Hybrid Dynamical–Statistical S2S Forecasts with Weather Types over North America

Dynamical forecast models provide a foundation for S2S forecast systems, but systematic errors may arise for various reasons, including insufficient spatial resolution, insufficient ensemble size, and errors in physical parameterizations. Despite these flaws, the ability of dynamical models to simulate sources of prediction skill and their large-scale circulation responses allows us to draw from empirical predictor/large-scale circulation relationships to compensate for these shortcomings. In this study we use the framework known as weather types (WTs) to act as the mediator for a hybrid dynamical-statistical S2S forecast system. WTs are large-scale, quasi-stationary circulation patterns that, in this application, are determined by k-means clustering of geopotential height. We generate both subseasonal (weeks 3-4) and seasonal forecasts for December – February over eastern North America by taking dynamical model forecasts of WTs and then using empirical relationships to translate these WT forecasts into probabilistic temperature and precipitation forecasts. We generate and evaluate this forecast system with hindcasts from the Geophysical Fluid Dynamics Laboratory (GFDL) Forecast-oriented Low Ocean Resolution (FLOR) dynamical model for the period of 1981-2016. Preliminary results suggest that this hybrid dynamical-statistical approach substantially reduces seasonal precipitation forecast error over the raw model forecasts. Pattern-dependent circulation biases and sources of predictability associated with the dominant WTs are investigated.