NG41C-0666
Surface-sediment dynamics in a dust source from thermal infrared remote sensing
Characteristics of surface sediments are significant factors in modeling dust entrainment and wind erosion, and it is of interest to monitor them using remote sensing in source areas at high spatial and temporal resolution. A time-series of Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) data were acquired for Soda Playa (CA), a modern depositional environment associated with dust emission. Analysis of the multispectral thermal infrared (TIR) images indicates that the type and distribution of the surface sediments can be mapped by linear spectral unmixing techniques. Image-based spectral endmembers extracted from the ASTER five-band surface emissivity data were used to drive fraction images. The spectral-mixture analysis reveals that the mosaic-like pattern of the main sediment types - silica-rich, clay-rich, and salt-rich, changes in time as a consequence of interactions between hydrologic and geomorphic processes in the playa environment. The results highlight the dynamic response of the playa-surface to wind erosion, and suggest that this technique is useful for continuously detecting dust emission potential in sources characterized by a small extension and a complex surface.
NG41C-0667
Integration of ASTER thermal infrared data and the Google Earth application to examine the relationship between sand transport pathways and dust emission hot-spots
The identification and characterization of major mineral dust source areas from the Sahara Desert and the composition of these particulates are critical to our understanding of global dust production, models of atmospheric emission and transport, and determination of radiative properties of dust plumes from desert regions. The locations of many of these dust hot-spots have been determined using data such as the Total Ozone Mapping Spectrometer aerosol index (TOMS AI), showing that the largest sources of global dust are located in the Sahara Desert of west and central Africa. Studies using geochemical tracers, surface observations, analysis of meteorological data, and remote sensing have, however, arrived at different and sometimes conflicting delineations of Saharan dust sources. Geomorphic environment is a major control on dust emission source and process. Surface and remote sensing observations of the Bodele Depression, the largest dust source in the Sahara, show that sand saltation on the playa surface and the abrasion of material are the primary processes in generating dust. In order to test the hypothesis that other major natural global dust emission sources are linked to the interaction of sand transport along defined pathways and dust sources, multispectral thermal infrared (TIR) data from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) are integrated with the high resolution, mosaicked image data available from the Google Earth application. The geomorphic environment, surface composition and their relative importance as a control on dust emissions are assessed using high resolution data at dust hot-spots that were identified by the TOMS AI at the much lower spatial resolution of one pixel/degree.
NG41C-0668
A framework of field observations and spatial data for understanding dust emissions in the Mojave Desert
Modeling dust events at landscape to regional scales requires field observations of dust-source characteristics, mapping of source types by remote sensing, and wind fields representing the conditions that mobilize dust from the surface. A conceptual framework has been built for understanding dust-source types and their dynamics in the Mojave Desert. Observations of dust events in the Mojave indicate five general source types: 1) Sparsely vegetated surfaces that are vulnerable during periods of drought; 2) wet playas where a near-surface groundwater table generates "fluffy" (very soft sediment) conditions; 3) transitional playas where groundwater extraction has lowered the water table, and playa surface composition produces sediments that are vulnerable to erosion; 4) ephemeral flood deposits; and 5) anthropogenic sources where off-road vehicles, military training exercises, and dirt roads create a disturbed surface. Some sources are perennial and others are strongly influenced (sometimes in opposite ways) by precipitation cycles. A multi-year study of precipitation, vegetation, winds, and saltation at several plots in the Mojave National Preserve shows that blooms of annual vegetation in wet years can leave biomass that protects the surface for more than a year after the rains. Monitoring of the wet Franklin Lake Playa shows that a shallow ground-water table is associated with more vulnerable conditions for dust emission. Repeat photography of the relations between winds and dustiness at transitional Mesquite Lake Playa shows that dust is mobilized during the spring when winds are greater than about 5 m/s. Satellite images reveal dust emission from ephemeral fluvial systems, such as the Mojave River Sink, at the end of wet spring seasons. Satellite images also document dust emissions from areas of heavy military and off-road vehicle activity. Landsat imagery was used to map perennial vegetation cover for the Mojave Desert, calibrated to 250 field transects. The perennial vegetation map was used in conjunction with a Landsat-derived surface-sheltering map to identify areas vulnerable to dust emissions due to sparse vegetation. A map of playa lakes, derived from USGS 1:100,000 maps, satellite imagery, and field observations establishes the boundaries of wet, transitional, and dry playas. Wind fields generated from a regional climate model show that intense winds in vulnerable regions are associated with contemporaneous dust plumes observed from geostationary satellites during the major dust storms of January 5, 2007 and March 27, 2007. Through systematic collection of data from the scale of field plots to satellite images, a framework for the modeling of dust generation for a geographically complex and temporally dynamic region like the Mojave Desert is being created.
NG41C-0669
The Relationship of Land Cover to Aeolian Dust Production at the Jornada Basin, New Mexico, USA
Vegetation tends to reduce aeolian transport of surface sediments. However, not all vegetation types act in the same way to do so. In general, the more land cover the less erosion will occur; thus grasslands should experience less aeolian erosion than shrublands, which are characterized by patchy cover with open intershrub spaces. Five major ecosystem types are described at the Jornada Basin Long- Term Ecological Research site (Jornada LTER) in south-central New Mexico, USA: mesquite dunes, black grama grasslands, creosote bush shrublands, tarbush alluvial flats, and grass-dominated playas. Here we investigate the dry particle size distribution of material collected by BSNE aeolian particle samplers in 2006 in these five different vegetation types, allowing us to estimate dust production at sites with different land cover. As mesquite and creosote bush continue replacing historical grasslands at Jornada, understanding the characteristics of wind erosion will be important for future management plans. The mesquite sites had the greatest horizontal mass flux, although with substantial variation. M-NORT, a site with large sand dunes, had much greater mass flux than other mesquite sites. For most sites, the dry particle size distributions at 5, 10 and 20 cm heights above the land surface were very similar, dominated by sand, while the distributions for 50 and 100cm heights shifted towards a greater percentage of silt and clay (dust) particles. The playa site and one of three tarbush sites stand out as having the greatest percentages of dust particles, between 33- 52 % of total mass at all heights. After taking into account the differences in mass flux, the mesquite site with the larger dunes and the playa site had the greatest flux of dust-sized particles. These two sites demonstrate different mechanisms of producing dust at the Jornada LTER. The playa is a relatively major dust producer due to its high proportion of fine particles, whereas the mesquite site is a major dust producer because of its far greater overall mass flux. Considering the differential mass flux of aeolian particles produced by each vegetation type, the greatest total mass of dust per unit area at the Jornada LTER occurs at the playa site, followed by the mesquite site with large dunes and tarbush, while creosote bush, grassland, and the other two mesquite sites produce the least dust. Since mesquite dunes cover about 34% of the Jornada Basin, we conclude that mesquite dunes have the potential to produce the most dust from the area overall, with playas and tarbush-dominated alluvial flats (which cover about 7% each) having the potential to emit large amounts of dust if the conditions are correct. These results suggest that the continuing replacement of grasslands with shrubs will likely increase dust emissions from the Jornada Basin.
NG41C-0670
Post-fire Wind Erosion in a Semiarid Shrub Steppe
The objective of this project is to determine and describe the effects of fire on the potential for soil erosion by wind in a semiarid shrub steppe in southeastern Idaho. We apply a previously developed method for determining the threshold wind speed (critical threshold) required to initiate saltation of soil particles at several sampling locations within wildfires that burned in summers 2006 and 2007. Sampling locations have data collection stations with anemometers, a piezoelectric sensor mounted at 5 cm above the ground surface that records impacts from saltating soil particles, temperature and relative humidity sensors mounted at ground level, soil moisture sensors, and soil erosion bridges. Analysis is intended to identify key controlling variables of post- wildfire eolian transport with the intention of continued monitoring over longer time periods with the incorporation of LiDAR data for surface characterization. Initial field results indicate substantial differences in wind erosion potential between burned and unburned sites. Saltation activity is greater and more frequently detected at the burned sites in comparison to the unburned sites. Comparison between the burned sites indicates periods of similar saltation activity and threshold wind speeds. Comparison also shows periods when saltation is detected at one burned site but not the other, and periods when greater wind speeds are required at one burned site to initiate saltation. This suggests that within-burn variability exists in wind erosion potential and requires more complete characterization to accurately model and predict post-fire eolian transport. In parallel to the field-based research, we are developing remote sensing techniques to characterize the soil surface and vegetation communities at a scale appropriate for wind erosion modeling. These techniques include hyperspectral and LiDAR analysis of the soil surface and structural information of the vegetation. Soil and vegetation surface roughness information from LiDAR will be used to compare wind speed thresholds in the field areas. This work has important implications for land rehabilitation, soil conservation efforts, and human health.
NG41C-0671
The Non-Linear Response of Aeolian Sand Dunes to Changing Climatic Conditions and Vegetation Cover
Partially vegetated dunes have been viewed as relics from a drier and/or windier past. However, research in the southwest Kalahari (e.g. Wiggs et al, 1995; Bullard et al, 1996) confirmed our understanding that sediment movement can occur under vegetation canopies and that dune morphology is tied to the dynamics of vegetation, showing a staggered and/or damped response to a changing climatic, grazing or burning regime. Evidence suggested that dunes which are stripped of vegetation through natural or human disturbance exhibit an order of magnitude increase in surface sediment activity. A threshold vegetation cover below which dune activity markedly increased was recognised at about 14%. Questions remain as to the rate of vegetation recovery after such disturbances and how quickly the regulatory effect of vegetation on dune dynamics can be re-established. This paper reports findings of a 15-year programme monitoring dune activity under a re-establishing vegetation cover after a severe disturbance in the southwest Kalahari. Results indicate that even in prime growing conditions it can take up to a decade for disturbed vegetation to re- establish and regain control of dune dynamics. However, there is a contrast in the recovery rate on different landscape components with vegetation renewal in interdunes being twice as rapid as that measured in crestal regions. These results are discussed in the context of (i) the non-linear response of dunes to changing climatic forcings; and (ii) the calibration of models of dune dynamics with regard to future predicted climatic warming and drying. http://www.ouce.ox.ac.uk/research/arid-environments/
NG41C-0672
Aeolian Sand Transport by Boundary Layer Turbulence
The erratic and intermittent nature of wind-driven sand transport challenges our current transport models, which lack physical mechanisms for explaining and taking into account this spatio-temporal variability. This paper presents a collective overview of results from investigations into the nature of spatio-temporal variability in sand transport generally, and the formation and behaviour of aeolian streamers specifically. This includes three principal studies. First, the results of field investigations into the formation and behaviour of aeolian streamers in coastal and desert environments, where spatio-temporal transport variability and associated turbulence characteristics were assessed with an extensive instrument array. Streamers were measured with a transverse array of Safires, while the wind field and associated turbulent structures were monitored with cup-anemometry and a rake of hot-film probes. Second, these field data were used to assess the statistical trends in transport variability as a function of spanwise scale of measurement and the temporal scale of time-averaging transport rates. Third, spectral wavelet analysis of high-frequency collocated wind speed (hot- film probes) and transport flux (Safires) time-series revealed distinct forcing-response regimes at different temporal scales. The transitions between these regimes and their ranges compare favourably with physically meaningful scales, such as the minimum temporal scale of saltation response to wind speed fluctuations, and the integral time-scale of the observed internal boundary layer turbulence dynamics. The paper concludes with a tentative conceptual framework that attempts to integrate the results and insights from these studies towards an improved understanding of aeolian sediment transport processes.
NG41C-0673
Point Pattern Analysis of Star-Dune Fields
Star dunes are among the largest and most complex aeolian dunes in nature. Varying morphologies of star dunes are well documented; however, the dune-field scale properties of the pattern have received relatively little attention. This study addresses the spatial organization of star dune field patterns in the Erg Oriental, Edeyen Murzuq, Rub-al-Khali and the Gran Desierto. Areas targeted in each dune field display a transition from a simple dune pattern, in which only star forms occur, to a complex dune pattern where star dunes occur superimposed on relict linear and crescentic dune topography. Star-dune peaks determined from SRTM 90 digital elevation data are treated as points for point pattern analysis. Nearest-neighbor statistics are calculated across each dune field over 2500 sqkm intervals to characterize changes in the pattern. Dune peak spacing in simple star-dune patterns is highly disperse (R = 1.8), indicating a significant departure from a random point pattern. Simple star patterns also show a strong correlation between nearest neighbor spacing and height. Complex star dune patterns show a lower degree of dispersion and a weaker correlation between nearest neighbor spacing and height. Ultimately, these differences reflect both the control of the relict-dune pattern on the organization of the superimposed star-dune pattern and the overall maturity of the star-dune pattern.