136

.171

**

**

-.739

.899**

Sig.

.080

.536

.268

.003

.949

.949

.044

.041

.633

.650

.158

.036

.875

.681

.569

.471

.000

.000

Cor.

-.436

-.066

-.182

.661**

-.055

-.055

-.431

-.437

-.152

-.149

-.298

.534*

.078

.111

-.178

.162

-.796**

.947**

Sig.

.054

.784

.442

.002

.819

.819

.058

.054

.524

.530

.202

.015

.744

.641

.454

.496

.000

.000

*

Cor.

-.436

-.066

-.182

.661

-.055

-.055

-.431

-.437

-.152

-.149

-.298

.534

.078

.111

-.178

.162

-.796

.947**

Sig.

.054

.784

.442

.002

.819

.819

.058

.054

.524

.530

.202

.015

.744

.641

.454

.496

.000

.000

-.283

*

.559

.021

.059

Cor.

-.426

-.099

-.211

**

**

.608

-.073

-.073

*

-.462

*

-.095

-.467

-.089

-.132

.147

**

**

-.778

.935**

Sig.

.061

.677

.373

.004

.759

.759

.040

.038

.692

.708

.226

.010

.930

.804

.579

.537

.000

.000

Cor.

-.442

-.070

-.185

.662**

-.077

-.077

-.448*

-.454*

-.113

-.111

-.275

.573**

.040

.071

-.149

.140

-.810**

.953**

Sig.

.051

.768

.434

.001

.747

.747

.048

.044

.634

.641

.240

.008

.867

.765

.530

.557

.000

.000

*

Cor.

-.440

-.078

-.194

.663

-.058

-.058

-.443

-.449

-.129

-.126

-.292

.546

.055

.091

-.158

.153

-.800

.947**

Sig.

.052

.742

.411

.001

.807

.807

.051

.047

.589

.596

.212

.013

.817

.702

.506

.519

.000

.000

-.307

*

.517

.084

.121

Cor.

-.435

-.074

-.190

**

**

.661

-.042

-.042

-.429

*

-.435

-.157

-.155

-.180

.169

**

**

-.789

.941**

Sig.

.055

.758

.421

.002

.860

.860

.059

.055

.508

.514

.187

.020

.725

.612

.448

.476

.000

.000

Cor.

-.421

-.103

-.220

.670**

-.031

-.031

-.440

-.446*

-.139

-.136

-.315

.502*

.063

.111

-.161

.164

-.781**

.929**

Sig.

.064

.665

.352

.001

.896

.896

.052

.049

.558

.566

.176

.024

.791

.642

.496

.490

.000

.000

*

**

Cor.

-.433

-.081

-.196

.655

-.063

-.063

-.446

-.452

-.124

-.121

-.290

.548

.050

.086

-.156

.151

-.797

.946**

Sig.

.056

.734

.407

.002

.792

.792

.049

.045

.601

.610

.214

.012

.835

.718

.512

.526

.000

.000

*

**

*

*

*

*

**

Cor.

-.427

-.101

-.215

.624

-.050

-.050

-.451

-.456

-.120

-.115

-.302

.528

.046

.088

-.149

.162

-.773

.931**

Sig.

.060

.670

.363

.003

.835

.835

.046

.043

.616

.629

.195

.017

.847

.711

.532

.494

.000

.000

[ A13 ]

**

Appendix A

Spectral Bands C + B + G + R + NIR

C/B

C/G

C/R

C / NIR

B/C

B/G

B/R

B / NIR

G/C

G/B

G/R

G / NIR

R/C

PH

EC

TDS

T

NTU

TSS

NO3

NO3-N

PO4

TP

DO

BOD

OWQI

CWQI

AWWQI

NFSWQI

SD

TSI

Cor.

-.432

-.086

-.202

.656**

-.049

-.049

-.441

-.447*

-.135

-.132

-.301

.529*

.061

.100

-.162

.160

-.790**

.941**

Sig.

.057

.719

.393

.002

.837

.837

.051

.048

.570

.578

.197

.016

.798

.673

.495

.500

.000

.000

Cor.

*

.463

.091

.195

**

-.773

-.013

-.013

.402

.408

.155

.164

.349

-.326

-.110

-.121

.160

-.152

**

.775

-.829**

Sig.

.040

.704

.409

.000

.956

.956

.079

.074

.513

.489

.132

.161

.644

.612

.499

.523

.000

.000

Cor.

.432

-.049

.056

-.883**

.190

.190

.391

.396

.086

.100

.202

-.552*

-.028

.031

.148

.015

.927**

-.902**

Sig.

.057

.838

.815

.000

.422

.422

.088

.084

.717

.675

.393

.012

.907

.898

.533

.950

.000

.000

*

*

*

*

Cor.

.385

.017

.093

-.704

.466

.466

.485

.487

-.059

-.057

.112

-.569

.115

.279

.102

.104

.903

-.820**

Sig.

.093

.945

.695

.001

.038

.038

.030

.029

.803

.810

.638

.009

.628

.234

.667

.662

.000

.000

-.399

.344

.161

-.056

Cor.

-.051

.251

.225

**

.033

.333

.333

.265

*

.258

-.555

**

-.565

**

**

**

-.232

-.415

.568

.590

**

Sig.

.830

.286

.341

.891

.151

.151

.260

.272

.011

.009

.325

.069

.009

.006

.082

.137

.498

.815

Cor.

-.477*

-.074

-.186

.776**

.014

.014

-.405

-.411

-.175

-.183

-.344

.373

.123

.144

-.178

.175

-.792**

.868**

Sig.

.033

.757

.433

.000

.955

.955

.076

.071

.460

.439

.138

.106

.606

.545

.453

.460

.000

.000

Cor.

.274

-.174

-.099

-.718

.331

.331

.270

.271

.000

.014

-.006

-.622

.057

.156

.099

.161

.792

-.712**

Sig.

.242

.463

.678

.000

.154

.154

.250

.247

1.000

.954

.979

.003

.811

.510

.679

.497

.000

.000

*

*

*

*

Cor.

**

**

.338

-.017

.051

-.624

.521

.521

.463

.464

-.086

-.086

.055

**

**

-.590

.143

.319

.091

.134

**

**

.852

-.765**

Sig.

.145

.943

.832

.003

.018

.018

.040

.039

.717

.719

.819

.006

.547

.171

.703

.573

.000

.000

Cor.

-.221

.190

.125

.319

.301

.301

.087

.079

-.558*

-.569**

-.328

-.234

.550*

.575**

-.421

.374

-.149

.266

Sig.

.348

.423

.599

.171

.198

.198

.715

.740

.011

.009

.158

.320

.012

.008

.065

.104

.531

.257

*

**

**

Cor.

-.463

.042

-.071

.867

-.130

-.130

-.406

-.411

-.128

-.139

-.217

.573

Sig.

.040

.859

.765

.000

.586

.586

.076

.072

.591

.558

.359

.008

Cor.

-.320

.170

.087

**

.760

-.271

-.271

-.293

-.295

-.029

-.043

-.010

.647

**

.063

.044

-.170

.057

-.914

.939**

.793

.853

.475

.810

.000

.000

-.031

-.103

-.108

-.115

**

**

-.824

.764**

Sig.

.169

.473

.714

.000

.247

.247

.210

.207

.902

.858

.967

.002

.898

.667

.650

.629

.000

.000

Cor.

.384

.038

.108

-.596**

.503*

.503*

.527*

.529*

-.079

-.084

.108

-.565**

.135

.304

.097

.071

.861**

-.801**

Sig.

.095

.873

.651

.006

.024

.024

.017

.016

.740

.724

.651

.009

.570

.192

.684

.765

.000

.000

*

*

Cor.

-.331

.238

.143

.590

.152

.152

-.044

-.053

-.553

-.568

-.325

.013

.517

.505

-.462

.325

-.456

.549*

Sig.

.154

.313

.548

.006

.523

.523

.853

.825

.012

.009

.162

.956

.019

.023

.040

.163

.043

.012

*

**

**

*

*

*

**

**

Cor.

-.479

-.007

-.114

.693

-.180

-.180

-.492

-.497

-.087

-.086

-.237

.617

Sig.

.033

.978

.632

.001

.446

.446

.028

.026

.716

.718

.315

.004

[ A14 ]

*

*

**

.020

-.002

-.151

.109

-.882

.972**

.933

.993

.524

.648

.000

.000

Appendix A

Spectral Bands R/B

R/G

R / NIR

NIR / C

NIR / B

NIR / G

NIR / R

B / NIR + C

B / NIR + B

B / NIR + G

B / NIR + R

B / NIR + NIR

B/R+C

B/R+B

PH

EC

TDS

T

NTU

TSS

NO3

NO3-N

PO4

TP

DO

BOD

OWQI

CWQI

AWWQI

NFSWQI

SD

TSI

Cor.

-.462*

.007

-.092

.670**

-.256

-.256

-.511*

-.515*

-.043

-.040

-.194

.647**

-.024

-.074

-.132

.067

-.893**

.954**

Sig.

.040

.976

.700

.001

.276

.276

.021

.020

.857

.867

.412

.002

.922

.755

.578

.779

.000

.000

Cor.

*

-.457

-.036

-.130

**

.610

-.295

-.295

*

*

-.550

-.015

-.546

-.009

-.211

.613

**

-.049

-.116

-.122

.075

**

-.870

.934**

Sig.

.043

.879

.585

.004

.207

.207

.013

.012

.948

.971

.372

.004

.837

.627

.608

.754

.000

.000

Cor.

-.488*

.099

-.018

.723**

-.094

-.094

-.373

-.380

-.309

-.313

-.308

.455*

.246

.210

-.325

.217

-.833**

.943**

Sig.

.029

.678

.941

.000

.693

.693

.106

.099

.185

.178

.187

.044

.296

.373

.162

.357

.000

.000

*

**

**

**

**

Cor.

.071

-.303

-.268

-.134

-.211

-.211

-.302

-.294

.665

.676

.256

.399

-.675

-.616

.538

-.308

-.021

-.057

Sig.

.765

.194

.253

.573

.371

.371

.196

.208

.001

.001

.277

.082

.001

.004

.014

.186

.931

.812

*

-.318

.211

-.300

Cor.

.201

-.249

-.187

-.348

-.196

-.196

-.151

-.142

**

.643

**

.656

.337

.260

**

-.639

**

-.591

.529

Sig.

.396

.290

.430

.133

.407

.407

.526

.550

.002

.002

.146

.268

.002

.006

.016

.172

.371

.199

Cor.

.263

-.296

-.218

-.515*

-.145

-.145

-.105

-.096

.630**

.646**

.324

.104

-.617**

-.554*

.531*

-.264

.394

-.460*

Sig.

.262

.205

.356

.020

.543

.543

.659

.686

.003

.002

.163

.664

.004

.011

.016

.260

.086

.041

*

*

**

**

Cor.

.440

-.208

-.107

-.763

.222

.222

.271

.280

.440

.447

.335

-.229

-.394

-.225

.479

-.181

.831

Sig.

.052

.378

.654

.000

.348

.348

.247

.232

.052

.048

.149

.331

.085

.340

.033

.446

.000

Cor.

-.378

-.126

-.239

**

.602

.010

.010

-.418

-.425

-.178

-.172

-.355

.410

.101

.161

-.187

.201

**

-.705

-.836** .000 .878**

Sig.

.100

.596

.310

.005

.965

.965

.067

.062

.452

.469

.125

.073

.670

.498

.431

.394

.001

.000

Cor.

-.423

-.103

-.220

.653**

.012

.012

-.412

-.419

-.183

-.181

-.344

.437

.112

.164

-.189

.201

-.745**

.904**

Sig.

.063

.666

.352

.002

.960

.960

.071

.066

.439

.445

.137

.054

.638

.490

.424

.396

.000

.000

*

Cor.

-.439

-.054

-.174

.714

-.046

-.046

-.419

-.425

-.167

-.168

-.299

.526

.093

.126

-.189

.156

-.815

.951**

Sig.

.053

.821

.463

.000

.847

.847

.066

.062

.480

.479

.201

.017

.698

.596

.425

.511

.000

.000

*

-.045

Sig.

.046

Cor.

-.232

Sig.

.324

.143

Cor.

-.451

**

**

-.157

.616

.851

.509

.004

.714

.714

-.339

-.395

.331

-.184

-.184

.085

.154

.438

.438

.008

**

-.087

-.087

-.436

-.267

.580

**

-.442

-.127

-.124

.058

.055

.051

.593

.603

.255

.007

-.575**

-.573**

.407

.420

-.068

.621**

.008

.075

.065

.776

.003

.036

.151

**

**

-.797

.950**

.079

-.163

.808

.742

.492

.525

.000

.000

-.471*

-.389

.291

-.094

-.543*

.612**

.090

.213

.693

.013

.004

Cor.

-.376

-.130

-.243

.598

.011

.011

-.420

-.426

-.173

-.166

-.353

.412

.096

.157

-.182

.200

-.703

.877**

Sig.

.102

.585

.302

.005

.965

.965

.066

.061

.465

.483

.127

.071

.686

.508

.442

.399

.001

.000

**

**

Cor.

-.422

-.106

-.222

.651

.012

.012

-.413

-.420

-.180

-.177

-.343

.439

.108

.161

-.186

.199

-.744

.904**

Sig.

.064

.658

.346

.002

.960

.960

.070

.065

.448

.455

.138

.053

.649

.497

.432

.400

.000

.000

[ A15 ]

**

Appendix A

Spectral Bands B/R+G

B/R+R

B / R + NIR

NIR / B + C

NIR / B + B

NIR / B + G

NIR / B + R

NIR / B + NIR

PH

EC

TDS

T

NTU

TSS

NO3

NO3-N

PO4

TP

DO

BOD

OWQI

CWQI

AWWQI

NFSWQI

SD

TSI

Cor.

-.439

-.055

-.175

.713**

-.046

-.046

-.420

-.426

-.165

-.166

-.298

.528*

.090

.125

-.187

.155

-.815**

.951**

Sig.

.053

.817

.460

.000

.847

.847

.066

.061

.486

.485

.202

.017

.705

.601

.429

.514

.000

.000

*

-.046

Sig.

.046

Cor.

-.213

Sig.

.366

.133

Cor.

-.450

**

-.158

.614

.847

.506

.004

.713

.713

-.348

-.399

.300

-.183

-.183

.082

.199

.440

.440

.009

-.088

-.437

-.442

-.125

-.121

-.266

.581

.056

.054

.051

.601

.610

.257

.007

-.565**

-.564**

.429

.443

-.052

.615**

.010

.059

.051

.828

.004

.028

.150

**

-.796

.950**

.077

-.161

.816

.748

.498

.528

.000

.000

-.491*

-.405

.312

-.107

-.513*

.579**

.076

.181

.653

.021

.007

Cor.

-.378

-.130

-.242

.601

.006

.006

-.422

-.428

-.171

-.165

-.352

.416

.094

.154

-.182

.198

-.707

.879**

Sig.

.101

.586

.303

.005

.979

.979

.064

.060

.470

.488

.128

.068

.692

.517

.443

.404

.000

.000

Cor.

-.423

-.105

-.222

**

-.088

**

**

.652

.009

.009

-.415

-.422

-.178

-.176

-.343

.441

.107

.159

-.186

.198

**

**

-.747

.906**

Sig.

.063

.658

.347

.002

.970

.970

.069

.064

.452

.458

.139

.051

.654

.504

.433

.403

.000

.000

Cor.

-.439

-.055

-.175

.714**

-.048

-.048

-.420

-.426

-.165

-.165

-.297

.529*

.089

.123

-.187

.154

-.816**

.952**

Sig.

.053

.817

.460

.000

.842

.842

.065

.061

.488

.487

.203

.017

.708

.605

.430

.516

.000

.000

*

**

**

Cor.

-.450

-.046

-.158

.615

-.089

-.089

-.437

-.443

-.124

-.121

-.266

.582

Sig.

.046

.847

.506

.004

.708

.708

.054

.050

.603

.613

.258

.007

**

**

**

Cor.

-.217

-.341

-.393

.309

-.194

-.194

-.567

-.566

.425

.439

-.051

.620

Sig.

.357

.141

.086

.184

.413

.413

.009

.009

.062

.053

.830

.004

* Correlation is significant at 0.05 level (2-tailed) as highlighted with Yellow color. ** Correlation is significant at 0.01 level (2-tailed) as highlighted with Red color.

[ A16 ]

.055 .819 *

**

.075

-.161

.149

-.797

.753

.499

.531

.000

.950** .000

*

-.488

-.408

.305

-.110

-.525

.587**

.029

.074

.191

.646

.018

.007

Appendix A

Table (A-3): Selection of variables for assessment of water quality in relation to non-industrial water use (Chapman, 1996). Water Quality Parameters

Background monitoring

Aquatic life and fisheries

Drinking water sources

Recreation and health

Agriculture Livestock Irrigation watering

General variables Temperature Colour Odour Suspended solids Turbidity/transparency Conductivity Total dissolved solids pH Dissolved oxygen Hardness Chlorophyll a

Xxx Xx

xxx

Xxx X Xx

xxx xx x x xx xxx x xx

Xx Xx Xxx Xx X X X X Xx Xx

xxx x

X Xxx

Xxx Xxx X

x xx xx xxx xx

x

x xxx xx x

x

xx

Nutrients Ammonia Nitrate/nitrite Phosphorus or phosphate Organic matter

X Xx Xx

TOC COD BOD

Xx Xx Xxx

X xx xxx

xx

x

Xx

Major ions Sodium Potassium Calcium Magnesium Chloride Sulphate

X X X Xx Xx X

X

xxx x

X X X

x

xxx x

Other inorganic variables Fluoride Boron Cyanide

Xx x

X

xx xx

Xxx Xx

x x x xx x

Xx Xxx Xx Xx X

xx

Xxx Xxx Xxx

xxx xxx xxx

x xx

x x

x x

x x

x

x x x x x

Trace elements Heavy metals Arsenic & selenium Organic contaminants Oil and hydrocarbons Organic solvents Phenols Pesticides Surfactants

x

Microbiological indicators Faecal coliforms Total coliforms Pathogens

xxx x x

xx

x - xxx: (Low to high) likelihood that the concentration of the variable will be affected and the more important it is to include the variable in a monitoring programme. Variables stipulated in local guidelines or standards for a specific water use should be included when monitoring for that specific use.

[ A17 ]

‫الخالصة‬ ‫ليس من العملي رصد نوعية المياه بالطرق الميدانية التقليدية سوى ألجزاء صغيرة من‬ ‫البحيرات بسبب الكلفة الباهضة والوقت الالزم ألخذ النماذج وفحصها موقعيا ومختبريا وتحليل‬ ‫النتائج‪ .‬ان االستشعار عن بعد (التحسس النائي) بواسطة األقمار االصطناعية ذات الدقة العالية‬ ‫للصور هو اكثر مالءمة تطبيقها لجمع البيانات الالزمة لرصد وتقييم نوعية المياه فى‬ ‫البحيرات‪ .‬لذلك فان الهدف من هذه الدراسة هو تحديد بعض مؤشرات جودة المياه والخصائص‬ ‫الفزيائية والكيميائية للمياه (درجة الحرارة‪ ،‬االوكسجين الذائب‪ ،‬المتطلب البيوكيميائي‬ ‫لالوكسجين (‪ ،)BOD‬الحامضية‪ ،‬الكدرة او العكورة‪ ،‬المواد العالقة الكلية‪ ،‬المواد الذائبة الكلية‪،‬‬ ‫التوصيل الكهربائي‪ ،‬النترات‪ ،‬الفسفور وبكتريا ‪ )E. Coli‬باستخدام تكنولوجيا االستشعار عن‬ ‫بعد للقمر االصطناعي (‪ )Landsat 8 OLI‬ونظم المعلومات الجغرافية (‪ )GIS‬لعشرين نقطة‬ ‫فى بحيرة دوكان في اقليم كردستان‪ ،‬العراق لموسمين مختلفين من السنة‪.‬‬ ‫اربعة طرق حسابية قياسية مختلفة (‪)AWWQI, OWQI, CMEWQI, NSFWQI‬‬ ‫استخدمت اليجاد مؤشرات نوعية المياه فى النقاط العشرين في بحيرة دوكان‪ .‬ان نتائج الطريقة‬ ‫األولى (‪ )NSFWQI‬تشير الى التصنيف المتوسط لكل النقاط باستثناء النقاط ‪ 11‬و ‪ 12‬و ‪13‬‬

‫التي تشير الى التصنيف السيء لمياه البحيرة في موسم الربيع‪ .‬اما في الطريقة الثانية‬ ‫(‪ )CMEWQI‬فان جميع النقاط لها تصنيف جيد في موسم الخريف باستثناء النقطتين ‪ 1‬و ‪2‬‬

‫فلهما التصنيف الحد األدنى والنقطة ‪ 16‬ذات تصنيف مقبول‪ .‬فى الطريقة الثالثة (‪ )OWQI‬فان‬ ‫جميع النقاط لها تصنيف جيد في موسم الخريف باستثناء النقطتين ‪ 1‬و ‪ 8‬ذات تصنيف مقبول‬ ‫وان جميع النقاط لها تصنيف ضعيف جدا في موسم الربيع‪ .‬واخيرا‪ ,‬فان جميع النقاط في‬ ‫الطريقة الرابعة (‪ )AWWQI‬ذات تصنيف ضعيف باستثناء النقطتين ‪ 15‬و ‪ 19‬بانهما ضعيفتان‬ ‫جدا والنقطة ‪ 16‬غير صالحة للشرب‪ .‬اما بالنسبة لمؤشري الشفافية (‪ )SDT‬و (‪ )TSI‬فان النتائج‬ ‫تشير الى التباين الشديد فى جميع النقاط في البحيرة‪.‬‬ ‫في هذه الدراسة‪ ،‬تم استخدام الطيف الضوئي الجديد (الزرقاء الساحلي ‪)Coastal Blue‬‬

‫في القمر االصطناعي (‪ )Landsat 8 OLI‬فى وضع نماذج مؤشرات نوعية المياه لبحيرة‬ ‫الدوكان‪ .‬عالوة على ذلك‪ ،‬تم استخدام تقنية تحليل المكونات المستقلة الجديدة ( ‪Independent‬‬

‫‪ (ICA) )Component Analysis‬وتقنية الحد االدنى إلزالة الضوضاء ( ‪Minimum Noise‬‬

‫‪ )MNF) )Fraction‬وسبعة نسب مختلفة وستة عشر مزيجا مختلفا من االطياف الضوئية ايضا‬ ‫في وضع النماذج‪.‬‬ ‫تم اسخدام طريقة (‪ )Multiple Linear Regression‬لوضع النماذج الرياضية لمؤشرات‬ ‫نوعية المياه و الخصائص الفزيائية والكيميائية للمياه المذكورة أعاله وذلك بربطها مع االطياف‬ ‫الضوئية ذات الدقة العالية للقمر االصطناعي (‪ .)Landsat 8 OLI‬افضل نموذج تم الحصول‬ ‫عليه هو للطريقة (‪ )AWWQI‬والذي اعطى معامل ارتباط عالي ( ‪ )R2‬قدره ‪ 0.993‬للبيانات‬ ‫في موسم الخريف ومنخفض قليال قدره ‪ 0.612‬للبيانات في موسم الربيع‪ .‬اعلى معامل ارتباط‬ ‫بالنسية لمؤشري الشفافية (‪ )SDT‬و (‪ )TSI‬فهو ‪ 0.873 ،0.982‬في موسم الخريف و ‪،0.973‬‬ ‫‪0.951‬‬

‫في موسم الربيع على التوالي‪ .‬كذلك أظهرت النتائج قيم عالية لمعامل االرتباط قدره‬

‫‪ 0.982 ،0.832 ،0.982‬لكل من الكدرة (العكورة)‪ ،‬المواد العالقة الكلية واالوكسجين الذائب‬ ‫على التوالي‪ .‬عموما‪ ,‬للبيانات في موسم الربيع فان نتائج جميع النماذج تشير الى انخفاض في‬ ‫األداء بسبب التغييرات الموسمية‪ ،‬تباين المعايير والعوامل االخرى‪ .‬باإلضافة‪ ،‬فقد تم تسجيل‬ ‫معامل ارتباط قدره ‪ 0.862‬لنموذج درجة الحرارة‪.‬‬ ‫وبوضع النماذج بصورة خرائط ملونة يتيسر التكهن بتوزيع مؤشرات نوعية المياه في‬ ‫البحيرة وكل النتائج كانت معقولة‪ .‬االستنتاجات تشير الى ترابط معقول لجميع االطياف‬ ‫الضوئية للقمر االصطناعي (‪ )Landsat 8 OLI‬مع مؤشرات نوعية المياه‪ .‬ونقترح ان تكون‬ ‫هناك دراسات إضافية لالستشعار عن بعد لمؤشرات نوعية المياه العماق مختلفة فى بحيرة‬ ‫دوكان‪.‬‬

‫وزارة التعليم العالي والبحث العلمي‬ ‫جامعة السليمانية‬ ‫عمادة كلية الهندسة‬ ‫قسم هندسة الري‬

‫تقييم نوعية المياه في بحيرة الدوكان‬ ‫باستخدام صور القمر االصطناعي‬ ‫‪Landsat 8 OLI‬‬ ‫إطروحة‬ ‫مقدمة الى عمادة كلية الهندسة‪/‬جامعة السليمانية‬ ‫كجزء من متطلبات نيل درجة الماجستير‬ ‫في هندسة الموارد المائية‬ ‫من قبل‬ ‫هه ستى شوان عبدهللا‬ ‫بكالوريوس هندسة الري ‪2010 -‬‬

‫باشراف‬ ‫أ‪.‬م‪.‬د‪ .‬محمود صالح مهدي‬

‫تشرين االول ‪ 2015‬ميالدية‬

‫د‪ .‬حكمت مصطفى إبراهيم‬

‫گالريزان ‪ 2715‬كردي‬

‫محرم ‪ 1437‬هجرية‬

‫ثوخـتـــة‬ ‫چاودێریکردنی کوالیتی ئاو زياتر لە كەرتێكى بچووكى دەرياچەدا لە کارێکی ئاسايی‬ ‫مەیدانى دا کارێکی بێ كەڵكە بەهۆى خەرجى و پێويستبوونی كاتێکی زۆر‪ .‬بڕياردان بە‬ ‫هۆی ژیری هەستیاری دووردەستەوە کەلە ڕووی بڵندی مانگە دەسکرداکانەوە هەیە‪،‬‬ ‫گونجاوترە لە رووی بەکارهێنان و كۆ کردنەوەی زانيارى پێويست بۆ چاودێرى كردن و‬ ‫لەهەڵسنگاندنی کوالیتی ئاو لە دەرياچەيەكەدا‪ .‬بۆيە‪ ،‬ئامانجى ئەم توێژینەوەیە بریتیە‬ ‫لەدیاری کردنی پێرستی کوالیتی ئاو وچڕی هەندێک پارامیتەری (پلەى گەرمى‪،BOD ،‬‬ ‫‪ ،EC ،TDS ،TSS ،pH‬ناترات‪ ،‬فۆسفات‪ ،‬بەکرتریای ‪ )E.coli‬بە بەکار هێنانی ژیری‬ ‫هەستیاری دووردەستی مانگی دەستکردی ‪ Landsat 8 OLI‬و ‪ GIS‬دا بۆ ‪ 20‬خالی‬ ‫وێستگەیی سەر دەریاچەی دووکان لە کوردستانی عێراق لە دوو وەرزی جیاوەزدا‪.‬‬ ‫چواردەستەواژەی پێوانەیی بیرکاریانە بەکار هات لە دۆزینەوەی پێرستی کوالیتی ئاو‬ ‫لە بیست وێستگە لە دەریاچەی دوکان‪ .‬ئەنجامەکان بۆدەستەواژەی (‪ )NSFWQI‬دۆزریەوە‬ ‫کە ڕێژەیی ناوەند بۆ هەموو وێستگەکان بیجگە لە وێستگەی ‪ 13 ،12 ،11‬کە پۆلەکەیان‬ ‫خراپ بوو لە وەرزی پایزدا‪ .‬دەستەواژەی دووەم‪ ،(CCMEWQI) ،‬هەموو وێسگەکان پۆلی‬ ‫باش بوون تەنها وێسگەی ‪ 2 ،1‬نەبێت کە بە پۆلی پەراوێزخراو یاخود پۆلی کەم دادەنرێت‬ ‫وە پۆلی دادپەرورانە بۆ وێستگەی ‪ 16‬هەژمار دەکرێت‪ .‬لە دەستەواژەی سێیەم دا‬ ‫)‪ ،(AWWQI‬هموو ویستگەکان پۆلێن دەکرێن بە خراپ تەنها ‪ 15‬و ‪ 19‬نەبێت بە زۆر‬ ‫خراپ دادەنرێت‪ ،‬لە هەمان کاتدا ویستگەی ‪ 16‬بۆ خواردنەوە ناشێت‪ .‬دەرئەنجاماکان بۆ‬ ‫دیسکی ساکی رۆشنبین )‪ Secchi Disk Transparency (SDT‬و پێرستی دۆخی خولگەیی‬ ‫)‪ Trophic State Index (TSI‬دەرچوو کە گۆرانێكی زۆری بەرچاو هەبوو لە وێستگەکاندا‪.‬‬ ‫فرە ڕيگرێشنی هێڵى (‪ (Regression Multiple linear‬بەكار هێنرا تا مۆدێلێکى بيركارانە‬ ‫مەزەندە بكات بۆ نيشانەکردنی پێرستی چۆنيەتى ئاو و هەندێك پارامیتەر کە پشت‬ ‫دەبەستن بە شکانەوەو ڕەنگدانەوەی سپەكترالى الندسات ‪ 8‬ئۆلى‪.‬‬

‫لەم توێژینەوەیەدا‪ ،‬باندی نوێ )شینی کەناراوی ) ی ‪ Landsat 8 OLI‬ی وەرگیرا و‬ ‫بەکار هێنرا لە دروستکردنی مۆدێلکان ‪ .‬زیاد لەوەش تەکنیکی شیکردنەوەی سەربەخۆی‬ ‫باندەکان (‪ (ICA) )Independent Component Analysis‬و البردنی کەمترین کەرتی‬ ‫ژاوەژاوی باندەکان )‪ Minimum Noise Fraction (MNF‬و ‪ 7‬رێژەی باندی )شینی‬ ‫کەناراوی ) لەگەڵ باندەکان تێکەڵ کراو و بەکارهێنران‪.‬‬ ‫باشترين مۆدێل بۆ پیرستی دەستەواژەی پێوانەیی مۆدێلی (‪)AWWQI‬ە كە نيشانی‬ ‫دەدات لەگەڵ هاوكۆلكەى بەرزى سووربوون ‪ )R2( Coefficient of Determination‬بۆ‬ ‫وەرزى پێش باراناوى دەربارەى ‪ 0.993‬یە‪ ،‬وە بۆ وەرزى بەهار كەمێك نزمە کە ئەویش‬ ‫(‪ )0.612‬یە‪ .‬بەرزترين هاوكۆلكەى سووربوون بۆ (‪ )SDT‬و (‪ 0.982 )TSI‬و ‪ 0.873‬بۆ‬ ‫وەرزى پایز وە (‪ )0.973 ،0.951‬بۆ وەرزى بەهار بە ڕیز‪ .‬لەگەڵ ئەوەشدا‪ ،‬بەرزترین ‪R2‬‬ ‫ى ‪ 0.832 ،0.982 ،0.982‬بۆ ‪ ،TSS‬لێڵى و ‪ DO‬ئەنجامیان دەردەکەوێت بەڕيز‪ .‬بە‬ ‫شێوەيەكى گشتى‪ ،‬بۆ وەرزى بەهار هەموو مۆدێلەکان ئاستی وردی كاركردنیان دادەبەزێت‬ ‫بەهۆى گۆڕانى وەرزى و جياوازيى چری پارامیتەرەکان و هۆكارى ديكە‪ .‬لەگەڵ ئەوەشدا‪،‬‬ ‫ئەنجامەکان ‪ R2‬بەرز کە ‪ 0.862‬یە نيشان دەدەن بۆ پلەى گەرمى ئاو‪.‬‬ ‫لە کاتێک دا کە هەموو مۆدێلەکان بەکارهێنران بۆ ئەوەی نەخشەیەکیان هەبێ بە‬ ‫جیاوازی رەنگەکان تا بە شێوەیەک ئاسانکاری بکات لە پێشبینی کردنی ئەنجامەکان و‬ ‫هۆکارەکانی پەسەند بن بۆ باڵوبونەوەی چۆنێتی ئاوی دەریاچەکە‪ .‬دەرئەنجامەکە‬ ‫دەریداخات کە هاو پەیوەندی هەموو باندی مانگی دەستکردی ‪ Landsat 8 OLI‬گونجاوە‬ ‫بۆ خەمالندنی چۆنیتی ئاو‪ .‬پێشنیاردەکرێت کە توێژینەوەی زیاتر بکرێت کە چۆن ژیری‬ ‫هەستیاری دووردەست بەکاربهێنرێت بۆ پێرستی دەستەواژەی پێوانەیی بێت یاخود خەستی‬ ‫پارامیتەری چؤنێتی ئاو لە قواڵیی جیاواز دا لە دەریاچەی دوکان‪.‬‬

‫وةزارةتي خويَندني باالَ وتويَذينةوةي زانسيت‬ ‫زانكوَي سليَماني‬ ‫فاكةلَيت ئةندازياري‬ ‫بةشي ئةندازياري ئاوديَري‬

‫هەڵسەنگاندنی كواليتى ئاوی دەریاچەی‬ ‫دوکان بە بەکارهێنانی وێنەی مانگی‬ ‫دەستکردی ‪LANDSAT 8 OLI‬‬ ‫ماستةرنامةيةكة‬ ‫ثيشكةشكراوة بؤ فاكةلَيت ئةندازياري ‪ /‬زانكوَي سليَماني‬ ‫وةك بةشيَك لة ثيَداويسيت يةكاني بة دةستهيَناني ثلةي ماستةر‬ ‫لة ئةندازياري سةرضاوةكاني ئاو دا‬ ‫لةاليةن‬

‫هة سيت شوان عبداهلل‬

‫بةكالوريوس لة ئةندازياري ئاوديَري دا ‪2010 -‬‬ ‫بة سةرثةرشتيكردني‬

‫ث‪.‬ي‪.‬د‪ .‬حممود صاحل مهدي‬

‫تشريين يةكةم ‪ 2015‬زايين‬

‫د‪ .‬حكمت مصطفى ابراهيم‬

‫طةالريَزان ‪ 2715‬كوردي‬

‫موحةرم ‪ 1437‬كؤضي‬

Hasti Shwan Abdullah Fattah.pdf

A Thesis. Submitted to the Faculty of Engineering of the University of. Sulaimani in Partial Fulfilment of the Requirements. for the Degree of Master in Science of.

9MB Sizes 13 Downloads 575 Views

Recommend Documents

Abdullah-Smith_Hazim.pdf
of Israel in 2013. www.canarymission.org. 3/13. Page 3 of 13. Abdullah-Smith_Hazim.pdf. Abdullah-Smith_Hazim.pdf. Open. Extract. Open with. Sign In.

Abdullah-Smith_Hazim.pdf
There was a problem previewing this document. Retrying... Download. Connect more apps... Try one of the apps below to open or edit this item. Abdullah-Smith_Hazim.pdf. Abdullah-Smith_Hazim.pdf. Open. Extract. Open with. Sign In. Main menu.Missing:

Dlshad Abdullah Rashid.pdf
SULAIMANI IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE. OF MASTER OF SCIENCE IN. GENERAL GENETICS. By.

Dlshad Abdullah Rashid.pdf
Loading… Page 1. Whoops! There was a problem loading more pages. Retrying... Dlshad Abdullah Rashid.pdf. Dlshad Abdullah Rashid.pdf. Open. Extract. Open with. Sign In. Main menu. Displaying Dlshad Abdullah Rashid.pdf.

Mikrajuddin Abdullah - Fisika Dasar I 2016.pdf
Try one of the apps below to open or edit this item. Mikrajuddin Abdullah - Fisika Dasar I 2016.pdf. Mikrajuddin Abdullah - Fisika Dasar I 2016.pdf. Open. Extract.

knowledge Sheikh-abdul-aziz-ibn-abdullah-ibn-baz.pdf
Page 2 of 25. Publisher's Note. All praise is for Allaah; peace and prayers be upon Muhammad. his family, his Companions and all those who follow in their. footsteps until the Day of Judgement. In the following pages is a short essay by one of the gr

(The Basmalah) by Abdullah Arık.pdf
Beyond Probability, God's Message in Mathematics, Ser ... ment of the Quran (The Basmalah) by Abdullah Arık.pdf. Open. Extract. Open with. Sign In. Main menu. Displaying Beyond Probability, God's Message in Mathematics, Series 1 The Opening Statemen

Larki Shadi Kyun Karti Hai by Hafiz Muhammad Abdullah Roprri.pdf ...
Larki Shadi Kyun Karti Hai by Hafiz Muhammad Abdullah Roprri.pdf. Larki Shadi Kyun Karti Hai by Hafiz Muhammad Abdullah Roprri.pdf. Open. Extract.

jannati-romoni-by-mohd-abdullah-al-kafi.pdf
bvixi Dci cyi“‡li KZ©„Z¡ 12 المرأة على الرجل قوامة. mr igbx c„w_exi †kaô m¤ú` 14 الدنيا متاع خير الصالحة المرأة. mr ̄¿xi ̧Yvejx 15 الصالحة الزوجة صفات. ̄^vgxi ...

Adalat e Nabvi saw Ke Faislay by Abdullah Al-Qurtabi.pdf
Adalat e Nabvi s.a.w Ke Faislay by Abdullah Al-Qurtabi.pdf. Adalat e Nabvi s.a.w Ke Faislay by Abdullah Al-Qurtabi.pdf. Open. Extract. Open with. Sign In.

Jannat Ki Talash Main by Abdullah bin Ali Al Husaiman.pdf ...
Page 3 of 123. com.KitaboSunnat.www. ہبتکم نئال نآ تفم لمتشم رپ تاعوضوم درفنمو عونتم نیزم ےس لئالد ہمکحم. Page 3 of 123. Jannat Ki Talash Main by ...