Evaluating forecast accuracy of tropical cyclones undergoing rapid size changes in the North Atlantic

The ability to accurately forecast tropical cyclone (TC) track and intensity is of vital importance across the world both from an economic and a humanitarian perspective. Of similar importance is forecasting the size of the TC, as hazards such as high winds, heavy rain, and storm surge can occur well away from the TC center. While TCs in the North Atlantic have been shown to rapidly change in size, little work has been done to evaluate our ability to forecast these events.

In order to investigate the accuracy of TC size change forecasts, this study investigates the performance of the Global Ensemble Forecast System Reforecast (GEFS/R) dataset during periods of North Atlantic TC size changes. The GEFS/R was created by the Earth System Research Laboratory's Physical Sciences Division (ESRL/PSD) using the 2012 version of NCEP's Global Ensemble Forecasting System and consists of an 11-member ensemble of forecasts. In this presentation we will investigate the ability of the GEFS/R at different forecast lead times to accurately forecast size changes in three cases of TC size change. Using the ensemble members we will investigate the spread and uncertainty in the forecasts and examine the physical mechanisms associated with the forecast successes and failures.