Dust is known to aggravate respiratory diseases. This is an issue in the desert southwestern United States, where windblown dust events are common. The Public Health Applications in Remote Sensing (PHAiRS) project aims to address this problem by using remote-sensing products to assist in public health decision support. As part of PHAiRS, a model for simulating desert dust cycles, the Dust Regional Atmospheric Modeling (DREAM) system is employed to forecast dust events in the southwestern US. Thus far, DREAM has been validated in the southwestern US only in the lower part of the atmosphere by comparison with measurement and analysis products from surface synoptic, surface Meteorological Aerodrome Report (METAR), and upper-air radiosonde. This study examines the validity of the DREAM algorithm dust load prediction in the desert southwestern United States by comparison with satellite-based MODIS level 2 and MODIS Deep Blue aerosol products, and ground-based observations from the AERONET network of sunphotometers. Results indicate that there are difficulties obtaining MODIS L2 aerosol optical thickness (AOT) data in the desert southwest due to low AOT algorithm performance over areas with high surface reflectances. MODIS Deep Blue aerosol products show improvement, but the temporal and vertical resolution of MODIS data limit its utility for DREAM evaluation. AERONET AOT data show low correlation to DREAM dust load predictions. The potential contribution of space- or ground-based lidar to the PHAiRS project is also examined.