Geomorphology 135 (2011) 181–190

Contents lists available at SciVerse ScienceDirect

Geomorphology journal homepage: www.elsevier.com/locate/geomorph

Quantifying relationships of burning, roughness, and potential dust emission with laser altimetry of soil surfaces at submeter scales Joel B. Sankey a,⁎, 1, Jan U.H. Eitel b, 1, Nancy F. Glenn a, Matthew J. Germino c, 2, Lee A. Vierling b a b c

Department of Geosciences, Idaho State University- Boise, 322 E. Front St., Suite 240, Boise, ID 83702, USA Geospatial Laboratory for Environmental Dynamics, University of Idaho, Moscow, ID 83844-1135, USA Department of Biological Sciences, Idaho State University, 921 S 8th Ave, Stop 8007, Pocatello, ID 83209-8007, USA

a r t i c l e

i n f o

Article history: Received 26 January 2011 Received in revised form 5 August 2011 Accepted 12 August 2011 Available online 22 August 2011 Keywords: Aeolian transport LiDAR Potential dust emission Wildfire Shrub steppe Microsite

a b s t r a c t Decrease in surface roughness by the reduction of vegetation is one mechanism by which fire can promote aeolian transport in a variety of landscapes. The extent to which fire might alter the roughness of the soil surface at fine spatial scales and the effect of this alteration on post-fire aeolian response is not well known. We examined relationships in the field between dust emissions and subcentimeter-level soil surface roughness at submeter spatial scales with a terrestrial laser scanner and portable wind tunnel analog. Based on aeolian theory, we hypothesized that observed relationships would differ from those determined in previous studies with laser altimetry at landscape scales (meter–kilometer length scales). We examined four semiarid shrublands in southern Idaho/USA containing a distinct pattern of undershrub and interspace microsites, including recently burned and unburned conditions. Mixed models were used to determine effects of burning on surface roughness and the response of dust emissions to changes in surface roughness. Results indicated that burned soil surfaces were rougher than unburned soil surfaces. Dust emissions were enhanced by increases in roughness on burned soil surfaces and in the absence of nonerodible roughness elements (i.e., plants). This finding is expectedly in contrast to previous work that demonstrated an inverse relationship between soil erosion and surface roughness determined with LiDAR (light detection and ranging) at landscape scales (vegetated and unvegetated surfaces). Relationships of LiDAR-derived roughness and aeolian emissions are scale dependent and vary with environmental factors of fire and vegetation. These findings are integral for future research that attempts to parameterize model-based predictions of aeolian emissions with LiDARderived roughness. Published by Elsevier B.V.

1. Introduction Measurements of surface roughness are necessary for model- and experiment-based predictions of aeolian emissions on earth and other planetary surfaces (King et al., 2008; Okin, 2008; Sutton and McKenna-Neuman, 2008; Maurer et al., 2010; Neakrase and Greeley, 2010; Turpin et al., 2010; Munson et al., 2011). Approaches to quantifying roughness are limited at a large geographic extent and fine spatial resolution on heterogeneous surfaces, however. Laser altimetry (i.e., light detection and ranging, LiDAR, or ground-based terrestrial laser scanning, TLS) has undeniable utility for making measurements of ⁎ Corresponding author at: USGS Western Geographic Science Center, 520 North Park Avenue, Room 111, Tucson, AZ 85719, USA. Tel.: +1 520 670 6671; fax: +1 520 670 5113. E-mail addresses: [email protected] (J.B. Sankey), [email protected] (J.U.H. Eitel), [email protected] (N.F. Glenn), [email protected] (M.J. Germino), [email protected] (L.A. Vierling). 1 Authors Sankey and Eitel contributed equally to this manuscript. 2 Present address: USGS Forest and Rangeland Ecosystem Science Center, Boise, ID 83706, USA. 0169-555X/$ – see front matter. Published by Elsevier B.V. doi:10.1016/j.geomorph.2011.08.016

roughness (variability in surface elevation at millimeter–kilometer length scales), yet it is a relatively unexplored predictive tool for aeolian processes (Bullard, 2006; Baas, 2008; Hodge et al., 2009a,b; Pelletier et al., 2009; Sankey et al., 2010; Nield et al., 2011; Nield and Wiggs, 2011). LiDAR measurements of surface variability have potential utility for estimating important parameters for aeolian transport models such as aerodynamic roughness and zero plane displacement (Menenti and Ritchie, 1994; De Vries et al., 1997); yet the relationship between LiDAR-derived roughness and many fundamental aeolian processes has not yet been described (Nield et al., 2011). Soil erosion by wind has been shown to be inversely related to roughness derived from LiDAR acquired at relatively coarse (10 1– 10 3 m) geographic scales (Pelletier et al., 2009; Sankey et al., 2010). At such coarse spatial scales, aeolian erosion is expected to predominate on smooth surfaces with limited vegetation and microtopography, whereas deposition predominates on rough surfaces that are vegetated and/or have greater microtopographic relief (i.e., more nonerodible roughness elements). At finer spatial scales (e.g., 10 -3– 10 -1 m), the positive relationship between maximum heights achieved by saltating particles and surface roughness (Fryrear and

182

J.B. Sankey et al. / Geomorphology 135 (2011) 181–190

Saleh, 1993) was recently characterized by LiDAR on a sand beach and additionally attributed to the hardness of the moist surface (Nield and Wiggs, 2011). Increases in surface roughness probably can enhance the aeolian emission of fine-grained dust at fine (b1 m) spatial scales and in the absence of nonerodible roughness features such as plants. 1.1. Roughness and dust relationships at submeter scales Nonerodible roughness elements such as vegetation or rocks extract momentum from the wind and therefore reduce its erosive force. However, surface roughness can enhance dust emissions when nonerodible elements are present but at very sparse density because of increased turbulence and erosive wind flow adjacent to the roughness elements (Raupach et al., 1993; Neakrase and Greeley, 2010). Mineral dust (e.g., PM10, particles b 10 μ diameter, and/or some, though not all, larger suspension-sized grains) is often too fine to be entrained directly by wind (Bagnold, 1941; Pye, 1987). Abrasion of the surface by coarser (e.g., saltation-sized) grains and aggregates can be required to initiate emission of dust (Bagnold, 1941; Pye, 1987; Nickling, 1988; Gillete and Chen, 2001). Coarser aggregates and grains create raised roughness features on otherwise smooth surfaces that are normal, and therefore more susceptible, to the wind's erosive force (Bagnold, 1941). Coarser aggregates and grains therefore can create rougher surfaces and be a necessity for the emission of dust-sized particles. 1.2. Roughness, microtopography, fire, and dust The relationship of dust emission and surface roughness at submeter spatial scales is especially relevant for soil surfaces with vegetation-induced microtopography, such as is characteristic of desert shrublands (Charley and West, 1975; Knight, 1994; Schlesinger et al., 1996). Microtopographic positions beneath shrub vegetation (undershrub microsites) are often raised features with lower soil bulk density, greater abundance of organic materials (ranging from undecomposed-decomposed), and more stable soil aggregates (Blackburn, 1975; Roundy et al., 1978; Wood et al., 1978, 1982; Doescher et al., 1984; Johnson and Gordon, 1988). These characteristics of undershrub microsites create a ground surface that appears rougher, with a wide range of erodible particle and aggregate sizes and densities. Shrub-interspaces, conversely, are generally slightly lower in relative elevation, less vegetated, with substantially fewer organic inputs, higher bulk density, lower aggregate stability, and greater prevalence of surfaces that appear smooth yet crusted and therefore less susceptible to erosion than undershrubs (Blackburn, 1975; Roundy et al., 1978; Wood et al., 1978, 1982; Doescher et al., 1984; Johnson and Gordon, 1988). The roughness and morphology of desert shrublands have been previously characterized with airborne profiling (e.g., Ritchie et al., 1992) and scanning (e.g., Rango et al., 2000) LiDAR. However, to our knowledge the relationship between subcentimeter level surface roughness and dust emissions on shrub microsites has not been explicitly examined in the field. Furthermore, many desert shrublands are not susceptible to wind erosion unless the protective cover of vegetation has been removed by recent disturbance, such as fire (Sankey et al., 2009a). We therefore anticipated that burning would affect the surface roughness of microsites as well as the relationship of microsite roughness and dust emissions. For instance, the combustion of organic materials during burning might produce a more homogenous and therefore smoother surface consisting of largely mineral materials and ash. Burned surfaces are generally more erodible than unburned surfaces, however, and if erosion and roughness are positively correlated at the scale of individual microsites, it could be argued recursively that burning might increase the roughness of microsite surfaces. A possible mechanism for an increase in microsite surface roughness following burning might be that soil heating and combustion of

organic materials can be spatially heterogeneous, even at fine spatial scales (i.e., beneath individual shrubs) (Seefeldt et al., 2007). 1.3. Study objectives We studied surface roughness and dust emissions at four sites representing a range of wildland fire histories in the sagebrush steppe, a type of cold desert shrubland that is not susceptible to appreciable wind erosion unless recently burned (Sankey et al., 2009a, 2010). We examined the roughness and dust emission relationship at very fine spatial scales of shrub (sagebrush) and interspace microsites (b0.3 m diameter), in the absence of vegetation, using a TLS laser altimetry system (Clawges et al., 2007; Haubrock et al., 2009; Eitel et al., 2011) and the Portable In-Situ Wind Erosion Laboratory — a wind tunnel analog (PI-SWERL; Etyemezian et al., 2007; Sweeney et al., 2008). These tools provided a controlled and experimentally robust opportunity to examine the roughness and erosion relationship in situ, at submeter spatial scales. Four specific research questions with associated hypotheses were addressed. The first two questions pertained to microsite roughness and dust emissions, and the third and fourth questions examined elevation differences between microsites and dust emissions. Research questions and associated hypotheses were: Question 1: does surface roughness differ for undersagebrush and intersagebrush microsites, and what is the effect of burning? Hypotheses: (i) undersagebrush soil surfaces are rougher than intersagebrush soil surfaces, and (ii) the difference in roughness between undersagebrush and intersagebrush microsites is greater for unburned vs. burned surfaces. Question 2: does dust emission potential vary with surface roughness at the microsite scale, and what is the effect of burning? Hypotheses: (i) roughness is a significant predictor of dust emissions, and (ii) dust emissions are greater on rough microsites and less on smooth microsites for burned and unburned surfaces. Question 3: what is the effect of burning on elevation differences between undersagebrush and intersagebrush microsites? Hypothesis: the difference in elevation between microsites, in which undershrubs are higher relative to adjacent interspaces, is greater for unburned vs. burned surfaces. Question 4: does dust emission potential vary with elevation differences between paired undersagebrush and intersagebrush microsites? Hypothesis: elevation differences between microsite pairs are a significant predictor of dust emission potential. 2. Study area This study was conducted in sagebrush steppe (desert shrublands) of southern Idaho (Anderson and Inouye, 2001) and the Snake River plain (SRP) at four sites during 21–24 June 2010 (Fig. 1). Each site included a wildland fire and nearby unburned location (i.e., burned and unburned treatment). Fire history at each site represented a temporal gradient; burning occurred 5 years to 2 months prior to the study (Clover fire — burned July 2005; Moonshiner fire — August 2007; Noman fire — July 2009; Samaria fire — April 2010). Unburned plant communities were primarily big sagebrush (Artemesia tridentata) and bunchgrasses. Vegetation on burned surfaces was predominantly crested wheatgrass (Agropyron cristatum) at Clover and Noman, a mix of native herbs at Moonshiner (including Poa secunda, Elymus elymoides, Pseudoregneria spicata, Lappula occidentalis, Collinsia parviflora), and surfaces were mostly bare at the recently burned Samaria site.

J.B. Sankey et al. / Geomorphology 135 (2011) 181–190

183

Fig. 1. A map of the location of study sites in Idaho and relative to the western United States.

Soil surfaces were predominantly developed from loess parent materials (Busacca et al., 2004). In comparison to the low topographic relief of the Clover, Noman, and Moonshiner sites, the Samaria site was situated in hilly terrain on the margin of the SRP (Fig. 1) and had a greater presence of nonaeolian (colluvial) parent materials at the soil surface. The study sites had relatively well-developed soils characteristic of the sagebrush steppe; often with argillic and calcic subsurface horizons (enriched in illuvial clay and calcium carbonate, respectively) and less frequently with mollic epipedons (organic rich surface horizons) (NRCS Web Soil Survey, 2010). Soils are predominantly Xerollic and Haploxerollic Durargids and Camborthids at the Clover site, Sodic Xeric Haplocalcids at the Moonshiner site, Lithic Xerollic Camborthids and Xerollic Haploargids at the Noman site, and Lithic and Typic Calcixerolls at the Samaria site (NRCS Web Soil Survey, 2010). Texture, organic matter, and gravimetric water content of microsite soil surfaces were characterized during the study as part of an associated project. Texture for microsite soil surfaces ranged from clay loam to loam amongst study sites. Particle size distributions were very similar amongst the Clover, Noman, and Samaria sites with mean PM10 in the surface soil ranging from 43 to 46% and clay loam soil texture. Soils at the Moonshiner site were derived from somewhat coarser parent materials and had loam textures and mean PM10 of 28%. Soil texture generally does not vary significantly by microsite in this environment (Hoover, 2010). Organic matter (OM) generally does vary by microsite, however, and microsite soil surfaces ranged from ~ 4 to 20% amongst study sites with up to two times greater OM on undershrub microsites than interspaces. Microsite soil surfaces were dry and ranged from 0 to 7% gravimetric water content at all sites, with site averages b 3%. 3. Methods We made nearly simultaneous measurements of soil surface roughness and dust emissions at four paired undershrub and interspace microsites within each burned and unburned treatment at the four sites (4 microsite pairs × 2 burn treatments × 4 sites = 64 total measurements). Surface roughness was determined from light detection and ranging (LiDAR) data collected with a ground-based Leica ScanStation 2 terrestrial laser scanner (TLS; Leica Geosystems, Inc.)

(Eitel et al., 2010). Dust emissions were measured immediately following each TLS scan with the PI-SWERL, which is analogous to a portable wind tunnel (Etyemezian et al., 2007; DUST-QUANT LLC, 2008). The weather was dry and calm, with very little wind between measurements and throughout the study, in general. Prior to TLS and PI-SWERL measurements, any potential interaction between the shrubs and the surface underneath the shrubs was eliminated from the emission experiments. Existing sagebrush shrubs were removed with a handsaw to ground level and herbaceous vegetation was carefully clipped (Fig. 2). Loose sticks, rocks, and litter that might damage the PI-SWERL (loose materials with diameter N ~1 cm) were also removed. We avoided disturbing the experimental soil surfaces by not stepping, kneeling, or placing our hands on the ground within the microsites. The PI-SWERL has a circular, 32-cm-diameter footprint that easily fit within the spatial dimensions of each of the individual microsites we examined. For the purpose of efficiency, we used the entire footprint of the PI-SWERL as the spatial unit of comparison of measurements of dust emissions with TLS measurements. Nonetheless, we acknowledge that the effective area of the PISWERL is considered to be smaller than the footprint because the shear imparted by the PI-SWERL on the ground surface varies within the footprint (Etyemezian et al., 2007; DUST-QUANT LLC, 2008). 3.1. Surface roughness The time-of-flight, Leica ScanStation 2 TLS employs a pulsed green (532 nm) laser with a beam diameter of 4 mm at a range of 0–50 m, a scan rate of 50,000 points s −1, a maximum sample density of b1 mm, and a maximum range of 134 m at 18% albedo (http://hds.leicageosystems.com). Distance accuracy is quoted as 4 mm and position accuracy as 6 mm, and the instrument returns one data point per laser pulse. Each microsite was scanned with the TLS from a single scan position (Fig. 2). The center of each microsite was marked by a 50-cm-tall wire (1 mm diameter) survey flag to ensure that the microsite center could be later identified in the scan image. The TLS can produce shadowing behind the first object struck by each laser beam if the object (e.g., vegetation) is situated above the ground surface (Eitel et al., 2010). We thus carefully removed vegetation within and surrounding each

184

J.B. Sankey et al. / Geomorphology 135 (2011) 181–190

Fig. 2. (A) Photo of interspace and undershrub pair, (B) digital elevation model (DEM) for an interspace and undershrub pair, (C) DEM for an interspace, and (D) DEM for an undershrub. DEMs were derived from the terrestrial laser scanner data. Note that relative elevation (vertical axis) for panels (B) and (C) is based on the interspace base elevation, whereas in panel (D) it is based on the undershrub base elevation.

microsite to reduce shadowing of the soil surface without disturbing the latter (Eitel et al., 2011). The laser point spacing was 1 mm at 5 m (100 points cm−2), and the scan duration was about 30 s/plot. For each microsite, the TLS point cloud was subset into a N50 × 50 cm rectangle containing the circular area of interest (AOI) of the PI-SWERL footprint (diameter = 32 cm) centered on the visually identifiable survey flag. The subset point cloud was gridded onto an arbitrarily subdivided 10 × 10 mm regular grid digital elevation model (DEM) using a program written in the Interactive Data Language (IDL) software package (version 4.5, ITT Corp., New York, NY). The z value assigned to each grid cell of the DEM was the minimum z value contained within the search radius (search radius = square root (2)/2 x grid resolution) from the center of the grid cell or, if there was no z value within the search radius, linearly interpolated based on surrounding z values. The rationale for assigning the minimum z-value within the search radius to each grid cell was to decrease the likelihood that non-surface returns were included in the analysis (Guarneri et al, 2009). Non-surface returns in this study were mainly caused by remains of clipped vegetation or erroneous high points around grain edges that can occur when the laser beam is split on the edge of an object so that the laser return is a mix of signals from two or more objects.

Based on the 10 × 10 mm DEM, the local root-mean squared height (subsequently termed surface roughness or locRMSH) was calculated in millimeters for each cell location (i) in six separate calculation trials that each used a different moving window size (30 × 30, 50 × 50, 70× 70, 90 × 90, 110 × 110, and 130 × 130 mm), using the following equation (Haubrock et al., 2009): vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u Xl Yl u u ∑ ∑ ½zðxc ; yr Þ−μ 2 uc¼X r¼Y u f f   locRMSHi −ws ¼ t  Xl −Xf Yl −Yf

ð1Þ

where ws denotes the one-dimensional window size (e.g., 30, 50, 70, 90, 110, 130 mm), Xf is the index of the first column, Xl is the index of the last column, Yf is the index of the first row, Yl is the index of the last row, c is the column index, r is the row index, z(xc,yr) is the height value in millimeters at position xc,yr, and μ is the average height value (mm) within the moving window. Surfaces were not detrended prior to calculating locRMSH, because recent research suggests that the relationship between soil erosion and locRMSH is weakened if locRMSH are derived from detrended, high spatial resolution surfaces (b1 cm) (Eitel et al., 2011).

J.B. Sankey et al. / Geomorphology 135 (2011) 181–190

After calculating the surface roughness, a circular AOI with the dimension of the PI-SWERL footprint was created around the center of the DEM (i.e., survey flag) and used to extract the mean locRMSH within the AOI in vertical units of millimeters. The mean elevation (z) of the DEM within the AOI was extracted in units of centimeters. Elevation differences were calculated by subtracting the mean elevation of the interspace DEM and AOI from the mean elevation of the undershrub DEM and AOI for each microsite pair. 3.2. Dust emissions The PI-SWERL imparts an erosive force on the soil surface with a blade that rotates above the ground within the instrument. At each PI-SWERL measurement, we used the same modified ramp test which lasted a total of 500 s, during which rotation of the blade was increased incrementally from 0 to 4000 revolutions per minute (RPM) with three sustained periods of time when the blade rotated at 2000, 3000, and 4000 RPM (DUST-QUANT LLC, 2008). Equivalent friction velocities ranged from 0 to 0.691 m s −1 based on the RPM levels used in our experiment relative to empirical measurements and calibration performed by Etyemezian et al. (2007). Etyemezian et al. (2007) used a flat plate for calibration, and we acknowledge that the RPM and equivalent friction velocity relationship might vary on rough soil surfaces. The PI-SWERL uses a DustTrak® (TSI, Inc.) optical sensor (nephelometer) to measure concentrations of PM10 entrained from the soil surface beneath the rotating blade. The data variable recorded by the PI-SWERL that we analyzed was the total mass of PM10 entrained from the surface during the 500-s test (heretofore referred to as PM10 or PM10 emissions or dust emissions) (DUST-QUANT LLC, 2008). For three measurements (two burned undershrubs and one burned interspace at the Samaria site) PM10 concentrations approached the recommended upper limit for the DustTrak sensor and we turned off the sensor for the final portion of the measurement. In these instances we used a linear regression (all r 2 N 0.93) of PM10 concentrations and optical gate sensor (OGS) counts to estimate the PM10 concentrations for the portion of the measurement that the DustTrak was turned off (Dust-Quant LLC, 2008). The OGS are additional sensors within the PI-SWERL that provide an estimate of saltation at the same temporal resolution and are strongly correlated with the PM10 concentrations estimated by the DustTrak sensor (DustQuant LLC, 2008). Dust emission is a portion of the total wind erosion process, however, because of the strong correlation between PM10 concentrations measured with the DustTrak® and OGS estimates of saltation, the PM10 variable was not only a measure of dust emission potential but also a proxy of the relative erosion potential (e.g., saltation and suspension processes) of microsite surfaces. 3.3. Statistical analysis Some variables (e.g. locRMSH, PM10 concentration) collected at a given site (Clover, Noman, Moonshiner, Samaria) were not independent of each other which would have resulted in an incorrect variance–covariance structure if ordinary regression would have been used. In this study we thus chose to use mixed effects modeling since this enables the correct modeling of the variance–covariance structure if the assumption of independence is violated. The following mixed effects models were used to test hypotheses associated with research questions 1 through 4 (see 1.3 Study objectives), respectively: Question 1 : locRMSHi−ws ¼ β0 þ β1 MS þ β2 DS þ γ

ð2Þ

where β0 is the intercept, β1 and β2 are fixed slopes, MS is the microsite (interspace or undershrub), DS is disturbance status (burned or unburned), and γ is a random effect entering the model as an

185

intercept term. The random effect (γ) adjusts the intercept for each site (Clover, Noman, Moonshiner, Samaria) by a random value. Question 2 : PM10 ¼ β0 þ β1 MS2 þ β2 DS þ β3 locRMSHi−ws þ γ ð3Þ where β0 is the intercept, β1, β2, and β3 are fixed slopes, MS is the microsite (interspace or undershrub); DS is disturbance status (burned or unburned), locRMSHi-ws is the surface roughness, and γ is a random effect entering the model as an intercept term. The random effect (γ) adjusts the intercept for each site (Clover, Noman, Moonshiner, Samaria) by a random value. Question 3 : ED ¼ β0 þ β1 DS þ γ

ð4Þ

where ED is the elevation difference between each microsite pair, β0 is the intercept and β1 is a fixed slope, DS is disturbance status (burned or unburned), and γ is a random effect entering the model as an intercept term. The random effect (γ) adjusts the intercept for each site (Clover, Noman, Moonshiner, Samaria) by a random value. Question 4 : PM10 ¼ β0 þ β1 ED þ γ

ð5Þ

where β0 is the intercept, and β1 is a fixed slope, ED is the elevation difference between each microsite pair, and γ is a random effect entering the model as an intercept term. The random effect (γ) adjusts the intercept for each site (Clover, Noman, Moonshiner, Samaria) by a random value. All models were characterized and model assumptions were checked using the open-source software package R 2.8.1 (R Development Core Team, 2008) with the nlme library (Pinheiro and Bates, 2000). Effects were considered to be significant at a p-value of b0.05. 4. Results We addressed questions 1 and 2 separately using locRMSH-30, -50, -70, -90, -110, -130, and found that the differences among microsites and burned/unburned sites were most significant for the locRMSH-90 determination (i.e., surface roughness calculated with the 90 × 90 mm moving window, as determined by smallest p-values; results not shown). Thus, we present and discuss findings only for locRMSH-90. 4.1. Microsite differences in surface roughness Undershrubs had 52% greater surface roughness than interspaces on average amongst sites (Table 1, microsite coefficient; Fig. 3). Burned surfaces had 24% greater surface roughness than unburned surfaces on average as well (Table 1, burn treatment coefficient; Fig. 3). However, microsite differences (i.e., undershrub vs. interspace) in roughness were similar in the burned and unburned surfaces, which was contrary to our expectations (Table 1). 4.2. PM10 emissions and surfaces roughness PM10 emissions varied significantly with surface roughness (Table 2). PM10 emissions were greater on rough relative to smooth surfaces (Table 2). Relationships of PM10 emissions and surface roughness differed significantly (or nearly so) by microsite and burned/unburned surfaces (Table 2). Greater PM10 emissions were associated with increased surface roughness consistently amongst the four burned sites (Fig. 4, upper right panel), but less consistently amongst unburned sites or microsites (Fig. 4, upper left and lower panels).

186

J.B. Sankey et al. / Geomorphology 135 (2011) 181–190

4.3. Elevation differences between microsites The relative elevation of soil surfaces tended, with exceptions, to be greater on undershrub compared to interspace microsites, similarly for burned and unburned sites (Fig. 5). Elevation differences between paired undershrub and interspace microsites were pronounced for the Clover site but not for the Samaria site, for example, for both burned and unburned surfaces (Fig. 5). The mixed model analysis furthermore indicated that burning was not a significant predictor of relative elevation for microsite pairs (intercept value = 5.04, burn treatment value= −0.28, p = 0.89, degrees of freedom= 27). 4.4. PM10 emissions and elevation differences Elevation differences between microsite pairs were not a significant predictor of PM10 emissions in the mixed model analysis (intercept value = 0.37, microsite pair value = 0.00, p = 0.11, degrees of freedom = 27). 5. Discussion 5.1. Microsite roughness, morphology, and burning The expectation that undershrub microsites generally occupy raised microtopographic positions relative to adjacent interspaces in desert shrublands was confirmed with laser altimetry. Findings supported the hypothesis that undershrub microsites have greater soil surface roughness than interspaces (Table 1; Fig. 3). We expected that undershrub soil surfaces would be rougher than interspaces based on abundant literature that describes a wider range of particle and aggregate sizes and densities for undershrubs relative to interspaces (Blackburn, 1975; Roundy et al., 1978; Wood et al., 1978, 1982; Doescher et al., 1984; Johnson and Gordon, 1988; Blackburn et al., 1990, 1992; Bolton et al., 1990; Goff et al., 1993; Pierson et al., 1994, 2001, 2002; Davies et al., 2007, 2009; Aanderud et al., 2008; Hooker et al., 2008) as well as the greater proportion of organic matter on the undershrub soil surfaces that we characterized. Furthermore, interspaces often have a physical crust with a relatively flat, smooth appearance that is in sharp contrast, visually, to raised undershrub mounds (Blackburn, 1975; Roundy et al., 1978; Wood et al., 1978, 1982; Doescher et al., 1984; Johnson and Gordon, 1988; Blackburn et al., 1990, 1992; Bolton et al., 1990; Goff et al., 1993; Pierson et al., 1994, 2001, 2002; Davies et al., 2007, 2009; Aanderud et al., 2008; Hooker et al., 2008). Laser altimetry indicated that burned surfaces were rougher than unburned surfaces, irrespective of microsite, in the absence of vegetation (Table 1; Fig. 3). Findings did not indicate that the difference in roughness between undershrub and interspace microsites was greater for unburned vs. burned surfaces (Table 1). Burning could either increase or decrease roughness through several different mechanisms. The combustion of organic materials during the burning process might result in a more homogenous and therefore smooth surface following burning. Conversely, spatial heterogeneity in Table 1 Coefficient values and significance levels (p) from mixed model analysis (57 degrees of freedom and n = 64 measurements) for response of surface roughness (at a moving window size of 90 × 90 mm, locRMSH-90) determined with the terrestrial laser scanner to the fixed effects of microsite (undershrub, interspace), burn treatment (burned, unburned), and interaction term (microsite x burn treatment), and with random effect of site. Statistically significant p-values have been denoted by *. Effect

Intercept Microsite Burn treatment Microsite × burn treatment

Surface roughness (locRMSH-90) Coefficient value

p

5.21 − 1.52 1.24 − 0.85

0.00* 0.01* 0.04* 0.33

Fig. 3. Surface roughness (locRMSH-90) aggregated by microsite (left column) and burn treatment (right column) for each study site. Values in the box plots are median (center bold line), 25th, and 75th percentile values (bottom and top of box, respectively), 1.5 times the interquartile range or, if smaller, the maximum value (error bars). Data points larger than 1.5 times the interquartile range are plotted individually as outliers.

Table 2 Coefficient values and significance levels (p) from mixed model analysis (53 degrees of freedom and n = 64 measurements) for response of the natural log of total mass of PM10 (particle diameter b 10 μ) emitted from the PI-SWERL (Portable In Situ Wind Erosion Laboratory) tests to the fixed effects of surface roughness (at a moving window size of 90 × 90 mm, locRMSH-90), microsite (undershrub, interspace), burn treatment (burned, unburned), and interaction terms, and with random effect of site. Statistically significant p-values have been denoted by *. Effect

Intercept locRMSH-90 Microsite Burn treatment locRMSH-90 × microsite locRMSH-90 × burn treatment Microsite × burn treatment locRMSH-90 × microsite × burn treatment

ln PM10 Coefficient value

p

0.68 0.63 0.40 0.50 − 0.58 − 0.68 − 0.46 0.64

0.00* 0.00* 0.03* 0.03* 0.03* 0.05* 0.09 0.10

J.B. Sankey et al. / Geomorphology 135 (2011) 181–190

187

Fig. 4. PM10 emissions vs. surface roughness (locRMSH-90) for each burned site. Each datum is a single PI-SWERL trial on either undershrub or interspace microsites (n = 8 trials per burned or unburned surface, including 4 on each microsite type). Upper panels show data stratified by burn effect, and lower panels show data stratified by microsite.

surface heating associated with wildland fire might result in burned microsites that are rougher than their unburned counterparts. It is possible that the greater surface roughness observed for burned surfaces, irrespective of microsites, was related to the patchy nature of soil heating and incomplete combustion of soil surface materials. Second-order effects that occur after fire could also alter roughness. Historic erosion during the time between wildfire and our measurements (which varied from 5 years to 2 months among sites) could potentially smooth microsites by removing erodible peaks or could increase roughness by creating microlocalized blowouts. Burned surfaces are less protected by vegetation and therefore more susceptible to impacts from raindrops that might roughen the surface. We did not detect any significant differences in the relative elevation of undershrubs and adjacent interspaces between burned and unburned surfaces. Therefore, there did not appear to be a substantial effect of burning on microsite morphology. This suggests that physical and/or biological mechanisms (i.e., wind or water transport, ungulate trampling, burrowing, annual grass, or woody plant invasion) in addition to burning might be required to alter microsite morphology (Pierson et al., 1994, 2001, 2002; Hilty et al., 2003; Li et al., 2008; Ravi et al., 2009; Field et al., 2010; Sankey et al., 2010). Nonetheless, roughness and the average height difference between undershrub and interspace microsites appeared to vary amongst sites (Figs. 4 and 5). The ecological and geomorphic drivers that produce variability in microsite morphology such as we observed amongst the study sites are a recommended focus for future research, which might be revealed through using TLS and the PI-SWERL or wind tunnel to

detect responses to experimental manipulation of surface roughness. Spatial and temporal variability in microsite morphology has implications for important environmental issues in desert shrublands, such as the success of rehabilitation treatments (Boyd and Davies, 2010) and wildlife (e.g., sage grouse, Centrocercus urophasianus) habitat (Crawford et al., 2004).

5.2. Dust emissions Surface roughness was a significant predictor of dust emissions (Table 2). Results reinforced the expectation that the fundamental relationship between erosion and surface roughness derived from laser altimetry is scale dependent. At landscape scales (meter–kilometer length scales), previous research has demonstrated that erosion is inversely related to surface roughness derived from laser altimetry for measures of wind erosion (Pelletier et al., 2009; Sankey et al., 2010). At submeter spatial scales and in the absence of nonerodible roughness features such as vegetation, however, we observed that a positive relationship between surface roughness and dust emissions was evident in particular for burned surfaces in this study (Fig. 4). We hypothesized that a positive relationship might exist at submeter spatial scales based on aeolian theory that indicates that a relatively wide range of grain and aggregate sizes, and therefore greater surface roughness, is required to produce the abrading action that drives emission of fine-grained dust (Bagnold, 1941; Pye, 1987; Nickling, 1988; Gillete and Chen, 2001).

188

J.B. Sankey et al. / Geomorphology 135 (2011) 181–190

Noman, and Moonshiner sites (Fig. 4, upper right panel). While variability in ash concentrations amongst recently burned surfaces might also contribute to PM10 emissions and therefore affect the relationship of dust emissions to surface roughness, the burned sites we studied appeared to have predominantly mineral soil materials at the surface, suggesting that ash produced by burning had been removed by wind relatively rapidly, post-fire. It is possible that some variability in dust emissions observed amongst the PI-SWERL trials might have been influenced by variability in surface shear that resulted from inherent differences in surface roughness between microsites and the fact that the PI-SWERL was calibrated for a flat plate (Etyemezian et al., 2007). In this study, we employed terrestrial LiDAR to provide insight into the scale-dependent nature of the relationship between soil erosion and surface roughness, as anticipated from fundamental theory of aeolian geomorphology. LiDAR has undeniable utility for quantifying a wide range of geomorphic processes (Tratt et al., 2008; Bauer, 2009). In aeolian research, future studies are required to determine how LiDAR roughness can be translated to estimates of aeolian model parameters such as aerodynamic roughness and zero plane displacement, as well as to directly incorporate LiDAR-derived roughness into model-based predictions of aeolian emissions. 6. Conclusion

Fig. 5. Height differences between microsite pairs (undershrub and interspace) aggregated by burn treatment for each study site.

Why the positive relationship between surface roughness and dust emissions was most consistently evident on the burned surfaces we studied is not entirely clear. Burned surfaces in desert shrublands are generally more erodible and are expected to have a greater supply of erodible mass than unburned surfaces (Ash and Wasson, 1983; Wasson and Nanninga, 1986; Zobeck et al., 1989; Wiggs et al., 1994, 1995, 1996; Whicker et al., 2002, 2006; Vermeire et al., 2005; Breshears et al., 2009; Ravi et al., 2009; Sankey et al., 2009a,b). Surface crusting is evident in the interspace microsites and is likely to inhibit sediment supply. However, these crusts appear to be resistant to wildfire, so an increase in sediment supply or decrease in erodibility due to disturbance of the crusts by fire is not a likely explanation for the apparent stronger relationship between surface roughness and dust emissions on burned surfaces (Hoover, 2010). While all sites were generally loamy surfaces, differences did exist in the proportion of PM10 in the soil surface which might have influenced the sediment supply and the relationship between roughness and dust emissions observed at sites. The Moonshiner site had a somewhat coarser soil texture (smaller fraction of PM10) compared to the other three sites, for example, nonetheless PM10 emissions from the PI-SWERL trials at this site were some of the largest observed and differences between emissions on burned and unburned surfaces were particularly substantial (Fig. 4). It is possible that the supply for dust emission was so low in the unburned surfaces we examined, that incremental changes in microsite surface roughness had very little effect on the dust emission response relative to the response observed on burned surfaces. Amongst burned surfaces, for example, we noted that the response of dust emissions to incremental increases in microsite surface roughness appeared to be flattest for the least recently burned Clover site and steeper for the more recently burned and more emissive Samaria,

This study examined relationships of dust emissions and surface roughness at submeter spatial scales experimentally in the field. Desert shrubland surfaces subject to a range of undisturbed as well as wildland fire conditions were examined. Results demonstrated that a positive relationship between surface roughness and dust emissions existed at submeter spatial scales for burned surfaces in particular, and that the strength of the relationship varied with the scale for which roughness was calculated. This finding is expectedly in contrast to previous findings that have demonstrated an inverse relationship exists between soil erosion and landscape surface roughness determined with laser altimetry at much coarser spatial scales. Results therefore provide an example of how the erosion and surface roughness relationship varies fundamentally across spatial scales. Acknowledgments This material is based on work supported in part by the U.S. Army Research Laboratory and the U.S. Army Research Office under grant number W911NF-07-1-0481 and by the Bureau of Land Management under grant number L08AC14585. Funding to acquire the TLS was provided by the University of Idaho, Idaho NSF EPSCoR, and by the National Science Foundation under award number EPS-0814387. We thank Michelle Wiest for her assistance with statistical analysis. We thank Joanna Nield, Robert Washington-Allen, and one anonymous reviewer for their comments on previous versions of the manuscript. References Aanderud, Z.T., Shuldman, M.I., Drenovsky, R.E., Richards, J.H., 2008. Shrub-interspace dynamics alter relationships between microbial community composition and belowground ecosystem characteristics. Soil Biology and Biochemistry 40, 2206–2216. Anderson, J.E., Inouye, R.S., 2001. Landscape-scale changes in plant species abundance and biodiversity of a sagebrush steppe over 45 years. Ecological Monographs 71, 531–556. Ash, J.E., Wasson, R.J., 1983. Vegetation and sand mobility in the Australian desert dunefield. Zeitschrift für Geomorphologie Supplementband 45, 7–25. Baas, A.C.W., 2008. Challenges in aeolian geomorphology: investigating aeolian streamers. Geomorphology 93, 3–16. Bagnold, R.A., 1941. The Physics of Blown Sand and Desert Dunes. Methuen, New York. Bauer, B.O., 2009. Contemporary research in aeolian geomorphology. Geomorphology 105, 1–5. Blackburn, W.H., 1975. Factors influencing infiltration and sediment production of semiarid rangelands in Nevada. Water Resources Research 11, 929–937.

J.B. Sankey et al. / Geomorphology 135 (2011) 181–190 Blackburn, W.H., Pierson, F.B., Seyfried, M.S., 1990. Spatial and temporal influence of soil frost on infiltration and erosion of sagebrush rangelands. American Water Resources Association 26, 991–997. Blackburn, W.H., Pierson, F.B., Hanson, C.L., Thurow, T.L., Hanson, A.L., 1992. The Spatial and Temporal Influence of Vegetation on Surface Soil Factors in Semiarid Rangelands, Trans. ASAE 35, 479–486. Bolton, H., Smith, J.L., Wildung, J.E., 1990. Nitrogen mineralization potentials of shrubsteppe soils with different disturbance histories. Soil Science Society of America Journal 54, 887–891. Boyd, C.S., Davies, K.W., 2010. Shrub microsite influences post-fire perennial grass establishment. Rangeland Ecology and Management 63, 248–252. Breshears, D.D., Whicker, J.J., Zou, C.B., Field, J.P., Allen, C.D., 2009. A conceptual framework for dryland aeolian sediment transport along the grassland–forest continuum: effects of woody plant canopy cover and disturbance. Geomorphology 105, 28–38. Bullard, J.E., 2006. Arid geomorphology. Progress in Physical Geography 30, 542–552. Busacca, A.J., Beget, J.E., Markewich, H.W., Muhs, D.R., Lancaster, N., Sweeney, M.R., 2004. Eolian sediments. In: Gillespie, A.R., Porter, S.C., Atwater, B.R. (Eds.), The Quaternary Period in the United States. Elsevier, Amsterdam, The Netherlands, pp. 275–310. Charley, J.L., West, N.E., 1975. Plant-induced soil chemical patterns in some shrubdominated semi-desert ecosystems of Utah. Journal of Ecology 63, 945–963. Clawges, R., Vierling, L.A., Calhoon, M., Toomey, M.P., 2007. Use of a ground-based scanning lidar for estimation of biophysical properties of western larch (Larix occidentalis). International Journal of Remote Sensing 28, 4331–4344. Crawford, J.A., Olson, R.A., West, N.E., Mosley, J.C., Schroeder, M.A., Whitson, T.D., Miller, R.F., Gregg, M.A., Boyd, C.S., 2004. Ecology and management of sage-grouse and sage-grouse habitat. Journal of Range Management 57, 2–19. Davies, K.W., Bates, J.D., Miller, R.F., 2007. The influence of Artemesia tridentata ssp. wyomingensis on microsite and herbaceous vegetation heterogeneity. Journal of Arid Environments 69, 441–457. Davies, K.W., Bates, J.D., James, J.J., 2009. Microsite and herbaceous vegetation heterogeneity after burning Artemisia tridentata steppe. Oecologia 159, 597–606. De Vries, A.C., Ritchie, J.C., Menenti, M., Kustas, W.P., 1997. Aerodynamic roughness estimated from surface features for a coppice dune area using laser altimeter data. 12th Symposium on Boundary Layers and Turbulence, 12, pp. 289–290. Doescher, P.S., Miller, R.F., Winward, A.H., 1984. Soil chemical patterns under eastern Oregon plant communities dominated by big sagebrush. Soil Science Society of America Journal 48, 659–663. Dust-Quant LLC, 2008. User's Guide for the Miniature Portable In-situ Wind Erosion Lab (PI-SWERL): Version 1.2. Dust-Quant LLC, Las Vegas, NV, USA. Eitel, J.U.H., Vierling, L.A., Long, D.S., 2010. Simultaneous measurements of plant structure and chlorophyll content in broadleaf saplings with a terrestrial laser scanner. Remote Sensing of Environment 114, 2229–2237. Eitel, J.U.H., Williams, C.J., Vierling, L.A., Al-Hamdan, O.Z., Pierson, F.B., 2011. Suitability of terrestrial laser scanning for studying surface roughness effects on concentrated flow erosion processes in rangelands. Catena 87, 398–407. Etyemezian, V., Nikolich, G., Ahonen, S., Pitchford, M., Sweeney, M., Purcelle, R., Gillies, J., Kuhns, H., 2007. The portable in situ wind erosion laboratory (PI-SWERL): a new method to measure PM10 windblown dust properties and potential for emissions. Atmospheric Environment 41, 3789–3796. Field, J.P., Belnap, J., Breshears, D.D., Neff, J.C., Okin, G.S., Whicker, J.J., Painter, T.H., Ravi, S., Reheis, M.C., Reynolds, R.L., 2010. The ecology of dust: local- to global-scale terrestrial perspectives. Frontiers in Ecology and the Environment 8, 423–430. Fryrear, D.W., Saleh, A., 1993. Field wind erosion — vertical-distribution. Soil Science 155, 294–300. Gillete, D.A., Chen, W., 2001. Particle production and aeolian transport from a “supply limited” source area in the Chihuahuan desert, New Mexico, United States. Journal of Geophysical Research 106 (D6), 5267–5278. Goff, B.F., Bent, G.C., Hart, G.E., 1993. Erosion response of a disturbed sagebrush steppe hillslope. Journal of Environmental Quality 22, 698–709. Guarneri, A., Vettore, A., Pirotti, F., Menenti, M., Marani, M., 2009. Retrieval of smallrelief marsh morphology from terrestrial laser scanner, optimal spatial filtering, and laser return intensity. Geomorphology 113, 12–20. Haubrock, S.-N., Kuhnert, M., Chabrillat, S., Güntner, A., Kaufmann, H., 2009. Spatiotemporal variations of soil surface roughness from in-situ laser scanning. Catena 79, 128–139. Hilty, J.H., Eldridge, D.J., Rosentreter, R., Wicklow-Howard, M.C., 2003. Burning and seeding influence soil surface morphology in an artemisia shrubland in southern Idaho. Arid Land Research and Management 17, 1–11. Hodge, R., Brasington, J., Richards, K., 2009a. In situ characterization of grain-scale fluvial morphology using terrestrial laser scanning. Earth Surface Processes and Landforms 34, 954–968. Hodge, R., Brasington, J., Richards, K.S., 2009b. Analysing laser-scanned digital terrain models of gravel bed surfaces: linking morphology to sediment transport processes and hydraulics. Sedimentology 56, 2024–2043. Hooker, T.D., Stark, J.M., Urszula, N.A., Leffler, J., Peek, M., Ryel, R., 2008. Distribution of ecosystem C and N within contrasting vegetation types in a semiarid rangeland in the Great Basin, USA. Biogeochemistry 90, 291–308. Hoover, A.N., 2010. Spatial heterogeneity persists among soil microsites and native, but not exotic, plants following post-fire wind erosion. M.S.Thesis, Department of Biological Sciences, Idaho State University, Pocatello, ID, USA. Johnson, C.W., Gordon, N.D., 1988. Runoff and erosion from rainfall simulator plots on sagebrush rangeland. Transactions of ASAE 31, 421–427. King, J., Nickling, W.G., Gillies, J.A., 2008. Investigations of the law-of-the-wall over sparse roughness elements. Journal of Geophysical Research 113, F02S07. doi:10.1029/2007JF000804.

189

Knight, D.H., 1994. Mountains and Plains: The Ecology of Wyoming Landscapes. Yale University Press, USA. Li, J., Okin, G.S., Alvarez, L.J., Epstein, H.E., 2008. Effects of wind erosion on the spatial heterogeneity of soil nutrients in two desert grassland communities. Biogeochemistry 8, 73–88. Maurer, T., Herrmann, L., Stahr, K., 2010. Wind erosion characteristics of Sahelian surface types. Earth Surface Processes and Landforms 35, 1386–1401. Menenti, M., Ritchie, J.C., 1994. Estimation of effective aerodynamic roughness of Walnut Gulch watershed with laser altimeter measurements. Water Resources Research 30, 1329–1337. Munson, S.M., Belnap, J., Okin, G.S., 2011. The responses of wind erosion to climate induced vegetation changes on the Colorado Plateau. Proceedings of the National Academy of Sciences 108, 3854–3859. Neakrase, L.D.V., Greeley, R., 2010. Dust devils in the laboratory: effect of surface roughness on vortex dynamics. Journal of Geophysical Research 115, E05003. doi:10.1029/2009JE003465. Nickling, W.G., 1988. The initiation of particle movement by wind. Sedimentology 35, 499–511. Nield, J.M., Wiggs, G.F., 2011. The application of terrestrial laser scanning to aeolian saltation cloud measurement and its response to changing surface moisture. Earth Surface Processes and Landforms 36, 513–522. Nield, J.M., Wiggs, G.F.S., Squirrell, R.S., 2011. Aeolian sand strip mobility and protodune development on a drying beach: examining surface moisture and surface roughness patterns measured by terrestrial laser scanning. Earth Surface Processes and Landforms 36, 273–278. NRCS Web Soil Survey, 2010. Accessed 7/29/2010 http://websoilsurvey.nrcs.usda. gov2010. Okin, G.S., 2008. A new model of wind erosion in the presence of vegetation. Journal of Geophysical Research 113, F02S10. Pelletier, J.D., Mitasova, H., Harmon, R.S., Overton, M., 2009. The effects of interdune vegetation changes on eolian dune field evolution: a numerical-modeling case study at Jockey's Ridge, North Carolina, USA. Earth Surface Processes and Landforms 34, 1245–1254. Pierson, F.B., Blackburn, W.H., Van Vactor, S.S., Wood, J.C., 1994. Partitioning small scale spatial variability of runoff and erosion on sagebrush rangeland. Water Resources Bulletin 30, 1081–1089. Pierson, F.B., Robichaud, P.R., Spaeth, K.E., 2001. Spatial and temporal effects of wildfire on the hydrology of a steep rangeland watershed. Hydrological Processes 15, 2905–2916. Pierson, F.B., Carlson, D.H., Spaeth, K.E., 2002. Impacts of wildfire on soil hydrological properties of steep sagebrush rangelands. International Journal of Wildland Fire 11, 145–151. Pinheiro, J.C., Bates, D.M., 2000. Mixed-Effects Models in S and S-PLUS. Springer, New York, NY, U.S.A. Pye, K., 1987. Aeolian Dust and Dust Deposits. Academic Press Inc., Orlando, FL. R Development Core Team, 2008. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria3-900051-07-0. URL http://www.R-project.org. Rango, A., Chopping, M., Ritchie, J., Havstad, K., Kustas, W., Schmugge, T., 2000. Morphological characteristics of shrub coppice dunes in desert grasslands of southern New Mexico derived from scanning LiDAR. Remote Sensing of Environment 74, 26–44. Raupach, M.R., Gillette, D.A., Leys, J.F., 1993. The effect of roughness elements on wind erosion threshold. Journal of Geophysical Research 98 (D2), 3023–3029. Ravi, S., D'Odorico, P., Wang, L., White, C.S., Okin, G.S., Macko, S.A., Collins, S.L., 2009. Post-fire resource redistribution in desert grasslands: a possible negative feedback on land degradation. Ecosystems 12, 434–444. Ritchie, J.C., Everitt, J.H., Escobar, D.E., Jackson, T.J., Davis, M.R., 1992. Airborne laser measurements of rangeland canopy cover and distribution. Journal of Range Management 45, 189–193. Roundy, B.A., Blackburn, W.H., Eckert, R.E., 1978. Influence of prescribed burning on infiltration and sediment production in the pinyon-juniper woodland, Nevada. Journal of Range Management 31, 250–253. Sankey, J.B., Germino, M.J., Glenn, N.F., 2009a. Aeolian sediment transport following wildfire in a sagebrush steppe. Journal of Arid Environments 73, 912–919. Sankey, J.B., Germino, M.J., Glenn, N.F., 2009b. Relationships of post-fire aeolian transport to soil and atmospheric conditions. Aeolian Research 1, 75–85. Sankey, J.B., Glenn, N.F., Germino, M.J., Gironella, A.I.N., Thackray, G.D., 2010. Relationships of aeolian erosion and deposition with LiDAR-derived landscape surface roughness following wildfire. Geomorphology 119, 135–145. Schlesinger, W.H., Raikes, J.A., Hartley, A.E., Cross, A.F., 1996. On the spatial pattern of soil nutrients in desert ecosystems. Ecology 77, 364–374. Seefeldt, S., Germino, M.J., DiChristina, K.M., 2007. Prescribed fires have minor and transient effects on herbaceous vegetation cover and composition. Applied Vegetation Science 10, 249–256. Sutton, S.L.F., McKenna-Neuman, C., 2008. Variation in bed level shear stress on surfaces sheltered by nonerodible roughness elements. Journal of Geophysical Research 113, F03016. doi:10.1029/2007JF000967. Sweeney, M., Etyemezian, V., Macpherson, T., Nickling, W., Gillies, J., Nikolich, G., McDonald, E., 2008. Comparison of PI-SWERL with dust emission measurements from a straight-line field wind tunnel. Journal of Geophysical Research 113 (F01012). doi:10.1029/2007JF000830. Tratt, D.M., Neff, J.M., Valinia, A., 2008. Analysis of laser remote sensing technology needs in the Earth sciences: a decadal-scale outlook. Journal of Applied Remote Sensing 2 (023546). doi:10.1117/1.3036940. Turpin, C., Badr, T., Harion, J.L., 2010. Numerical modelling of aeolian erosion over rough surfaces. Earth Surface Processes and Landforms 35, 1418–1429.

190

J.B. Sankey et al. / Geomorphology 135 (2011) 181–190

Vermeire, L.T., Wester, D.B., Mitchell, R.B., Fuhlendorf, S.D., 2005. Fire and grazing effects on wind erosion, soil water content, and soil temperature. Journal of Environmental Quality 34, 1559–1565. Wasson, R.J., Nanninga, P.M., 1986. Estimating wind transport of sand on vegetated surfaces. Earth Surface Processes and Landforms 11, 505–514. Whicker, J.J., Breshears, D.D., Wasiolek, P.T., Kirchner, T.B., Tavani, R.A., Schoep, D.A., Rodgers, J.C., 2002. Temporal and spatial variation of episodic wind erosion in unburned and burned semiarid shrubland. Journal of Environmental Quality 31, 599–612. Whicker, J.J., Pinder, J.E., Breshears, D.D., 2006. Increased wind erosion from forest wildfire: implications for contaminant-related risks. Journal of Environmental Quality 35, 468–478. Wiggs, G.F., Livingstone, I., Thomas, D.S.G., Bullard, J.E., 1994. Effect of vegetation removal on airflow patterns and dynamics in the southwest Kalahari desert. Land Degradation and Rehabilitation 5, 13–24.

Wiggs, G.F., Thomas, D.S.G., Bullard, J.E., Livingstone, I., 1995. Dune mobility and vegetation cover in the southwest Kalahari. Earth Surface Processes and Landforms 20, 515–529. Wiggs, G.F., Livingstone, I., Thomas, D.S.G., Bullard, J.E., 1996. Airflow and roughness characteristics over partially vegetated linear dunes in the southwest Kalahari desert. Earth Surface Processes and Landforms 21, 19–34. Wood, M.K., Blackburn, W.H., Eckert, R.E., Peterson, F.F., 1978. Interrelations of the physical properties of coppice dune and vesicular dune interspace soils with grass seedling emergence. Journal of Range Management 31, 189–192. Wood, M.K., Eckert, R.E., Blackburn, W.H., Peterson, F.F., 1982. Influence of crusting soil surfaces on emergence and establishment of crested wheatgrass, squirreltail, thurber needlegrass, and fourwing saltbush. Journal of Range Management 35, 282–287. Zobeck, T.M., Fryrear, D.W., Pettit, R.D., 1989. Management effects on wind-eroded sediment and plant nutrients. Journal of Soil and Water Conservation 44, 160–163.