Bias-correction of Tropical Cyclone Structure in a Global Ensemble Prediction System

Ensemble prediction offers significant benefits to many applications of weather forecasting, especially in conditions of high risk. Global ensemble prediction systems (EPS) demonstrate significant value in tropical cyclone track prediction, but their utility for quantitative intensity and structure prediction is limited because the relatively coarse resolution imposes systematic biases. We are diagnosing these biases, and developing a system to adjust predicted cyclones to remove these biases from the surface wind and pressure fields. These adjusted wind fields then provide better-calibrated wind exceedance probability guidance than the raw model output, and will in future be used to force a wave model and generate similarly better-calibrated wave probabilities.