Time Consistency of Track Prediction Aids for Western North Pacific Tropical Cyclones

Abstract

The time consistency of three tropical cyclone track prediction aids available at the Joint Typhoon Warning Center (JTWC), Guam is examined by comparison of successive cross-track and along-track errors relative to the best track (postseason analysis) positions. The cross-track errors are divided into three groups (terciles) that are referred to as Left, Center or Right categories. Generally, the aids have more time consistency when the error in the previous forecast is in the Right category than in the Left category. A forecast verification within the Center category of 24-b path errors generally is followed by equal chances of Left, Center or Right error categories within 24?48 h. The One-way influence Tropical Cyclone Model (OTCM) provides the most consistent forecasts in time. In general, the Climatology and Persistence (CLIPER) 24-h verification provides little information as to how the subsequent CLIPER forecast wig verify. The wobble about the actual track is defined as the difference in cross-track errors between successive 12-h forecasts. The OTCM has the smallest wobble in the 24-h and 48-h forecasts and the same wobble as JTWC at 72 h. It is concluded that the time consistency of successive forecasts should be another measure of the value of tropical cyclone track aids.