Precipitation forecasting with adaptive parameterization selection

The current approach many operational ensemble prediction systems utilize is a single set of parameterization schemes for all model forecasts. However, previous studies have shown that variability resulting from changes in model physics is important to ensemble forecasts, suggesting that certain parameterization choices may be more skillful than others for a given situation. The question arises of whether a training dataset composed of previous model forecasts could be used to select parameterizations for an upcoming forecast, improving the skill of that forecast. Similar work has demonstrated that the incorporation of such a dataset can allow for the creation of ensemble statistics from deterministic models with similar performance to that of a traditional initial condition ensemble. It has additionally been shown that this type of dataset can be useful for model post-processing and bias correction. This study seeks to find if predictions of optimized ensemble parameterizations from previous verification will lead to an adaptive ensemble with improved forecast skill compared to a static-parameterizations ensemble for both climatological and extreme precipitation events. Analogue techniques were used to select analogues to ensemble mean precipitation forecasts. These analogues were then be used to predict the optimized set of parameterization schemes for the given upcoming pattern, and a new forecast was generated using this optimized set. This new forecast was compared to control systems with static physics to determine if a reduction in error is observed. Comparisons of analogue predictors and threshold sensitivity were investigated on a 12-month training dataset at synoptic resolution, which was the basis for an additional three months of adaptive forecasts. Forecasts were verified by multiple statistical methods, both grid-based and object-based, to quantify any change in forecast precipitation error.