Tropical Cyclone Size: Climatology, Physics and Prediction

Tropical cyclones (TCs), generally known as typhoons, affect many parts of Asia everyyear, causing loss of lives and property. In issuing warnings to the public when a TC isapproaching, a weather agency will need to predict not only the strongest winds associatedwith the TC but also the extent of the damaging winds, such as gale-force winds. However,the predictions of the latter, commonly known as TC size, are currently done empiricallybecause of our limited understanding of the physical processes responsible for causingchanges in the size. A recently completed project by the PI, through the support of the RGC,has identified one possible process and generated the climatology of TC size for two oceanbasins. However, the results and the conclusions are based on a dataset that has only 11years of data, although it is already the largest dataset available.To ascertain these results, identify other possible physical processes, establish morerobust size climatologies for all ocean basins, search for possible interannual or eveninterdecadal variations of size, it is necessary to use a dataset that spans a much longer periodof time. This proposal is therefore to address these objectives by using such a dataset. Thesecond part of this project is to investigate the possibility of using an operationalnumerical-weather-prediction model for real-time prediction of TC size, and verify thepredictions against this dataset. Such predictions have never been attempted before.Specifically, the project has the following objectives:1) to validate the use of the new size estimates against existing observational datasets2) to establish the spatial climatology of TC size in all ocean basins3) to examine the seasonal, interannual and interdecadal variations as well as possibletrends in TC size for all ocean basins4) to verify the dynamic mechanisms proposed in previous studies and identify othermechanisms (especially thermodynamic) related to TC size change5) to explore the possibility of real-time prediction of TC size and to evaluate theaccuracy of such predictions, and6) to make detailed diagnostics of model outputs to further identify possible causes ofTC size changeResults of this study will therefore be of tremendous benefit to the operationalcommunity in terms of real-time forecasts of TC size, and to the scientific community interms of an improved understanding of the physical processes responsible for TC sizechanges.