Increasing springtime dust activity reported over the southwestern United States is a risk factor for visibility, air quality, human health, and water resources. The impact on water resources is primarily through a possible shift in the Colorado River basin’s hydrological cycle due to early springtime snowmelt from dust deposition on mountain snowpack. These risks are likely exacerbated with landcover change across the southwest due to climate change and land use practices. Understanding spatiotemporal trends in indicators of surface roughness (e.g., us*/Uh) and the wind friction velocity at the soil surface (us*) over the different land cover types across the southwest is a first and critical step for understanding the controls on aeolian processes and downwind dust deposition. We apply a recently developed albedo-based drag partition scheme to calculate us*/Uh and us*. We use the 20 year (2001-2020) Moderate Resolution Imaging Spectroradiometer daily albedo (500 m pixel) to calculate us*/Uh. The us*/Uh are multiplied by wind speed (at 10 m height) from GridMet (Gridded Surface Meteorological Dataset: daily at 4 km pixels) to provide us* for the spring months (March-May)-the season of greatest dust activity. These data are harmonized within the Google Earth Engine to provide daily maxima at 500 m and examine how surface roughness (primarily due to vegetation) has changed over the past two decades across the southwest and its relation with increasing dust activity. We performed decadal and trend analyses using the Mann-Kendall trend test which identified spatiotemporal patterns of us*/Uh and us* across the study area within different ecoregions and land cover types. The results enabled the identification and characterization of springtime dust hotspots over the southwestern United States during the last two decades. The dust hotspot trend analysis provided insights needed to predict the future state of the dust activity and its relation with human health, dust on snow, and the hydrological cycle in the Colorado River basin. Importantly, this study also informs how the albedo-based drag partition scheme can improve regional and global dust emission modeling sensitivity to subgrid-scale land cover dynamics.