资源环境科技发展态势分析平台

Compared with drought onset, the quantification and attribution of drought demise are not well studied. Meteorological droughts usually terminate more rapidly than they initiate, making it hard to define a transition period leading to drought demise with monthly data. In this study, methods of quantifying and attributing drought demise are applied to the Modern-Era Retrospective Analysis for Research and Applications version 2 (MERRA-2) using modified Standardized Precipitation Index representing meteorological drought calculated at pentad intervals to resolve subseasonal demise events. Methodologies to attribute three specific causes of drought demise in nine climate regional divisions over conterminous U.S. (CONUS) are developed. The three phenomenological causes are tropical cyclones, atmospheric rivers, and changes in land atmospheric feedback. Atmospheric river is the most common of the factors for drought demise, over most of the regions in the eastern, and central U.S. tropical cyclones are important causes over the Southwest, the South in fall, and the Southeast in summer. Evolving land atmospheric feedback is a factor mainly over the central and southwestern United States. These attributions estimated may not be representative of the long-term climatologies of drought demise due to the short duration of MERRA-2. A representativeness test is conducted for estimating the three impacts on drought demise using subsamples from a large ensemble of climate model simulation including several centuries of data. Thirty to forty years is not long enough to be representative of local long-term statistics for the attribution of the causes of demise of extreme events like drought but may be adequate at regional scales.