Applying self-organizing map to improve predictive understanding of sub-seasonal variability and its impact on summer droughts over the US Great Plains

Although the influence of sub-seasonal variability on warm season droughts, such as the 2012 Great Plain droughts, has been well documented, the variation and evolution of the large-scale dynamic and thermodynamic patterns of the sub-seasonal variability, and their linkage with land surface feedbacks have not been adequately characterized. This study uses a multivariate self-organizing maps (SOM) approach to diagnose precipitation anomalies and the associated evolution of the anomalous large-scale atmospheric patterns over the United States’ Great Plains (US GP) during the warm season (April-August), as represented by standardized anomalies of 500 hPa geopotential (Z500′), integrated vapor transport (IVT′), and convective inhibition index (CIN′). Six circulation patterns that control warm season rainfall variability over the US GP are identified. The Southern GP droughts are attributable to more frequent and persistent northern low-southern high as well as dominant high circulation patterns over the US GP, and is connected to larger negative CIN’ instead of lower IVT’. In contrast, Central GP droughts are attributable to more frequent and persistent western low-eastern high, or northern high-southern low, or dominant high patterns, and are also notably linked to a larger negative IVT′ instead of larger CIN′. Overall, these results suggest that land surface feedbacks may play a more important role in warm season droughts over the Southern GP than reduced moisture transport. However, reduced moisture transport may play a more important role than land surface feedbacks in droughts over the Central GP.