Modelled direct causes of dust emission change (2001–2020) in southwestern USA and implications for management

Abstract

North American observed atmospheric dust has shown large variability over the last two decades, coinciding with regional patterns of vegetation and wind speed changes. Dust emission models provide the potential to explain how these direct causes of vegetation and wind speed changes are related to changing dust emission. However, those dust models which assume land cover types are homogeneous over vegetation classes and fixed over time, are unlikely to adequately represent changing aerodynamic roughness of herbaceous cover, woody cover, and litter. To overcome these model limitations and explain changing (2001–2020) dust emission, we used a new MODIS albedo-based dust emission model calibrated to satellite-observed magnitude and frequency of dust emission point source (DPS) data. We focused our work on four regions of southwestern USA, identified previously as the main dust emission sources. We classified the interplay of controlling factors (wind speed and aerodynamic roughness) which created disturbance regimes with dust emission change consistent with diverse land use and management drivers. Our calibrated model results show that dust emission is increasing or decreasing, in different regions, at different times, for different reasons, consistent with the absence of a secular change of observed atmospheric dust. Our work demonstrates that using this calibrated dust emission model, sensitive to changing vegetation structure and configuration and wind speeds, provides new insights to the contemporary factors controlling dust emission. With this same approach, the prospect is promising for modelling historical and future dust emission responses using prognostic albedo in Earth System Modelling.