Precipitation Variability under different Sea Surface TemperatureScenarios and Implications for Decision Making

Abstract:

The current generations of seasonal forecast systems are also accompanied by an extensive set of hindcasts. For example, Climate Forecast System version 2 (CFSv2) at National Centers for Environmental Prediction has hindcasts starting 1981 and includes an ensemble of four 9-month forecast every 5^th day. Hindcast datasets are generally used for (a) bias correction and calibration of real-time forecasts, and (b) assessment of prediction skill. Hindcasts also provide a large sample of seasonal means that can be used for analysis of climate variability – CFSv2 hindcasts have a sample of ~ 5000 for each season. As an example of the utility of the hindcast dataset, this large sample data set is analyzed to document seasonal mean precipitation variability over California, and how it depends on the sea surface temperature (SST) variability in tropical Pacific. An overarching conclusion of the analysis is a wide range of seasonal mean precipitation outcomes that can occur under similar SSTs. Possible implications for precipitation variability (for a fix forcing) on decision making are also discussed.