Collaborative Research: NSFGEO-NERC: Aeolian dust responses to regional ecosystem change

Lead Research Organisation: Cardiff University
Department Name: Sch of Earth and Environmental Sciences

Abstract

Overview: The proposed research will establish dust emission responses to regional ecosystem change across North America due to three disturbance regimes: 1) numerous discrete, small scale (<10 ha), abrupt ecosystem changes associated with energy (oil and gas) development; 2) discrete, large scale (>1000 ha), abrupt ecosystem change associated with fire and ecological feedbacks that promote invasive grasses and altered fire-wind erosion regimes; and 3) diffuse, large scale (>1000 ha), pervasive ecosystem changes associated with sustained shrub invasion of desert grasslands. New dust emission models driven by remote-sensing and broad-scale standardized ecological datasets will be used to resolve the mechanics of soil and vegetation-aeolian transport interactions. The models will be calibrated using data from National Wind Erosion Research Network observatory sites, then applied to explore patterns of dust emission across the study regions. National ecological datasets will be used to analyze the drivers and interpret dust emission responses to regional ecosystem changes across the disturbance regimes.

Intellectual Merit: This proposal examines the geomorphic mechanisms and impacts of regional ecosystem change (land use and land cover change) on the magnitude of North American dust emission. Dust originating from North American deserts is regionally important for its profound impacts on human health and transportation systems, water resources, agricultural production, and its feedbacks to biogeochemical cycling and climate. Dust source regions are experiencing changes in ecosystem structure and function due to drought, land use and management pressures, and climate change. Plot-scale research has explored the interactions between vegetation change and aeolian processes in drylands. However, the significance of ecosystem changes for current and future regional dust emissions has not been established over large areas. This research will transform our understanding of how different types of ecosystem change due to human land-use pressures and natural disturbances impact regional dust emissions. Findings will significantly improve our understanding of the coupling between the dust cycle and anthropogenic drivers of ecosystem change.

Planned Impact

Broader Impacts: Wind erosion of global drylands is a key land degradation process, impacting food security, human health, economic and political stability, and adaptation to climate change. This research will provide information critical for managing aeolian process interactions and mitigating dust impacts globally. Public and private land managers and agencies will benefit from the information produced to assess co-benefits and trade-offs for dust mitigation options. The Investigators' existing relationships with the Bureau of Land Management, National Park Service, and Natural Resources Conservation Service will ensure immediate transfer of knowledge gained. Research results will be disseminated through workshops with stakeholders, the National Wind Erosion Research Network website, and traditional academic venues. In addition, multiple educational activities are included, such as training of underrepresented graduate and undergraduate students, and a K-12 education program.
 
Description ********************Feb 2022*******************************

Following on from the key findings during the previous period, the project has generated multiple research articles at different phases of publication. Our research on North American dust emission, using the albedo-based dust emission model (AEM) has been published in Aeolian Research. https://doi.org/10.1016/j.aeolia.2021.100766

This article including many of the finding covered in the previous reporting period, including the novel application of satellite-derived dust point source (DPS) observations for calibration of modelled dust emission magnitude and frequency. In addition, subsequent key findings were communicated in this article, including the identification of two distinct climate and land surface regimes, which in isolation generate similar quantities of dust emission. In the arid southwest, surface roughness is relatively smooth most of the time, with dust emission reacting to local and seasonal peaks in wind speed (wind limited system). In the Great Plains, wind speeds are large most of the time, with dust emission occurring only where and when vegetation roughness reduces sufficiently (roughness limited system). These key insights highlight the potential impact of land use practices or alterations to climate (changing wind/precipitation patterns) on the dust emission potential in each region. In a further article, we assessed the weaknesses in dust emission modelling, and the potential improvements using the AEM compared to tradition dust emission modelling (TEM) across North America. We highlight the benefits of using DPS for calibration against traditional approaches, specifically measurements of dust-aerosol optical depth (DOD). Here, we firstly demonstrated that DOD frequency (DOD > a given threshold) shows no correlation with observed DPS frequency. Indicating a limitation of using transported atmospheric dust (i.e., DOD) as an indicator/calibration of dust emission. AEM frequencies were generally closer (RMSE = 0.6) to observations than TEM frequencies (RMSE = 0.76). However, model performance is most distinct beyond DPS regions (i.e., across a larger range of land cover and climatic conditions). The AEM model retains dynamic information on vegetation roughness, measured variance along a continuum of conditions. Here, dust emission is constrained by wind friction velocity at the eroding bed, generally in areas of limited vegetation roughness. The TEM has a static representation of roughness, where a fixed roughness length is applied homogenously within land cover classes. Consequently, TEM estimates larger dust emission across greater extents of North America, in areas less likely to produce dust. These findings have been submitted to Geoscientific Model Development, where it is at an advanced stage of review - public and peer-review complete, with comments responded too; a decision on publication is due from the editor in the coming weeks. https://doi.org/10.5194/gmd-2021-337

Our work on Dichotomous analysis, whereby we determine AEM performance against a global dataset of >37k unique DPS observed locations has also been submitted to Geoscientific Model Development (currently receiving public and peer review). https://doi.org/10.5194/gmd-2021-423 Performance is measured through the coincidence of dichotomous outcomes, specifically dust and no-dust events. Our key findings include, for the first time a comparison of dust emission probability, described by the collation of published DPS frequency data from each of the key dust producing regions. We identify the apparent rarity of dust emission, with regions thought to generate frequent dust emission (e.g., North Africa, Middle East). that AEM simulations coincided with dichotomous observations ~71% of the time but simulated dust emission ~27% of the time when no dust emission was observed. Our analysis indicates that the widely accepted dust entrainment thresholds (u*ts =0.2 m s-1) was typically too small, needed to vary over space and time, and at the grain-scale is incompatible with the wind friction velocities (us*) described at the modelled scale (MODIS 500 m). During observed dust emission, us* was too small because wind speeds were too small and / or the wind speed scale (ERA5; 11 km) is again incompatible with the MODIS 500m scale. The absence of any limit to sediment supply caused the AEM to simulate dust emission whenever P (us* > u*ts), producing many false positives when and where wind speeds were frequently large. Global dust emission estimates for AEM and TEM were compared using our novel (DPS) derived calibration (AEM) and traditional (DOD) calibration techniques, in a research article submitted to Nature Geosciences. The results show a dramatic shift in the AEM data, revealing a seasonally shifting dust emission predominance within and between hemispheres, not persistent North African dust emission primacy as widely interpreted from aerosol optical depth measurements. Earth's largest dust emissions progress from East Asian desert and grassland in boreal spring, to Middle Eastern desert in boreal summer and then Australian shrublands in boreal autumn-winter. Trends in dust emission over time are currently being investigated for two separate manuscripts, focused on North American and Global perspectives. In North America, vegetation and wind dynamics have caused alternating trends in dust emission magnitude between two 10-year periods since 2001. Regionally, dust emission over the Colorado Plateau has reacted predominantly to changes in wind speed intensity, while the Great Plains show a significant increase in dust emission eastward, as vegetation roughness has reduced significantly (p>0.05) and wind speeds have stilled. Overall, most dryland areas have shown an increase in vegetation roughness, indicative of global greening seen in other global dryland areas over the past 2 decades. Globally, dust emission trends appear to react to annual changes in snowfall extent, with years of increased snowfall reducing the annual probability of dust emission in cold deserts areas.

**********************March 2021********************************

Despite restrictions inflicted through the COVID-19 pandemic (see COVID), this on-going project has successfully made several key findings.

1) DEM calibration using DPS data. The main challenge in calibrating or verifying the skill of DEMs is acquiring significant dust emission observation data. Where dust emission typically occurs in remote, inhospitable locations observations rely on either spatially distant ground-station data, or regular satellite observations of aerosol optical depth (AOD). Each of these data inherit limitations, predominantly linked to the lack of spatial resolution or frequency of observation. Importantly, neither of these sources of observation pertain to actual observations of emission, with both capturing dust emission in later phases of transport (AOD) or deposition (ground station). Therefore, reliance on either will inevitable create spatial and temporal bias during calibration and subsequent forecast. We collated multiple datasets describing dust emission frequency, observed through manual identification of EO data. Dust emission from each of the dust point source (DPS) observations are forecast for the duration of the observation study. Calibration is determined through a linear regression model of the residual difference between dust emission forecasted during DPS observation , and on all - For the first time, dust emission forecasts were calibrated using the frequency of dust emission observations, acquired from subjective identification in remote sensing data. This novel calibration technique, utilising observed dust emission frequency data to correct inaccuracies in existing wind friction velocity (wind speed) thresholds, generating improved accuracy in forecast magnitude.

2) Dust emission modelling of North America. In North America, soil organic carbon and nutrient losses caused by wind erosion and particularly dust emission reduce soil fertility, clay content, moisture holding capacity and directly impact crop yield. These losses in soil quality may be hidden from crop yield by fertiliser applications but which causes increasingly expensive crop production. Not tackling the cause of nutrient losses increase the dispersion of nutrients across ecosystems causing pollution which may have consequences for marine eutrophication, biodiversity and environmental protection of marine food resources. Using the novel albedo-dust emission model (DEM) (Chappell and Webb 2016), we successfully produced a dust emission map for North America, including a 20-year climatology of emission frequency and magnitude. Unlike traditional models, the albedo DEM incorporates dynamic surface roughness to describe the amount of surface sheltering, due primarily to vegetation structure, allowing an accurate estimation of wind friction velocity at the eroding surface once attenuated by local aerodynamic roughness. Our map highlights areas of intense dust emission and has linked the relative frequency and magnitude of emission to specific characteristics of vegetation cover, and seasonal wind dynamics. These data reveal a spatial heterogeneity in dust emission, identifying a narrow corridor of increased dust emission from the eastern extent of the Great Plains, to the northern reaches of the Chihuahua Desert. These areas present ideal condition, with reduced aerodynamic roughness from the relative lack of vegetation, and increased winds following the Chinook winds (a type of foehn wind) descending the Central Rocky Mountains. These results demonstrated the impact of vegetation roughness on the spatial extent of dust emission. Using spatial analysis, we identified regions with similar wind regimes, which do not produce dust largely due to ecological differences, including increased surface roughness caused by variations in vegetation type and coverage. These finding highlight the dynamic interaction between land surface condition and dust emission, providing insight into potential impact of 21st century climate change on North American soil erosion.

3) Dichotomous analysis: Previously, DEM performance focussed on the ability to recreate volumetric observations of atmospheric dust. These observations typically occurred as a spatial and temporal aggregates from either point source locations (ground network data) or aerial AOD (satellite observations). Accordingly, as modelled and observed data measure two separate processes (emission and transport respectively) these types of observations are unsuitable for DEM performance assessment. Furthermore, ambiguity remains in how these observed volumes of dust arrived over a specified region. Indeed, dust can be transported over thousands of kilometres; likewise, DEMs arrive at a given approximation through multiple complex process-based equations, each with inherent compromises. It therefore is probable that any coincidence in approximation occurs due to chance, and not to the reliability of the model. By using dust point source (DPS) data, we are able to demonstrate a direct comparison of a DEM's performance with actual emission observation. A key advantage of using DPS frequency data, is the abilty to simplify the results through binary outcomes (Dust / no Dust). Dichotomous tests are used where the forecast and observation variable contain a maximum of two distinct outcomes. This categorical verification is used in weather forecasting, typically for specific meteorological events (e.g., tornado, rain or snow), where the verification question is "Did/Will this event occur"? In each instance, observation and forecast will provide a binary response, (i.e., 1 = Yes it will/did occur, 0 = No it did not / will not occur), these responses can be compared in a contingency table, where the responses are categorised as either Hit (Observation and Forecast = 1), Miss (Observation = 1 / Forecast = 0), False Positive (Observation = 0, Forecast = 1) or a Correct Negative (Observation and Forecast = 0). The results from these contingency tables provide the data for a number of categorical statistics, which describe the conditional performance of the model. Choice of statistics is determined by the purpose of the forecast, and the measure of 'goodness' within that purpose. As an example, accuracy score can be bias towards either the number of correct negatives or number of hits. If the purpose of the model was to determine a rare event (e.g., tornado/dust storm), a high accuracy score could be maintained even if no events were accurately forecast, and the number of correct negatives were proportionally high against the total number forecasts. Therefore, if the purpose of the model was to predict the occurrence of dust storms (for mitigation purposes), then the accuracy score only describes a fraction of the performance and does not fully define the goodness of the model. Combining multiple dichotomous statistics (e.g., accuracy + bias) provides a powerful metric for dissecting a model's ability to perform during scenarios where an event occurs and does not. Using these analyses, we identified significant bias in DEMs with traditional methods producing a dust emission frequency 3 orders of magnitude greater than observation. Using the albedo DEM, that bias was reduced to 1 order of magnitude. The enduring bias highlights the need for dynamic wind shear velocity thresholds, which vary spatially and temporal in reaction to changes in erodibility of the soil.

****************March 2020************************

The funding was awarded in early 2019 but a postdoc was not in place to start work until Nov, 2019 and so the project is in its early stages, on-going and has achieved the following:

(1) Identified indicators of ecosystem structure and function that can be used to assess disturbance impacts on vegetation communities and dust emissions across the three study ecoregions. Established approaches for identifying benchmarks to assess disturbance impacts and change in ecosystem structure and function and dust emission. Arising publication Webb et al. (2020). The work revealed benefits and limitations of different indicators of vegetation type, cover, structure (e.g., height) and spatial distribution for assessing dust emission responses to land uses and disturbances that drive regional ecosystem change. The review also addressed different approaches for establishing benchmarks for the indicators with a focus on indictors that can be used to describe the status, condition and trend of ecosystems as well as their dust emission potential (either directly or through physics-based models). We established how classification systems describing land potential (e.g., ecological site descriptions used by US land management agencies) can be used to inform benchmarks that can help both our research and land management agencies assess whether locations across landscapes are at risk of exceeding structural and/or functional thresholds of concern that may put them at risk of degradation due to disturbance processes and increase dust emissions.

(2) Established a remote-sensing driven regional dust model and determined the dust model sensitivity to land cover change associated with ecosystem disturbances and change relative to existing/available dust models. We conducted an in-depth analysis of the application of drag partition schemes in existing regional dust models and compared the sensitivity of the most commonly applied regional dust model (Marticorena and Bergametti, 1995) with a new remote sensing (albedo) based model established by Chappell and Webb (2016). Our evaluation of drag partition schemes (Webb et al., 2020) revealed that implementation of drag partition schemes in traditional dust models over the last 20 years has been incomplete, producing large (up to three orders of magnitude) hidden uncertainties in the magnitude of modelled dust emissions. Traditional dust models are largely insensitive to the impacts of land cover change (due to incomplete implementation of drag partition and lack of surface roughness dynamics). The Chappell and Webb (2016) model is sensitive to surface roughness dynamics and predicts measured dust emissions across the US within one order of magnitude. Comparison of albedo model estimates against dust concentration measurements at 109 Interagency Monitoring of Protected Visual Environments (IMPROVE) sites across the US showed that the albedo model is able to reproduce spatial and temporal patterns of dust emission observed in network measurements. An arising manuscript is under review in Nature Communications.

There are two significant emerging research questions from this preliminary work:

i) has the impaired nature of the traditional dust emission models developed a false impression of dust sources and their contribution and if so do those impression extend from regional sources to global predominance?

ii) the traditional dust emission models are coupled to earth system models to quantify dust-climate and other pervasive impacts of dust in earth systems, so does the impaired nature of the traditional dust emission models undermine current climate projections?
Exploitation Route *******************Feb 2022 ************************
To date, the achievements of this project include the publication of works in Aeolian Research, and the submission of three other articles to Geoscientific Model Development and Nature Geosciences. Furthermore, these analyses and associated data lay the foundation for future research, both in dust emission patterns around the world, and changing dynamics of vegetation roughness and wind patterns. The former directly effects how dust emission is recognized in general circulation models (GCMs), enabling us to apply dynamic changes to vegetation patterns and simulate how changing climate (precipitation, wind) may impact dust emission potential of the soils. For the latter, observed changes in vegetation roughness patterns around the globe point towards global greening, where previously spectral analyses of the Earth's surface have revealed a net increase in photosynthetic activity. These new data highlight a change in the physical structure of global vegetation, providing greater insight into dynamic reactions to a changing climate (e.g., species invasion / migration) and its impact on surface wind patterns through aerodynamic drag properties. The last two papers currently in development will begin to address these issues, providing sufficient evidence to investigate these data further. At this point, the project has only 2 months remaining before it is complete. In these two months, it is expected that research articles which have been submitted, but not yet published will complete the review phase and will be made available as Open Access articles. During this period, two further manuscripts will be submitted for review, with publication expected in the months following the completion of the project. Results from the Dichotomous analysis will be presented at the EGU 2022 conference in Vienna (late May), following on from a successful presentation of the North American dust emission climatology at EGU 2020.


*******************March 2021 ************************

The achievements of project so far are described by how they will be received by interested stakeholders below.

1) Novel calibration of a dust emission model: By using dust point source observation data to calibrate frequency of emission, we provide a calibration technique that is more appropriate to dust emission models (see next section for details). These DPS data describe the largest source of global dust emission observation data. They have all been peer-reviewed and provide the only existing method of realistically calibrating dust emission models. This calibration technique will be of significant interest to earth system modellers, as dust emission plays an important role in many earth systems and bares a large impact on the radiation budget in global circulation models (GCMs). The subsequent calibrated dust emission forecasts provide the cutting-edge in DEM, and will be used within this project, future projects and interested parties to further understand land surface processes linked to dust emission, including the erodibility of soils, spatial and temporal variability in wind speed thresholds, vegetation structure and ecological diversity.

2) Dust emission map of North America: For some time, there has existed a need for a high-resolution map of North American dust sources and associated emission climatology. Landowners and policy makers require information on soil erosion and dust emission to inform decisions on land management practices, to prevent the continual degradation of top-soil productivity (proportion of soil-organic carbon), damage to crop production (sand blasting, loss of sol nutrients), and danger to human health (respiratory illness, reduced visibility for transportation). It is expected that these potential stakeholders and relevant research groups will take particular interest in these findings, as the albedo approach provides a significant departure from existing methods, both in design of model but also in forecast output. The results are expected to generate academic discourse, as the spatial distribution of dust emission in specific locations contradicts previous findings, while the seasonal trends are expected. However, we have demonstrated through comparison with existing techniques, and verification by accepted ground station network data, that the albedo approach provides greater accuracy in forecasting spatial and temporal variability in dust emission (see below). We look forward to discussions with interested parties, to establish means to make these data accessible, to inform on policy and land management practices.

3) Model performance analysed through dichotomous statistics: We pioneer a new form of DEM analysis, using DPS data to compare the observed emission frequency with model forecast (see next section for details). We expect that many researchers in dust emission community will use these data as a new standard for DEM analysis, as these results specifically highlight our ability to model wind shear velocity and their respective thresholds. These parameters arguably pose the greatest challenges to DEM skill, and are also the areas of greatest importance to policy makers and landowners who are invested in the protection of topsoil. Regional differences in model performance highlight dust emission potential across global dryland areas, providing an avenue for comparisons of dust emission magnitude as well as frequency. These data will be further explored within this project, and affiliated publications. They will likely generate debate amongst dryland researchers and climatologists, as they hold the potential to challenge accepted understanding of global dust emission sources.

****************March 2020 ********************

The project is in the early stages and the project is still ongoing. However, the research achievements so far indicate a need to evaluate critically the extent to which previously unknown uncertainty in traditional dust emission models have influenced the magnitude and relative contributions of dust and how this has influenced current understanding of the impact of dust on earth systems.

We have produced a new global dataset of roughness dynamics which will become available for earth system modelling to improve dust emission modelling and climate interactions. The arising method also enables earth system models to develop a prognostic approach to dust emission modelling which is likely to reduce uncertainty in dust-climate interactions and may produce a different understanding of the dust impacts on future climate.
Sectors Agriculture, Food and Drink,Education,Environment
URL https://doi.org/10.1016/j.aeolia.2021.100766
 
Description The research has provided middle school students in Las Cruces, NM with the opportunity to learn about aeolian sediment transport processes, work with sediment transport data collected across the Dona Ana county, and build their own analyses and conclusions from the data by participating in the Asombro Desert Data Jam. Twenty middle school science classes participated in the Desert Data Jam in 2019.
First Year Of Impact 2019
Sector Agriculture, Food and Drink,Education,Environment
Impact Types Societal
 
Description NERC Advisory Network
Geographic Reach National 
Policy Influence Type Participation in a guidance/advisory committee
 
Title Dust emission map of North America 
Description Dust emission map of North America: For some time, there has existed a need for a high-resolution map of North American dust sources and associated emission climatology. Landowners and policy makers require information on soil erosion and dust emission to inform decisions on land management practices, to prevent the continual degradation of top-soil productivity (proportion of soil-organic carbon), damage to crop production (sand blasting, loss of sol nutrients), and danger to human health (respiratory illness, reduced visibility for transportation). 
Type Of Material Data analysis technique 
Year Produced 2021 
Provided To Others? Yes  
Impact It is expected that these potential stakeholders and relevant research groups will take particular interest in these findings, as the albedo approach provides a significant departure from existing methods, both in design of model but also in forecast output. The results are expected to generate academic discourse, as the spatial distribution of dust emission in specific locations contradicts previous findings, while the seasonal trends are expected. However, we have demonstrated through comparison with existing techniques, and verification by accepted ground station network data, that the albedo approach provides greater accuracy in forecasting spatial and temporal variability in dust emission (see below). We look forward to discussions with interested parties, to establish means to make these data accessible, to inform on policy and land management practices. The Google Earth Engine Java script code is available to run using the links below for the traditional dust emission model (TEM) and the albedo-based dust emission model (AEM). 
URL https://code.earthengine.google.com/9726348d2fc3e81381e8a9229667afdd
 
Title Global dust point source data for calibrating dust emission 
Description Novel calibration of a dust emission model: By using dust point source observation data to calibrate frequency of emission, we provide a calibration technique that is more appropriate to dust emission models (see next section for details). These DPS data describe the largest source of global dust emission observation data. They have all been peer-reviewed and provide the only existing method of realistically calibrating dust emission models. 
Type Of Material Data analysis technique 
Year Produced 2022 
Provided To Others? Yes  
Impact This calibration technique will be of significant interest to earth system modellers, as dust emission plays an important role in many earth systems and bares a large impact on the radiation budget in global circulation models (GCMs). The subsequent calibrated dust emission forecasts provide the cutting-edge in DEM, and will be used within this project, future projects and interested parties to further understand land surface processes linked to dust emission, including the erodibility of soils, spatial and temporal variability in wind speed thresholds, vegetation structure and ecological diversity. The data used are identified in the main text and below using the Google Earth Engine data description and catalogue references, link and DOI. The satellite observed dust emission point source (DPS) data are available from a Zenodo repository 
 
Title Model performance analysis 
Description Model performance analysed through dichotomous statistics: We pioneer a new form of DEM analysis, using DPS data to compare the observed emission frequency with model forecast (see next section for details). We expect that many researchers in dust emission community will use these data as a new standard for DEM analysis, as these results specifically highlight our ability to model wind shear velocity and their respective thresholds. These parameters arguably pose the greatest challenges to DEM skill, and are also the areas of greatest importance to policy makers and landowners who are invested in the protection of topsoil. Regional differences in model performance highlight dust emission potential across global dryland areas, providing an avenue for comparisons of dust emission magnitude as well as frequency. 
Type Of Material Data analysis technique 
Year Produced 2021 
Provided To Others? No  
Impact These data will be further explored within this project, and affiliated publications. They will likely generate debate amongst dryland researchers and climatologists, as they hold the potential to challenge accepted understanding of global dust emission sources. 
 
Description NMSU 
Organisation New Mexico State University
Department New Mexico State University Grants
Country United States 
Sector Academic/University 
PI Contribution Albedo-based wind erosion and dust emission modelling as part of NSF-NERC funded collaboration
Collaborator Contribution Instrumented wind erosion network measurements including the Jornada Experimental Range as part of NSF-NERC funded collaboration
Impact None so far
Start Year 2019
 
Description US Army field validation of albedo-based wind erosion and dust emission model 
Organisation Desert Research Institute
Country United States 
Sector Charity/Non Profit 
PI Contribution Advice on albedo-based wind erosion and dust emission modelling. Advice on sampling design for field experiment designed to test the albedo-based model.
Collaborator Contribution Field measurements of shear stress approximated by the albedo-based model which confirm the validity of the model.
Impact US Army funding for the field experiment and experimental equipment to validate the albedo-based wind erosion and dust emission model. Manuscript submitted to Journal of Geophysical Research.
Start Year 2019
 
Description US Army field validation of albedo-based wind erosion and dust emission model 
Organisation US Army Research Lab
Country United States 
Sector Public 
PI Contribution Advice on albedo-based wind erosion and dust emission modelling. Advice on sampling design for field experiment designed to test the albedo-based model.
Collaborator Contribution Field measurements of shear stress approximated by the albedo-based model which confirm the validity of the model.
Impact US Army funding for the field experiment and experimental equipment to validate the albedo-based wind erosion and dust emission model. Manuscript submitted to Journal of Geophysical Research.
Start Year 2019