GIScience has long been dominated by the naturalistic and scientistic ontological constraints championed by Willard Van Orman Quine. This has relegated dust storms and even mountains to the so-called slum of possibles that Quine had sought to clear in favor of a more aesthetic desert landscape. Paradoxically, mountains and dust storms often are real constituents of Earth’s deserts and semi-arid regions. The New Mexico Department of Health (NMDOH) is studying air quality and its impact on human health in the U.S.-Mexico border region. Emissions of fine particulate matter (PM10) from the Pleistocene pluvial Lake Palomas basin in the Chihuahuan Desert are a primary concern. Through Graham Harman’s object-oriented philosophy and rejection of the abiotic and biotic as ontologically distinct realms, “species” distribution models (SDMs) generated in Maxent become possible with just a small number of dust source presences located in MODIS visible band imagery. The first models from 2013 used multiple geoecologically relevant terrain objects to replace the geographically uncertain point sources. Although the terrain objects had been segmented with eCognition® from only two of the three ASTER GDEM parameters used, the early SDMs nevertheless suggested that dust storms are accessible on geomorphometric terms alone. Wind directions were derived in 2014 using dust plume image objects segmented from the MODIS thermal band translations of three dust storms. Three ‘wind-related terrain attributes’ were then generated for each storm in Whitebox GAT from the SRTM DEM to complement the new SRTM terrain objects. When comparably biased multi-object “background data” also replace the 10,000 pixel-level background samples in Maxent, the model AUCs decrease expectedly but still remain high enough for the SDMs to be potentially useful. The probability distributions are now “projected” over the same extents to the successively lower object and pixel levels in this novel and geographically scalar approach.