The new hurricane-land coupled system and its test for landfalling tropical cyclones

The tropical cyclone prediction system at NCEP is currently coupled with a slab land surface model that only predicts land surface temperature with other land conditions, such as the land surface wetness and heat capacity, fixed. The use of fixed surface wetness and heat capacity may cause the land surface temperature prediction and thus the surface thermodynamic processes to be significantly biased where the changes of these conditions are large. In addition, the current prediction system uses the land surface wetness derived from its surface vegetation while the wetness in reality is largely dependent on precipitation, which has large temporal and spatial variability. These deficiencies of the operational system may limit the predictive skills of the prediction system for landfalling tropical cyclones. For the new system developed, the comprehensive Noah land surface model is used. The new system was tested for the landfalling tropical cyclones in the Atlantic basin for the season of 2003. The results indicate that the impacts of Noah land surface coupling are generally small on landfalling tropical cyclone track and intensity depending on the formulation of the surface exchange coefficients while the land surface temperature and soil moisture may be considerably different from those with the slab land surface model. However, the precipitation differences are more pronounced, implying the relative importance of land surface coupling in predicting inland rainfall and flooding. To examine the generality of the findings, a significant number of recent-landfall cases are being conducted. The preliminary results will also be presented.