Attribution of North American Subseasonal Precipitation Prediction Skill
Subseasonal prediction of precipitation is one of the current grand challenges. Over the United States, NOAA’s official U.S. monthly precipitation forecast skill has been historically low. In this study we aimed to understand seasonal and regional variations of the North American subseasonal (weeks 3 – 6) precipitation skill. We utilized five subseasonal reforecast datasets (ECMWF-IFS, NCEP-CFSv2, NCEP-GEFSv12, CESM1, CESM2) as well as uninitialized simulations with atmospheric components of 4 of the models (ECMWF-IFS, NCEP-CFSv2, CESM1, CESM2). We evaluated precipitation skill over North America utilizing the anomaly correlation coefficient (ACC).
We demonstrate that the highest precipitation skill for weeks 3-6 over N. America is found in December, February, March (DJF) in the Southwest United States and Central America, and in March, April, May (MAM) over the Western United States and Central America. Precipitation skill for weeks 5-6 has similar amplitude and variability to that derived from AMIP simulations. Comparison of precipitation skill derived from the reforecasts to AMIP simulations elucidates that boundary forcing, from prescribed sea surface temperatures (SSTs) in the Pacific (from El Nino Southern Oscillation (ENSO)) is the largest contributing factor to precipitation skill in DJF over the Southwest United States and Central America, as well as parts of Florida for weeks 3-6. This connection is not clear in other seasons.
In this study, we quantify the contribution to the subseasonal, weeks 3-6, precipitation skill from boundary forcing, and more specifically, ENSO. We find that ENSO is the largest driver of DJF precipitation in Southwest United States, Central America, and in Florida for weeks 3-6. The connection between ENSO and N. American precipitation in other seasons is not clear.