The Use of Meteorological Data in the Prediction of Tropical Cyclones

Recent studies have shown that the ability to forecast tropical cyclone motion and intensity has not improved significantly over the past decade, in spite of apparent advances in the statistical and dynamical modeling of the motion- and intensity-change processes. This static situation has focused attention on the meteorological data and analysis techniques that provide input to these models. The quality and quantity of data have apparently not kept pace with the modeling efforts; in fact, increased sophistication in modeling alone may be counterproductive. Steps are being taken in the meteorological community to correct this problem. This study examines the entire data problem, discusses the remedial measures being taken, and points out some of the pitfalls in the statistical use of meteorological data.