Precipitation and Temperature Forecast Performance at the Weather Prediction Center

As the skill of numerical weather prediction (NWP) and associated post-processed guidance continues to improve, recent debate asks to what degree can expert human forecasters add value to NWP? The Weather Prediction Center (WPC) has a broad mission to serve as a center of excellence in quantitative precipitation forecasting, medium range forecasting, winter weather forecasting, surface analysis, and the interpretation of operational NWP. Historically, forecasters at the WPC have had access to a large portion of the available international model guidance suite, including multi-model ensemble information from international partners. WPC's unique national forecast mission coupled with its access to state of the art model guidance provides a rare opportunity to assess human performance relative to ever-improving NWP. This talk will examine precipitation and temperature forecast performance at WPC using multi-seasonal verification and case study approaches. Stringent contemporary verification accounting for ensemble approaches, bias-correction, bias sensitivity, and statistical significance confirm that the human can make statistically significant improvement over competitive deterministic model guidance for precipitation and maximum temperature. However, downscaled, bias-corrected ensemble forecasts of these elements available near the same time as the human-modified forecasts exhibit similar skill as the human forecasts. The implications of these results will be discussed.