The Development of High Resolution Super Probabilistic Forecast (HRSPF)

The NAEFS 5km CONUS and 6km Alaska probabilistic products in NCEP daily operation (4 times per day, every 6-hour, out to 16 days) are produced by the processes of 1). Bias correction of NCEP's GFS and GEFS respectively, the hybrid bias corrected high resolution deterministic and lower resolution ensemble forecasts, bias correction of CMC's GEFS; 2). Combination of two separated processed forecasts (NCEP and CMC) with systematic (statistical) adjustment; 3). Downscale to CONUS and Alaska regions from proxy truth (RTMA- Real Time Meso-scale Analysis). It has been demonstrated that the additional value has been added from each stated process. This study will focus on high resolution (5km and 2.5km) CONUS probabilistic forecast for surface wind and temperature by combining other comparable operational ensemble forecasts (such as bias corrected, downscaled SREF – Short-range Ensemble Forecast, bias corrected ECMWF global ensemble forecasts and et al.). The both of reliability and resolution are expected to be further improved through this multi-model, multi-center, multi-process and multi-resolution super ensemble.