Cooperative Institute for Research in Environmental Sciences

A Conditional Skill Mask for Improved Seasonal Predictions

Kathy Pegion(1,2) Arun Kumar (3)

(1) CIRES (2) NOAA/ESRL/PSD (3) NOAA/NWS/NCEP/CPC

In Spring 2011, the Missouri and Ohio river valleys experienced precipitation greater than 200% of normal during the season, while much of Texas and New Mexico received only 25% of normal precipitation. The National Centers for Environmental Prediction (NCEP), Climate Forecast System (CFS) model successfully forecasted many aspects of the U.S. temperature and precipitation anomaly pattern during this time. However, the official NCEP/Climate Prediction Center (CPC) outlooks did not include the CFS precipitation forecast because the model was not considered to be skillful in these regions based on an average of historical forecast skill.

Is it possible to know ahead of time when forecast skill would be higher than average and use that information to make better seasonal predictions? For seasonal predictions, higher skill is likely related to the phase and amplitude of the El Nino Southern Oscillation (ENSO). This project investigates the potential for improving the skill of seasonal temperature and precipitation forecasts for the continental United States by constructing and applying a skill mask that is conditional on a commonly used index of ENSO, thereby enabling the forecaster to use the CFS during times when the forecast is expected to be more skillful.