Development and Beta-Test of the Systematic Approach Expert System Prototype as a Tropical Cyclone Track Forecasting Aid (SAFA)

Abstract

This report summarizes two advancements of the Systematic Approach to tropical cyclone track forecasting in the western North Pacific. First, a knowledge-based expert system that assists the forecaster in dealing with thee information management, visualization, and decision-making challenges in track forecasting. A prototype module for evaluation of the dynamical model track guidance is described. This Systematic Approach Forecasting Aid (SAFA) leads the forecaster through a decision process to detect and exclude model tracks that are likely to have 72-h errors exceeding 300 n mi, and thus form a selective consensus (SCON) that is more accurate than a non-selective consensus (NCON) track. Second, a simulated real-time beta-test of the prototype SAFA module was carried out by a team at the Naval Postgraduate School for the 1998 cyclones 19W through 30W. Among the results from the beta-test are: (i) the team was able to use SAFA to recognize error mechanisms based on conceptual models from previous research; and (ii) the SCON track errors were smaller than the NCON errors. Recommendations for further improvements and testing are included.