Surface temperatures of sea ice and ice sheets from satellites
Jacob Høyer, Gorm Dybkjær, Rasmus Tonboe and Rune Larsen Center for Ocean and Ice, DMI
Presentation outline • Motivation for IST data production • IST satellite level 2 product – Sea ice – Ice sheets • Validation of satellite IST • Future plans – IST level 4 product – Reanalysis
• Conclusions • Recommendations
GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Motivation
GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
DMIs modeling activities Operational modeling • Ocean/Ice model • Hycom and CICE • Atmosphere: • HIRLAM NWP + Harmonie • Climate models • Ice sheet models
GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Potential impact of IST • Assimilation in NWP (HIRLAM) to improve • Polar Lows • Piteraq (Katabatic winds, 30-90 m/s within 30 min) • Arctic surface fluxes • Assimilation: Ocean and Sea Ice models (Hycom + Cice) • Ice managing • Sea ice break up forecasts • Servicing the Oil industry for oil exploration • Coupled climate models • Improve Greenland ice sheet modelling GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Requested by the sea ice user community • ESA CCI project on Sea Ice • User survey • 91 respondents • IST included in survey, not in the project • IST ranking 4 out of 22 parameters.
GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Existing operational temperature observations • Very few near real time observations • Arctic Sea Ice • Greenland Ice sheet
GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
O&SI-SAF Metop-A IST product • • • • •
3 minute granules of AVHRR data from the NWC SAF PPS software. Input data: AVHRR-TIR, cloud mask, geometry, ice concentration (osi-saf)+ ECMWF NWP. Output: NetCDF files for each 3 minute granules in ~1 km resolution in swath projection Version 1 : EU project MyOcean-1 Version 2 : Eumetsat Ocean & Sea Ice SAF project, CDOP II
Inglefield Bredning on 3 April 2011: left 13:45UTC, right 22:22UTC GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Algorithm description •
Integrated HL SST, IST and Marginal Ice Zone Temperature product, based on Metop AVHRR (Tonboe et al., 2011, Dybkjær et al., 2012)
•
SST for HL – Regional algorithm coefficients (separate day/night ) IST – V1 taken from Key et al., 1992 MIZ – Scaled linearly between IST and SST, using T4 – Vincent et al., 2008
• •
Tonboe, R.T., Dybkjær, G., and Høyer, J.L. Simulations of the snow covered sea ice surface and microwave effective temperature. TELLUS, 63A, 1028-1037, 2011. Dybkjær, G., R. Tonboe, and J. Høyer, 2012. Arctic surface temperatures from Metop AVHRR compared to in situ ocean and land data. Submitted to: Ocean Sci. Discuss., 9, 1–35, 2012, doi:10.5194/osd-9-1-2012 GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Comparison with in situ observations Buoys and ships
Field site, Qaanaaq
Summit 2mt, at ~3200m GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Qaanaaq campaigns • IST from self-calibrating infrared radiometer (ISAR) • Perfomance ~1 deg C • Large errors on NWP STD of Errors: 1.02K ISAR vs Metop-A IST 2011.
Ice mass balance buoys. IR120(Campbellscientific) vs Metop-A IST vs NWP, 2012. GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Validation - buoys and ships in Arctic - Complicated observations - Snow covered
- Cloud screening is probably the largest error source for satellite based IST estimation.
Metop-A IST-OBSarctic STDError (C) = 3.69; Bias (C) = -2.76, R = 0.89
GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Validation Greenland Ice Sheet • T_2m is not IST • IST compare better than ECMWF T 2m that assimilates obs.
Metop-A IST-T_2msummit STDError (C) = 3.14; Bias (C) = −3.35, R = 0.96
GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Future plans
GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Level 4 ST + Reanalysis • DMI will develop an Arctic level 4 product – SST, MIZ and IST – Spatial resolution ~5 kilometers. – Temporal resolution, probably 6 hourly
• DMI will develop an IST reanalysis – EU FP7 project NACLIM – AVHRR GAC data set from 1989 to 2009.
GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Collect more validation data An ISAR delivered to DMI within 1 month
ODEN cruise August-September 2012. A science of opportunity project - w LOMROGIII exp.
Arina Arctica in Disko bay, March 2012
• 7 weeks, Ice breaker cruise with ISAR and surface sampling (2012). • Golf of Bothnia – ice breaker opportunities, winter 2012/13. • 3rd Field campaign Qaanaaq, March 2013. • Semi-permanent reference station/ship track ?
GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Conclusions • • • •
1 km Operational Sea ice and Greenland IST from Metop-A Sparse observations -> large NWP errors Large user request and potential for IST Comparisons with in situ observations show differences 1-4 degrees • 20 years reanaysis available in NACLIM EU FP 7 project • Data collection ongoing, but more radiometer observations needed.
GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012
Recommendations for Globtemp • Include IST in Globtemperature project (Not SST !) – Operational – Reanalysis • IST relatively new type of observations for operational applications. To speed up usage: – Impact assessment studies using IST – NWP – Ocean/Sea ice models • Users want reliable uncertainties: – Stimulate gathering of accurate in situ observations from radiometers
GlobTemperature User Consultation Meeting, Edinburgh, June 27-28, 2012 2012