Verification of Official Monthly Mean 700-hPa Height Forecasts: An Update

View More View Less
  • 1 Climate Prediction Center, NCEP/NWS/NOAA, Camp Springs, Maryland
  • | 2 Research and Data Systems Corporation, Greenbelt, Maryland
© Get Permissions
Full access

Abstract

Quality analyses have been performed on a 21-yr record of monthly mean Northern Hemisphere extratropical 700-hPa height anomaly forecasts issued by the National Weather Service. A positive trend in skill noted a decade ago is shown to have continued to recent years. This trend is present in terms of overall reduction in squared error as well as individually in reduction of both phase and amplitude errors for all three subdomain sectors examined. The higher skill in the last decade principally is concentrated in forecasts for winter months and particularly over the oceans and at high latitudes and is attributed to advances in global numerical weather prediction. Prior to the 1980s, average forecast bias varied from region to region and overall was not large. Since then it has tended to be negative for all subsectors, mainly as a result of negatively biased height anomalies in midlatitudes for forecast months in the winter and spring. This bias is, perhaps, a reflection of the anomalous observed warmth during the period.

An attempt to improve the quality of the forecasts with a principal component filter had modest success in the sense that reductions in squared error were achieved aside from those that would have been expected from simple smoothing.

Abstract

Quality analyses have been performed on a 21-yr record of monthly mean Northern Hemisphere extratropical 700-hPa height anomaly forecasts issued by the National Weather Service. A positive trend in skill noted a decade ago is shown to have continued to recent years. This trend is present in terms of overall reduction in squared error as well as individually in reduction of both phase and amplitude errors for all three subdomain sectors examined. The higher skill in the last decade principally is concentrated in forecasts for winter months and particularly over the oceans and at high latitudes and is attributed to advances in global numerical weather prediction. Prior to the 1980s, average forecast bias varied from region to region and overall was not large. Since then it has tended to be negative for all subsectors, mainly as a result of negatively biased height anomalies in midlatitudes for forecast months in the winter and spring. This bias is, perhaps, a reflection of the anomalous observed warmth during the period.

An attempt to improve the quality of the forecasts with a principal component filter had modest success in the sense that reductions in squared error were achieved aside from those that would have been expected from simple smoothing.

Save