An Analysis of the Uncertainties in Precipitation Season Forecasts over the US West Coast

With an extensive data base of seasonal hindcasts and real time forecasts available as part of the North American Multi-Model Ensemble (NMME) effort, this study explores whether the low skill of seasonal precipitation forecasts over the US west coast in the NCEP Climate Forecast System version 2 (CFSv2) because of issues that are specific to CFSv2 or other NMME models also have the same feature. The data base also allows us to assess the signal-to-noise and inherent predictability across a range of models. In the analysis, we also examine the uncertainties in the model forecast precipitation responses to anomalous SST across all El Niño events in 1982-2017, and the uncertainties is response across models for specific El Niño events. In general, over the US west coast the precipitation forecast skills from all the NMME models are low and the uncertainties in El Niño response across models are large as well. The results for the west coast of the US are contrasted with that over the southeast US, where responses to El Niño SSTs show much more consistency across the same models. The linkage of the low skill and lack of consistency in El Niño responses to the inference of low signal-to-noise ratio for the precipitation over the US west coast is also explored.