An analysis of along- and cross-track forecast errors and error biases for TCs in the Atlantic Basin

Official track forecasts for a selection of Atlantic tropical cyclones that underwent extratropical transition in the 2000s have been decomposed into along- and cross-track components to identify forecast errors and error biases. We also investigate the seasonal and storm dependence on these error statistics. Decomposition of track forecast errors into motion-relative components permits easy identification of forecast biases and construction of the shape of the error distribution at various forecast lead times. Results from the study can be used to define an elliptical shape to the error distribution versus the traditional circular distribution. Stratification of the motion-relative error distributions into latitudinal bands, for example, will help refine the track forecast error cone currently displayed with forecast track maps. Consequently, wind speed probability swaths centered about a track forecast graphic could also be refined by recasting the track error statistics in this manner.