Probability Forecast Model of Tropical Cyclone Heavy Rainfall Area Based on Multielement Diagnosis

In this paper, a multi-element heavy rainfall area diagnosis scheme for Tropical Cyclones (TC) of different intensity was presented, and an improved probability forecast model based on this scheme was built and verified. The results show that when the relationship between heavy rainfall area of TC and physical elements was discussed from the aspects of relative centroid and accuracy, not only how the physical elements were neighbour to rainfall area, but also how they were matched each other, were considered. Then the ability of different elements at different levels to characterize TC heavy rainfall area can be distinguished. The elements, such as specific humidity, helicity, wet enthalpy, pressure energy vorticity, pseudo-equivalent temperature, humidity and pressure energy enthalpy advection, can be sorted in a descending order to characterize the heavy rainfall area. Furthermore, the optimal levels of different elements are different. Therefore, the multi-element diagnosis schemes will change with TC intensity. The improved forecast model considered the effect of TC intensity evolution on severe rainfall process, and its forecasted 0.5 probability contour matches the heavy rainfall area best. The probability forecast results can decrease the displacement error of the heavy rainfall area of TC Soulik forecasted by ECMWF model. The probability forecast will be improved with the increase of the TC intensity.

Keywords: Tropical Cyclone; Helicity; Centroid; Accuracy; Heavy rainfall Area; Probability Forecast