In 2001 the Environmental Modeling Center (EMC), one of the National Centers for Environmental Prediction (NCEP), established the Short Range Ensemble Forecast (SREF). Currently it consists of 21-members, with forecasts to lead time of 87 hours produced 4 times per day. Individual members as well as basic ensemble products are available at 40km resolution (the 212 grid). Probabilistic forecasts of exceeding certain thresholds, ensemble mean, spread are generated for over 70 selected variables.

A statistical downscaling method will be implemented at NCEP for SREF in early 2012. The approach to be used is similar to that used in the North American Forecast Ensemble System (NAEFS), and is an attempt to establish the relationship between coarse and fine resolution data and project coarse resolution information onto finer grid. The operational Real-Time Mesoscale Analysis (RTMA) for CONUS with 5-km resolution (NDGD grids) is chosen as the truth when compared with analysis, which is from NCEP NAM Data Assimilation System (NDAS) to 5-km resolution. The differences between the two analyses are accumulated via a decaying-average algorithm and using a “Downscaling Vector” in the downscaling process.

When applied to NAEFS, the results show that this downscaling method can effectively reduce systematic forecast errors. The downscaled forecast also displays more detailed forecast information than the raw forecast. For SREF, we will show results of the testing of the application of this downscaling to: a comparison of raw, bias corrected, and bias corrected downscaled NCEP SREF forecasts to observations.