Detection of tropical cyclone track changes from the ECMWF ensemble prediction system

Authors

  • Hsiao-Chung Tsai,

    1. Graduate School of Engineering and Applied Sciences, Department of Meteorology, Naval Postgraduate School, Monterey, California, USA
    Search for more papers by this author
  • Russell L. Elsberry

    1. Graduate School of Engineering and Applied Sciences, Department of Meteorology, Naval Postgraduate School, Monterey, California, USA
    Search for more papers by this author

Corresponding author: R. L. Elsberry, Department of Meteorology, Naval Postgraduate School, 589 Dyer Rd., Monterey, CA 93943, USA. (Elsberry@nps.edu)

Abstract

[1] A cluster analysis of the 51 member European Center for Medium-range Weather Forecasts ensemble forecast tracks of existing tropical storms from the THORPEX Interactive Grand Global Ensemble archive is utilized to detect multiple track solutions, and weighted-mean vector motion (WMVM) tracks are calculated for the ensemble tracks and for the individual track clusters. When three track clusters were present, larger errors occurred in the corresponding deterministic model forecast. Reliability of the European Center for Medium-range Weather Forecasts ensemble forecast tracks when two distinct clusters exist is examined in terms of the hit rates when the cluster with the larger number of members is selected. A larger hit rate is achieved if track clusters with at least 70%, 80%, or 90% of the members are selected. In these situations, the forecaster can select the WMVM tracks for those clusters and have confidence that a more accurate track than the overall WMVM track will generally be predicted.

Ancillary