Identification of dust sources and hotspots in East Asia during 2000–2015: implications for numerical modeling and forecasting

Abstract. More detailed knowledge regarding recent variations in the characteristics of East Asian dust events and dust sources can effectively improve regional dust modeling and forecasts. Here we reassess the accuracy of previous predictions of trends in dust variations in East Asia, and establish a relatively detailed inventory of dust events based on satellite observations from 2000 to 2015. More than 2000 Moderate Resolution Imaging Spectroradiometer (MODIS) images of 462 sand and dust storm events over East Asia were collected and analyzed, and individual events were tracked back to their sources through a combination of color RGB images, brightness temperature difference, and trajectory simulations using the HYSPLIT model. Decreased dust event frequency in spring but increased frequencies in summer and autumn were observed. Of the identified dust emission sources, sandy lands and lake beds, rather than the sandy and stone deserts, were found to be the dominant dust sources. Dust hotspots in East Asia are mainly dry lake and river beds and alluvial fans. Recent changes in land use associated with anthropogenic activities (mining and excessive exploitation of water resources) are revealed as one of the major factors leading to an expansion of dust source regions, especially for the northeastern part of Taklimakan desert. Trajectory analysis also shows that dust can even be transported northwards by the Mongolia Cyclone, to the Far East region and even the Arctic Circle, potentially affecting the climate and ecosystem of the Arctic region. Recent physically-based dynamic approaches adopted in dust models reduce the reliance on empirical source functions in dust modeling; however, the validity of down-scaling these schemes to regional scale needs to be further verified with "ground-truth" information as reported here.