A statistical post-processing technique is developed and tested to reduce the Navy global model (NOGAPS) track forecast errors for tropical cyclones in the Atlantic, North and South Pacific, and South Indian Oceans. In addition to the current storm position, intensity, and date, the set of 24 predictors includes various segments in the 00-36h NOGAPS forecast track as well as a 00-36h backward extrapolation that is compared with the known recent track positions. Another key piece of information is the offset of the initial NOGAPS position, which is an extrapolation made about 2 h before the model run time, relative to an updated position that will be known by 6 h after the synoptic time, which is when the NOGAPS forecast is actually available for use by the forecaster.
An initial study using a development sample of western North Pacific tropical cyclones during 1992-96 produced promising results. For this development sample, the adjusted NOGAPS track errors are reduced by about 50 n mi (93 km) at 12 h, 33 n mi (61 km) at 36 h, and 24 n mi (44 km) at 72 h. Independent tests with a 1997 western North Pacific sample, 1995-97 Atlantic sample, and 1996-97 eastern and central North Pacific sample of NOGAPS forecasts have similar improvements from this post-processing technique. Oftentimes, NOGAPS forecasts that the tropical cyclone will move away from a persistence track immediately after the initial time, and the entire NOGAPS forecast might be considered suspicous. The impressive error reduction at shorter forecast intervals demonstates this technique's ability to use the updated initial position and knowledge of past motion in order to "restart" the NOGAPS forecast in a more plausible direction. The forecaster is then more likely to have confidence in using the adjusted NOGAPS forecast.