Client-Centered Calibration of the NOAA Climate Forecast System

Prescient Weather proposes four Phase I tasks to increase the value of the new NOAA Climate Forecast System to commercial customers. The capabilities demonstrated and explored in Phase I will be developed and integrated in Phase II as components of a new Seasonal Information and Decision Support System (SIDSS) for our World Climate Service customers. The Phase I tasks are: • Improve season forecast calibration with a new climate-conserving calibration algorithm that produces relatively flat rank probability diagrams; • Convert calibrated forecasts of meteorological variables into forecasts of impact and decision variables such as degree days or wind power availability’ • Explore calibration of two-to-four week forecasts with conditioning on expected flow patterns; • Explore the potential of model output statistics (MOS) to calibrate and improve weekly and monthly forecasts of seasonal variability. Several of the tasks will explore the use of principal component methods to project forecasts on the historical verification data and to define flow regimes for conditional calibration. The SIDSS development in Phase II will focus on the client decision context, present historical and predicted information including numerical and analog forecast, and facilitate client development of individualized forecasts. It is a key part of our commercialization strategy.