Predictability of Southwestern United States Precipitation from non-ENSO Drivers and their Interactions with ENSO

Precipitation prediction on subseasonal and seasonal timescales is plagued by low skill, particularly over the Southwestern United States (SWUS), a populous and agriculturally productive region for which the drivers of precipitation are not well understood. This issue is exemplified by the persistence and season-to-season fluctuations of the ongoing SWUS drought, which are unexplained by the most historically effective predictor of SWUS precipitation, the El Niño Southern Oscillation (ENSO). Due to this, we aim to identify non-ENSO drivers of SWUS precipitation and examine how those drivers interact with ENSO. We do this through use of reanalysis data, the fully coupled Community Earth System Model version 2 (CESM2) large ensemble, and the Community Atmosphere Model version 6 (CAM6) with prescribed historical sea surface temperatures. In both CESM2 and CAM6, we find that wavenumber 5 zonal wavetrains are a primary driver of SWUS precipitation. Residing in the subtropical and midlatitude jet streams, they are associated with subtropical deep convection over the western Pacific and northern Indian Ocean, but are not associated with tropical convection. Their propagation and influence on SWUS precipitation can be affected by the upper level wind basic state. For example, in CESM2 La Niña seasons, the jet stream is shifted northward and zonal waves effectively influence SWUS precipitation, with the opposite during El Niño. Similarly, CESM2 has a mean southward bias in the upper level Pacific jet compared to reanalysis and CAM6, reducing the influence of zonal wavetrains on SWUS precipitation in CESM2. Overall, our work highlights atmospheric patterns of interest for SWUS subseasonal and seasonal forecasting, as well as the modulation of teleconnections due to basic state variation from ENSO and model biases.