Sources, compositions, and effects of atmospheric dust from American Drylands

Science Center Objects

The Drylands Project undertakes studies to measure past and ongoing changes in dust sources, flux, and composition in the American West, and strives to understand the effects of atmospheric dust on pressing national and global issues of snowmelt acceleration, air quality, and human health. The project develops the capability to forecast future dust emission/deposition and effects on the basis of episodes of accelerated aeolian activity recorded in Quaternary and historical deposits.

We and our collaborators measure properties of geologic materials to identify minerals and to determine chemistry (including forms of carbon), isotopes for selected elements (mostly strontium, lead, and 137-cesium), texture (particle sizes), magnetic properties that identify iron-oxide minerals, radiative properties over UV to NIR wavelengths, and salt content by electrical conductivity. In addition, we observe samples using various kinds of microscopy, including reflected-light and scanning electron microscopy. Samples include atmospheric mineral dust, surficial sediments likely to become dust, soils, and rocks.

Dust on San Juan Mountains snowpack
Dust darkens the surface of snowpack in the San Juan Mountains, Colorado on March 12 2009. 

Effects of dust on mountain snow pack

Dust deposited onto mountain snow pack can greatly decrease snow albedo (the capacity of bright surfaces to reflect solar radiation). These radiative effects can significantly increase the rate of snow melt, and rapid spring-time snow melt translates into early melt runoff (see Painter, T. H., A. P. Barrett, C. Landry, J. Neff, M. P. Cassidy, C. Lawrence, K. E. McBride, and G. L. Farmer, Impact of disturbed desert soils on duration of mountain snowcover, Geophysical Research Letters, doi:10.1029/2007GL030284R).

During years of heavy dust fall, these effects may then result in water delivery to streams, rivers, and reservoirs earlier than is desirable. Water runoff may even exceed the capacity of water storage systems to retain water for use later in the summer when water is needed most for downstream arid lands and communities.

The problem can be acute for water resources in the Colorado River basins above and below the Glen Canyon Dam. Much of the water in these basins comes from snow pack in mountains close to the deserts of the Colorado Plateau. Throughout the 2000s, dust storms have been frequent on parts of the Plateau, such as along the Little Colorado River dust corridor.

Snow pit for collecting dust from snowpack
Snow pit for collecting dust from snowpack, San Juan Mountains, Colorado, 13 March 2009. The dark surface on the snow is caused by a dust layer. Other dust layers can be seen within snowpack. The Colorado Dust on Snow (CODOS) program runs the monitoring site and collects dust. See also information from the Center for Snow & Avalanche Studies.(Credit: Chris Landry. Public domain.)

Different minerals in dust will have differing effects on snow albedo. When exposed to the sun, dark-colored particles exhibit "black box" effects, heating surrounding snow crystals. Some minerals, including varieties of iron oxides (e.g., hematite and goethite having iron in the 3+ oxidized state), are strong absorbers of solar radiation and can thereby warm when exposed to sunlight. In our project, we will investigate connections between iron oxide minerals and the radiative properties of dust. Better understanding of dust mineralogy will likely improve comprehension about the effects of dust on snow melt. Such understanding may also bear closely on issues of human health (iron redox in lungs, as an example), atmospheric radiative effects, and perhaps even ocean fertilization.

Additional Information:  

Preliminary findings describing the properties of dust deposited to Colorado snow cover during Water Year 2013 are available under the Results tab.

 

Remotely Sensed Imaging to Locate Dust Sources

Objectives are to determine the location, size, frequency, duration, and transport patterns of dust events in the western North America. We employ both low- and high-temporal resolution digital images collected by satellite and long-term ground-based digital camera stations.

Remote web camera deployed in Canyonlands National Park
Remote web camera deployed in Canyonlands National Park captures a regional dust event on the north-central Colorado Plateau, March 21, 2011.(Public domain.)
Dust emission from many point sources in the Little Colorado River Valley
Dust emission from many point sources in the Little Colorado River Valley on the southern Colorado Plateau. Dust is being transported towards snow cover in the Colorado Rocky Mountains. MODIS image from April 3, 2009.(Public domain.)
Image showing dust on snow throughout the Rocky Mountains, especially in the San Juan Mountains
Image showing dust on snow throughout the Rocky Mountains, especially in the San Juan Mountains. This image shows conditions after the dust layers that were deposited over the WY2009 dust season have merged together due to snow melt. The dust in this image includes the dust emitted from the southern Colorado Plateau shown in the other MODIS image. This MODIS image is from May 18, 2009.(Public domain.)

Dust from Playas and Dry Lakes

The recognition and importance of dust emission from playas, saline lakes, and dry lakes is illustrated by many studies from the western United States and around the world. Dust emission from these source areas has been shown to have far-reaching effects on (1) the degradation of groundwater, (2) climate, (3) soil and ocean fertilization, and (4) human and ecosystem health. Dust composition plays a vital role in understanding the various effects of dust on the earth system. The number of playa-dust sources will likely increase as a result of expansion of dry zones with climatic change, the conversion of saline lakes to playas, and anthropogenic activities, such as diversion of inflows (e.g. Aral Sea and Salton Sea). We investigate the conditions that promote or suppress dust emission from playa and dry lake settings focusing on interactions among geology, hydrology, and climate.

Additional information on:

Dust emission from Playas (Dry Lakes)

Dust emission from Franklin Lake playa

Accelerated deflation on Mesquite Lake playa, Mojave Desert

Wind-tunnel studies of playa surfaces

LiDAR studies of wind erosion at Mesquite lake playa (California-Nevada)

 

Aeolian Response to Climate Variability

Aeolian processes are very responsive to climate change in dry, semiarid regions. In the Mojave Desert, the Drylands Project (link to project main page here) has documented aeolian activity on time scales of decades, millennia, and hundreds of thousands of years. Aeolian processes of deflation, sand sheet formation, and aeolian erosion of yardangs during the past century have been documented on Mesquite playa, located 55km southwest of Las Vegas (please see Whitney et al, 2014). A strong correlation was found between the aeolian activity and the decadal droughts related to the Pacific Decadal Oscillation.

The Drylands Project is currently studying:

1)     Aeolian deposits that originated from two Mojave playas. These aeolian sands and silts have been dated by luminescence techniques and provide a late Holocene chronology of aeolian activity in the northern Mojave Desert.

 2)   Sand ramps — the largest aeolian landforms in the Mojave. These deposits represent unusually long records of strong aeolian episodes and contain paleosols related to regional climate changes. Sand ramps can contain terrestrial climate records back over 700,000 years, which can be integrated with paleo-lacustrine and speleothem records in the southwestern United States.

3)   Late Pleistocene to Late Holocene loess history on the Eastern Colorado Plateau that builds upon previous work by including loess deposits in Mesa Verde National park and soil development processes on the Colorado Plateau. Additional information: Reheis, M., Goldstein, H., Reynolds, R., Forman, S., Mahan, S., and Carrara, P., 2017, Late Quaternary loess and soils on uplands in the Canyonalnds and Mesa Verde area, Utah and Colorado, Quaternary research, in press.

 

Desert Stream Dynamics

Recent Flooding and Environmental Change in Las Vegas Wash

Studies of sediment erosion and deposition over the past 3,000 years in Las Vegas Wash, as well as studies about how human activity affects river systems in the Southwest, have yielded new information on the responses of river basins to climate change and urban growth. In Las Vegas Valley, streamflow and precipitation data over the past decade show that urbanization has changed rainfall-runoff relations to produce more frequent flooding as well as more damage to private property and to the primary riparian ecosystem in the Valley.

During the past 20 years, several factors related to rapid urbanization of Las Vegas and surrounding areas have led to extensive changes in Las Vegas Wash, the main drainage from Las Vegas Valley into Lake Mead. The most damaging change is extensive erosion that in places has caused downcutting of the Wash by as much as 30 meters and lateral expansion of the channel by as much as several hundred meters. Most erosion has occurred during floods caused by monsoonal, summertime downpours. The flooding not only damages marshes, but can be so severe that it endangers the main water pipeline into the city from Lake Mead. USGS studies are documenting the erosional effects from floods of the past decade. This information can be used to predict future changes to the actively eroding channel system, taken in light of projections for continued urban growth.

Additional Information:

Whitney, J.W., Glancy, P.A., Buckingham, S.E., & Ehrenberg, A.C., 2015, Effects of rapid urbanization on streamflow, erosion, and sedimentation in a desert stream in the American Southwest: Anthropocene, v.10, p. 29-42. [doi:10.1016/j.ancene.2015.09.002]

Preliminary findings describing the properties of dust deposited to Colorado snow cover during Water Year 2013

Properties of dust deposited on Colorado mountain snow cover: effects on snow albedo

Harland L. Goldstein1, Richard L. Reynolds1, Chris Landry2 - 1U.S. Geological Survey, Denver, Colorado 80225 hgoldstein@usgs.gov; rreynolds@usgs.gov2Center for Snow and Avalanche Studies, Silverton, Colorado 81433 clandry@snowstudies.org

Samples of dust deposited on snow cover (DOS) are collected from 11 Colorado Dust on Snow program (CODOS) sites throughout Colorado (Figure 1) and are sent to the U.S. Geological Survey in Denver, CO for dust mass loading (at Senator Beck Basin - Swamp Angel Study Plot [SASP] only) and dust property determinations. Samples are collected near the end of spring prior to complete snow melt, when most dust layers have merged into one layer (all-layers-merged; ALM). Dust samples are analyzed to understand the properties of dust that affect snow albedo and to link DOS to dust-source areas. These properties are (1) reflectance as a function of the solar radiation spectrum, (2) mineral and chemical compositions, and (3) particle sizes.

Figure 1. Location map of the 11 CODOS sampling sites.
Figure 1. Location map of the 11 CODOS sampling sites.

Descriptions of analytical methods are in our publications (Goldstein and others, 2013; Reynolds and others, 2013) and are summarized at the end of this report. Briefly, we determine the reflectance properties, use magnetic measurements to determine the presence of, and estimate the amounts of, "dark" minerals and ferric oxide minerals, and measure the concentration of organic carbon, some of which resides in black carbon. We compare the amounts of these dust components to laboratory measurements of reflectance to assess the influence of these dust components on albedo.

Reflectance

Reflectance values of the ALM samples fall within a narrow range relative to our global data set which includes dust and dust-source sediments from the western U.S., Australia, and central and southern Africa (Figure 2). Total reflectance values of the ALM samples fall within the middle of the global total reflectance value range, whereas the visible reflectance values are among the lowest of the global values. Closer inspection of the ALM reflectance values reveals small differences in ranges and variability; CODOS WY13 ALM dust layers are less variable than the SASP multi-water year ALM samples. Despite these small differences, the overall similarity as compared to the global suite of reflectance data suggests that all ALM samples have a similar dust-particle effect on snow albedo.

Range of reflectance values
Figure 2. Range of reflectance values. SASP ALM and CODOS WY13 ALM samples have a relatively small range of reflectance values compared to global dust and dust-source area sediment. Total reflectance values fall within the middle of the range of global values, whereas visible reflectance values fall within the lower part of the global values. Reflectance values for CODOS WY13 ALM samples are less variable than SASP ALM samples from all WYs.

Dust Components

Dust components consist of many different materials, mostly particles of minerals and also organic matter (Figure 3). These components have varying capacities to absorb solar radiation. In the example illustrated in Figure 3, average total reflectance values (Rtot) are 32% for the untreated sample, 44% for the mineral fraction, and 29% for the organic fraction. Furthermore, from literature and our ongoing research, we recognize that the most important dust-component particles for accelerating snow melt are dark minerals, "black carbon", and certain iron oxide minerals. These dust components are present in all ALM samples and can vary in amounts from layer-to-layer during a water year (WY), from year-to-year at the same site, and within a WY from site-to-site across the CODOS sampling area. These components also have a range of sizes, shapes, colors, and other properties that control the effects of dust on snow-cover albedo. Different kinds of particles may reflect, scatter, or absorb solar radiation. Our studies focus on the particles that absorb solar radiation because these particles exert the strongest influence on albedo.

Mineral and organic matter components of SASP WY11 ALM dust
Figure 3. Mineral and organic matter components of SASP WY11 ALM dust. Untreated dust sample (A; Rtot; 32%). Mineral fraction (B; Rtot; 44%) and organic fraction (C; Rtot; 29%) after density-based separation. The color of the mineral fraction is due to the iron oxide minerals, goethite and hematite. Organic fraction contains some mineral matter due to incomplete separation; however, the dark organic matter is prominent.

Mineralogy

Dust mineralogy of the CODOS WY13 ALM dust is dominated by quartz (27-44%), total clays (illite, smectite, kaolinite; 21-36%), and feldspar (16-34%). Other minerals include carbonates (calcite, dolomite; 1-6%), and mica (muscovite, biotite; 2-12%). Reflectance spectroscopy identified goethite (FeO(OH)) as the dominant iron mineral in the 11 WY13 CODOS ALM samples. Iron oxides commonly occur as nanometer-sized domains on clay minerals and on clay-coated quartz grains as observed using Scanning Electron Microscopy (SEM; Figure 4).

SEM image showing one example of the size and occurrence of iron oxide minerals in dust
Figure 4. SEM image showing one example of the size and occurrence of iron oxide minerals in dust. This image shows a quartz grain coated with illitic and smectitic clays that was deposited at SASP during the D6 event in WY13 (4/8/2013). Bright spots are nanometer-sized iron oxide and titanium oxide domains. This type of mineral particle is common in dust deposited to SASP.

Dark Minerals

Many different types of dark mineral particles act as tiny "black boxes" that absorb solar radiation. Relatively high dark mineral content (inferred from the magnetic property of isothermal remanent magnetization (IRM-0.3T)) in ALM dust is strongly associated with relatively low total reflectance (increased absorption; Figure 5). One exception is low reflectance in the SASP WY11 ALM samples that appears to be largely controlled by the high abundance of carbonaceous matter as indicated by organic carbon concentration (see Figure 6).

Reflectance as a function of dark mineral matter
Figure 5. Reflectance as a function of dark mineral matter. Total reflectance values of the CODOS WY13 ALM and the SASP ALM samples (excluding the SASP WY11 outlier) are sensitive to dark mineral matter content (IRM). IRM-0.3T is a measure of magnetite content and is used as a proxy for total amounts of dark minerals.

Organic Carbon

We estimate amounts of carbonaceous matter, such as soot from the combustion of fossil fuels, from measurement of organic carbon in samples. Generally, higher contents of organic matter correspond to lower total reflectance (Figure 6).

Reflectance as a function of organic carbon
Figure 6. Reflectance as a function of organic carbon. There is a general association between organic carbon content, especially in SASP samples, and reflectance whereby total reflectance decreases as organic carbon content increases. The SASP WY11 reflectance value may be related to certain forms of carbonaceous matter, as suggested by the high abundance of organic carbon (~5.5 %).

Iron oxide minerals

The important iron oxide minerals are hematite and goethite in which their oxidized iron imparts reddish brown-orange-yellow colors to dust layers. When on surfaces of relatively large dust particles, iron oxide minerals such as hematite and goethite (as indicated by the magnetic parameter hard isothermal remanent magnetization; HIRM), will diminish reflectance values in dust samples (Reynolds and others, 2013). However this cause-and-effect is not clearly illustrated in this data set (Figure 7) because (1) the ALM samples have small ranges in their reflectance and HIRM values, (2) there are other influences on reflectance from black carbon and dark minerals, and (3) HIRM measurements are sensitive to the size and occurrence of iron oxide minerals. Nonetheless, iron oxide minerals strongly absorb solar radiation and have a large effect on reflectance and snow melt.

Visible reflectance and iron oxide content (HIRM)
Figure 7. Visible reflectance and iron oxide content (HIRM) of the ALM dust lack a strong correlation due to the small range of values, influence from other dust components, and sensitivityof HIRM measurements to sizes and occurances of iron oxide minerals.

Particle-Size Distribution

Particle-size distributions are similar among the WY13 ALM samples from eight of the eleven CODOS sites (separated into groupings of southwestern and northeastern sites; Figure 8). Three sites, Wolf Creek Pass, Spring Creek Pass, and Park Cone, are different than the other eight sites, but are similar to each other; exhibiting overall coarser particle-size characteristics. The particle-size distributions of the WY13 ALM samples from the northeastern and southwestern sites have median particle sizes of 18-19 micrometers, and about 90% of each sample is composed of particles that are silt sized and smaller (less than 63 micrometers; PM63; Table 1). The WY13 ALM samples from Wolf Creek Pass, Spring Creek Pass, and Park Cone have coarser median particle sizes and lower PM63, PM20, PM10, and PM2.5 than the other sites.

Wind-blown particles are typically sorted from source to depositional setting, along the transport path, with respect to size and density. When considering a single dust source, the deposited particles are generally coarser at sites closest to the source and become finer with increasing distance. The similarity in particle-size distributions among ALM samples from most CODOS sites, and the coarser (and similar) particle size from three of the sites are an unexpected and interesting result, and further investigations are warranted into relations between particle sizes and dust-source areas and the effects of particle size on reflectance.

Particle-size distribution of WY13 ALM samples
Figure 8. Particle-size distribution of WY13 ALM samples from all CODOS sites showing similar distributions for eight of the eleven sites (northeast and southwest sites). Wolf Creek, Spring Creek and Park Cone have different particle-size distributions than the other eight sites, but are similar to each other. CODOS sites were separated into two general groups, southwest and northeast, to display similarities in particle-size distributions across most sites. Southwest sites include SASP, Grand Mesa, and McClure Pass and the northeast sites include Hoosier Pass, Berthoud Pass, Rabbit Ears Pass, Grizzly Peak and Willow Creek Pass.
Particle-size results for the WY13 ALM from all CODOS sites
Table 1. Particle-size results for the WY13 ALM from all CODOS sites. Particulate matter (PM) values represent percentages under the specified grain size (e.g. 9.53% of the WY13 ALM samples from the northeastern sites is composed of particles 2.5 micrometers and smaller). The results illustrate the similarities among the northeastern and southwestern sites and the coarser particles sizes found in the ALM samples from Wolf Creek, Spring Creek, and Park Cone.

Sources of Dust

There are many sources of dust across the western U.S. Dust and dust-source areas can be tracked using remote sensing and particle-trajectory analyses and can also be linked by physical, chemical and mineralogical properties. Discussions regarding specific sources of dust are beyond the scope of this report; however, dust contributions to the CODOS sites from Four Corners region/southern Colorado Plateau (FC/sCP) dust-source sediments and locally derived materials are evident from SEM analyses. ALM samples from SASP and two CODOS sites (Grizzly Peak/Loveland Pass and Park Cone) were examined under SEM to identify dust-particle compositions and to loosely discriminate among likely sources. SEM revealed that (1) the majority of particles in the SASP ALM samples are derived largely from FC/sCP; and (2) the two CODOS sites contain mostly locally derived material with lesser amounts of particles that could possibly be from FC/sCP. Although certain types of particles are common in the FC/sCP dust-source region, they are not specific to that region alone. Therefore, their presence in deposited dust, particularly at sites further from the FC/sCP source region, is not completely diagnostic. Nevertheless, the combination of ground conditions, wind patterns, and direct observations indicates that FC/sCP dust sources are capable of contributing dust that is deposited at many of the CODOS sites. As one example, the large dust event D8WY13 observed at SASP from 4/15/2013 through 4/17/2013 originated in the Four Corners region (Figure 9). Dust fall was also observed across all CODOS sites during this time period. Particle-trajectory analysis demonstrates that dust emitted from the Four Corners region during the D8WY13 likely tracked across the CODOS sites, reaching the northern most sites within 12 hours (Figure 10).

MODIS satellite image showing D8WY13 dust event originating in the Four Corners region
Figure 9. MODIS satellite image showing D8WY13 dust event originating in the Four Corners region. Red outline denotes extent of major dust emission visible on this image. Circles are CODOS sites as labeled in Figure 1.
MODIS satellite image as in Figure 2, showing D8-WY13 dust event
Figure 10. MODIS satellite image as in Figure 2, showing D8-WY13 dust event. Colored lines represent particle trajectories modeled using HYSPLIT (www.arl.noaa.gov/HYSPLIT_info.php) at various altitudes for 12 hours. Black - 100 meters above ground level (MAGL); red - 200 MAGL; blue - 500 MAGL. Dust emitted in the Four Corners region during this event appears to have traveled northeast across the CODOS sites within a 12 hour period.

Summary

Dust on mountain snow cover changes snow albedo and enhances the absorption of solar radiation, thereby increasing rates of snow melt and leading to earlier-than-normal spring runoff and overall smaller late-season water supplies for tens of millions of people and industries in the American West (Deems and others, 2013; Skiles and others, 2012; Painter and others, 2007; 2010). These effects are largely due to the optical properties of dust particles deposited to snow cover. On the basis of studies of radiative properties of deposited dust, we document absorption of solar heat by dark minerals, black carbon, and certain iron oxide minerals. Further study of the observed variations in dust component amounts, including particle sizes, and their variable influence on reflectance will help us understand why dust has such large effects on accelerating snow melt and will be useful for modeling the range of dust snow-melt effects to be seen in the future.

Our observations of variations in dust components and particle sizes are also leading to new ways to determine sources of the dust by comparing the variable compositions of DOS-layer compositions with different compositions of fine-grained sediments and soils in dust-source areas. Deposited dust to Colorado snow pack can be generally linked to FC/sCP sources as well as contributions from locally derived material. The examination of dust from additional CODOS sites using SEM coupled with future isotopic analyses will help to discern the spatial extent of contributions from various dust sources, including the FC/sCP region. Furthermore, individual layers of deposited dust derived from known Four Corners dust events could be collected across the CODOS sites and analyzed to determine dust-source contributions to Colorado mountain snow cover. In summary, study of the physical, chemical and mineralogical compositions of ALM dust over more years is needed to (1) improve understanding and modeling of the effects of dust on snow-cover albedo and (2) discern clues about locations and behavior of the dominant dust sources. Understanding the dust properties that affect snow albedo and developing the ability to link deposited dust to dust-source areas may guide mitigation of dust emission that affects Colorado River water resources.

Methods

For detailed description of methodologies see Goldstein and others (2013) and Reynolds and others (2013).

Reflectance

Reflectance spectra of the dust samples are measured using bi-directional methods relative to a 100% reflective white reference panel to determine average reflectance values. Any deviations from the white reference panel reflects absorption by the dust sample and results in reflectance values less than 100%. For example, a reflectance value of 45% corresponds to an absorption of 65%. Reflectance values are reported as average total reflectance and average visible reflectance. Average total reflectance is the average reflectance over the entire measured wavelength range of 0.35 to 2.50 micrometers. Average visible reflectance is the average reflectance over the visible wavelength range of 0.40 to 0.70 micrometers.

Mineralogy

Semi-quantitative bulk mineralogy abundances were made using a powder X-ray diffraction (XRD) technique that is sensitive to abundances greater than about 3-5 percent. Because of the low abundances, magnetite, hematite, and goethite are typically not detectable in these samples using XRD. Iron mineralogy was identified from reflectance spectroscopy measurements where minerals are best fit to absorption features.

Magnetic Properties

Magnetic properties of isothermal remanent magnetization (IRM-0.3T) and hard isothermal remanent magnetization (HIRM) were measured to infer dark mineral and iron oxide mineral contents, respectively. Isothermal remanent magnetization measurements were made using a high-speed spinner magnetometer after IRM was generated at room temperature in an impulse magnetizer. First, IRM was imparted in a 1.2-tesla (T) induction (IRM1.2T). The samples then were magnetized in the opposite direction by using an induction of 0.3 T (IRM-0.3T).

Hard isothermal remanent magnetization (HIRM) was calculated as HIRM = (IRM1.2T + IRM-0.3T)/2 (King and Channel, 1991). The HIRM parameter is a measure of the amount of ferric oxide minerals, when these minerals are magnetically ordered at room temperature.

Organic Carbon

Organic carbon concentrations were determined by combustion under vacuum in an oxygen-rich environment at 980°C using a Costech 4010 CHNSO elemental analyzer after treatment with 15% HCl to remove carbonate.

Acknowledgments

We thank Kim Buck and Andrew Temple for administrative and field assistance, Eric Fisher for particle-size measurements, Jiang Xiao for reflectance spectroscopy and magnetic property measurements, Gary Skipp for performing X-ray diffraction, Ray Kokaly for data reduction and quality control of the reflectance spectroscopy measurements, Dan Fernandez for carbon content determinations, and George Breit for particle characterizations using the scanning electron microscope.

References

Deems, J.S., Painter,T.H., Barsugli,J.J., Belnap,J., and Udall,B.: Combined impacts of current and future dust deposition and regional warming on Colorado River Basin snow dynamics and hydrology, Hydrol. Earth Syst. Sci., 17, 4401-4413, doi:10.5194/hess-17-4401-2013, 2013.

Goldstein, H.L., Reynolds, R.L., Morman, S.A., Moskowitz, B., Kokaly, R.F., Goossens, D., Buck, B.J., Flagg, C., Till, J., Yauk, K., Berquó, T., 2013. Iron mineralogy and bioaccessibility of dust generated from soils as determined by reflectance spectroscopy and magnetic and chemical properties: Nellis Dunes Recreational Area, Nevada. U.S. Geological Survey Scientific Investigations Report 2012-5054, 15 p.

King, J.W., Channel, J.E.T., 1991. Sedimentary magnetism, environmental magnetism, and magnetostratigraphy. Reviews of Geophysics, Supplement, 358-370.

Painter, T.H., Barrett, A.P., Landry, C., Neff, J.C., Cassidy, M.P., Lawrence, C., McBride, K.E., Farmer, G.L., 2007. Impact of disturbed desert soils on duration of mountain snow cover. Geophysical Research Letters 34, L12502, doi:10.1029/2007GL030284.

Painter, T.H., Deems, J.S., Belnap J., Hamlet, A.F., Landry, C.C., Udall, B., 2010. Response of Colorado River runoff to dust radiative forcing in snow. Proceedings of the National Academy of Sciences 107 (40), 17125-17130; doi: 10.1073/pnas.0913139107.

Reynolds, R.L., Goldstein, H.L., Moskowitz, B.M., Bryant, A.C., Skiles, S.M., Kokaly, R.F., Flagg, C.B., Yauk, K., Berquó, T., Breit, G., Ketterer, M., Fernandez, D., Miller, M.E., Painter, T.H., 2013, Composition of dust deposited to snow cover in the Wasatch Range (Utah, USA): Controls on radiative properties of snow cover and comparison to dust-source sediments, Aeolian Research (http://dx.doi.org/10.1016/j.aeolia.2013.08.001).

Skiles, S.M., Painter, T.H., Deems, J.M., Bryant, A.C., Landry, C.C., 2012. Dust radiative forcing in snow of the Upper Colorado River Basin: 2. Interannual variability in radiative forcing and snowmelt rates. Water Resources Research 48, W07522, doi:10.1029/2012WR011986.