Weeks 3 & 4 averaged weather forecasts over the United States from Climate Forecast System Reanalysis and Reforecasts(CFSRR) at NCEP

In the seamless suite of operational forecast products starting from days out to many seasons, the hole that exists for weeks 3 & 4 forecasts have always remained an enigma and frustration. As part of an ongoing and recent reanalysis and forecast efforts at NCEP with the climate forecast system and reanalysis (CFSRR) using coupled ocean-atmosphere models, global forecasts out to 45 days are available four times daily from 1999 onwards. Even though it is well known for some time that daily weather is not predictable beyond ten days, we plan to use this unique set of data to examine if some aspects of circulation averaged over weeks 3 and 4 can be predicted at certain times. If there exists some pockets of periods with some useful skill, we will put them in the context of ‘favorable circulation regimes' if any, that enable these forecast periods of opportunity and whether the skill scores are related to weeks 1 & 2 forecast skill scores. Besides presenting results for 500 mb Geopotential height scores, we will pay particular attention to US Temperature and Precipitation fields.