Granulometry and geochemistry of dust emission from Owens (dry) Lake, California

Owens (dry) Lake, the terminus of Owens River, is located in the Basin and Range physiographic province in east-central California. The transformation of Owens Lake into a desiccated desert playa (approximately 280 km 2) via diversion of the river into the Los Angeles Aqueduct (in 1913) led to extensive wind erosion, making the playa one of the most intense sources of airborne dust in the Western Hemisphere. The processes that direct the chemical and mineralogical composition of surface sediments at Owens (dry) Lake and the physical mechanisms of wind erosion, saltation and dust emission at Owens (dry) Lake are well-understood. However, prior research has not investigated the effects of spatial and temporal variations of individual aeolian dust events on the particle size distribution and elemental composition of dust aerosols as they are emitted. This study investigates the variations of particle size (measured by laser diffraction) and elemental composition of individual aeolian sediment samples collected from four sequential wind events (approximately one week apart) at six different heights above the land surface at seven different locations along an 1.2 km upwind-downwind sampling transect on the Owens playa. Laser diffraction’s ability to measure grains in their dry state as collected from a dust trap, as was done here, allows us to measure the actual size of the often-agglomerated, salt-bearing particles moving in the wind, as opposed to other techniques which require the disaggregation of particles and/or dissolution of salts in liquid. The samples also presented the opportunity to inter-compare two elemental analysis techniques (Particle-Induced X-Ray Emission Spectrometry [PIXE] and Inductively-Coupled Plasma Atomic Emission Spectroscopy [ICP-AES]) as complementary methods for analysis of airborne dusts and sediments. Results show that dust from Owens (dry) Lake contains several potentially toxic elements including As, Cd, Co, Cr, Pb, Mo and Zn with possible detrimental health effects and environmental impacts. Spatial and temporal variations within and between individual dust storms were reflected in the volumetric particle size distributions and elemental compositions of the dust. For example, dust grain size was strongly correlated at a significance level of α = 0.01 to the height at which the particles were captured above the playa surface and showed a significant difference (p \u3c 0.05) between individual wind storms (becoming coarser with each sequential dust event), but did not present a significant correlation with location along the linear transect where samples were collected. The overall mean percent volume of finer particles (\u3c20 μm) was greater than the overall mean percent volume of coarser particles (\u3e250 μm). The analysis of the mean percent volume of dust particles in eleven size classes with respect to height allowed the determination of the height of transition between particles moving in saltation and suspension, an essential parameter in equations developed to investigate vertical and horizontal flux of particles from eroding surfaces. Elemental concentrations of the dust varied with height above the land surface, distance downwind of the dust source, and between individual dust storms, as well as with dust particle size. For example, Na and S concentrations (indicative of a layer of efflorescent salts atop the playa) covaried with the concentrations of submicron to siltsized particles during each event, increased with height above the ground, and decreased with distance downwind of the dust source and with each sequential dust storm, suggesting quick and efficient removal of fine salts from the playa. Al, Ti, Mn, K, Fe, and Rb concentrations covaried with the percentages of fine/medium sand particles in the first two events and coarse sand in the third event, suggesting these elements were borne on saltating particles of clastic sediments. The combination of techniques reveals particle size/chemical fractionation and spatial variability of sediment properties during dust emission at aeolian “hotspots,” with implications to dust emission modeling, geochemical cycling, and aerosol source/receptor relationships. Statistically significant differences (p \u3c 0.05) were shown between PIXE and ICP-AES analytical results for splits of the same dust samples. Variability between PIXE and ICP data is explained by inherent differences between the analytical techniques, sample preparation methods, and/or variability in sample matrices. For example, PIXE determined higher concentrations of Al, As, Cu, Fe, K, Mn, Ni, Sr, Ti, and Zn in all samples. Combined with low ICP-AES percent recoveries of Al, Fe, K, Mn, and Ti for NIST SRM 2710, these results indicated incomplete dissolution during digestion (EPA Method 3050B) of elements bound in silicate structures. These findings point out the care which must be taken in selection of elemental analysis methods for environmental samples and in evaluation of elemental concentrations obtained by different techniques