Extracting the tropospheric short-wave influences on subseasonal prediction of precipitation in the United States using CFSv2

Abstract

The development of subseasonal precipitation forecasts on regional scales is becoming an active area of research. Climate forecast models have shown deficiencies in predicting the extreme precipitation anomalies at medium to long-range timescales. This study explores the subseasonal relationships between tropospheric short-waves and regional precipitation anomalies across the continental United States and evaluates capabilities of the NCEP Climate Forecast System Version 2 (CFSv2) in resolving these relationships. A regional precipitation proxy is derived from the prediction of the upper tropospheric short-waves based on multiple linear regressions. Across the six climate regions defined by NOAA, the 30-day reforecasts of this short-wave based precipitation proxy are compared to identify the combinations of month and zonal wavenumber that exhibit the highest prediction score. Forecast of this precipitation proxy over certain regions is found to outperform the direct precipitation output of CFSv2 out to 4 weeks, suggesting a subseasonal predictability in precipitation that can be harvested from persistent circulation features.