| description abstract | This paper compares the performance of two multi-level high-resolution baroclinic tropical cyclone models which are currently in operational use. In the Atlantic, the National Hurricane Center utilizes the Movable Fine-mesh Model (MFM) for objective forecast guidance. In the western Pacific, the Navy's Nested Tropical Cyclone Model (NTCM) is employed by the Joint Typhoon Warning Center for this same purpose. However, the computer resource requirements, basic design and method of use are substantially different for the two models. Accordingly, two separate tests were conducted. In the first, or operational comparison, the NTCM was initialized with the Fleet Numerical Oceanography Center tropical analysis while the MFM was initialized with the National Meteorological Center (NMC) global analysis. Forecast errors for both models were very similar in this experiment. The other test isolated model differences by initializing both models with the NMC analysis only. Considerably degraded performance for the NTCM was noted. This result emphasizes the sensitivity of dynamic model performance to the large-scale analysis, and underlines the importance of analysis-prediction system co-development. Each model is briefly described and case-by-case performance evaluations are presented. The results indicate that dynamic models promise great potential for significant improvement in tropical cyclone movement forecasting, particularly for forecast intervals of greater than 36 h. | |
