Comparison of Daily Precipitation Statistics for the United States in Observations and in the NCEP Climate Forecast System

Abstract

An intercomparison of the statistics of daily precipitation within seasonal climate over the conterminous United States is carried out using gridded station data and output from the NCEP Climate Forecast System (CFS). Differences in the occurrence of daily precipitation between the observations and a set of CFS reforecasts are examined as a function of forecast lead time for 1982?2005. Difference patterns show considerable evolution depending on season and lead time, with positive biases in CFS at most locations and leads except along the southern tier of states during the spring and summer months. An examination of differences in daily precipitation statistics by ENSO phase and in the frequencies of wet and dry spells is also conducted using a longer period of gridded daily station data (1948?2006) and a pair of 100-yr CFS coupled simulations. These comparisons expose additional details of the regional and seasonal dependence of the bias in the CFS simulations and reforecasts over the conterminous United States. The analysis motivates additional synoptic studies aimed at improving the linkage between daily precipitation and related circulation features in CFS. Prospects for using this information to develop more reliable ensemble-based probabilistic forecasts in real time at leads of 2?4 weeks (e.g., risks of heavy rain events) are also considered.