Detection of New Dust Sources in Central/East Asia and Their Impact on Simulations of a Severe Sand and Dust Storm

Information regarding dust concentrations and size distributions is very important for determining air quality and aerosol–radiation–cloud interactions. Only by using a correct erosion database can the sectional dust emission scheme resolve detailed size distributions and determine where and how dust will be emitted. In this paper, the bias and reasons for dust emission in China Meteorological Administration Unified Atmospheric Chemistry Environment–Dust, an operational dust forecasting model, are analyzed using a heavy sand and dust storm episode. We used 18 years of routine sand and dust storm phenomena recorded at meteorological stations to retrieve and update the desertification details in the MBA sectional dust emission scheme adopted in China Meteorological Administration Unified Atmospheric Chemistry Environment–Dust. New desertification details include decreased erodibility in the area adjacent to Uzbekistan, Turkmenistan, and southern Kazakhstan, where Kyzylkum, Karakum, and Aralkum are located in central Asia, and in the Chinese deserts of Onqin Daga, Mu Us, and Gurbantungut. New desertification also results in increased erodibility in northern Mongolia. Comparisons show that the new desertification database decreases overestimation of dust emission in central Asia, including western Mongolia. It improves the underestimation of dust emission in northern Mongolia and the Gobi desert in southeast Mongolia, and the Taklimakan desert in China. Consequently, it corrects the overestimated dust cloud in the source area and in areas impacted by dust transportation. The timing, quantitative mass concentrations, and dust size distributions determined here are all more reasonable and rational than those of the original case. Key Points Dust emission in a heavy sand and dust storm (SDS) is overestimated in central Asia and underestimated in Mongolia by the sectional SDS model CUACE/Dust Desertification classification was improved using SDS data from meteorological stations and incorporating local land use changes Overestimations and underestimations of dust emissions/concentrations by CUACE/Dust were corrected using new desertification information