Systematic and Integrated Approach to Tropical Cyclone Track Forecasting. Part 2:Climatology, Reproducibility, and Refinement of Meteorological Knowledge Base

摘要

This report continues the development of the Systematic Approach to tropicalcyclone track forecasting by Carr and Elsberry (1994) with specific application to the western Pacific region. Five years (1989-93) of 12-h Naval Operational Global Atmospheric Prediction System (NOGAPS) analyses are examined for every tropical cyclone to establish a climatology of Environment Structure characterizations. Frequencies of the four Synoptic Patterns and the six Synoptic Regions are calculated, and characteristic tracks while in each of the Pattern/Region combinations are provided. A four-year subset of NOGAPS analyses is used in a reproducibility test in which three trainees attempted to achieve the same Environment Structure assignments as an experienced forecaster. Approximately 81% of the Synoptic Patterns, 86% of the Synoptic Regions, and 77% of the Pattern/Region combinations were correctly assigned by the trainees. However, certain Pattern/Region combinations were found to be poorly identified, which indicated a need for some new training materials. Approximately 81% of the Pattern/Region transitions to a correct or similar combination were detected by the trainees. About 72% of these correct/similar transitions were identified within + 12 h of the actual transition, and 88% were identified within + 24 h. Certain transitions were consistently missed, which again indicates a need for improved descriptions and training. Refinements of the Meteorological Knowledge Base of the Systematic Approach were made on the basis of the five year climatology and reproducibility test. These refinements, which are summarized in Chapter 4.9, include new transitional mechanisms associated with Monsoon Gyre Formation and Dissipation and Reverse oriented Trough Formation. A new conceptual model called Subtropical Ridge Modulation is defined in which superposition of a midlatitude trough o.