Abstract:The most common tropical cyclone (TC) center identification method for searching for minimum sea level pressure (MSLP) still has some shortcomings. In this paper, TC center identification method and intensity estimation method based on GFDL vortex tracker technique (GVT) are studied in the regional numerical weather prediction model application. Detailed comparisons are conducted between GVT and three traditional TC center location and intensity estimation techniques which include MSLP, the method of using maximum vorticity value at 500 hPa (VT500), and the method of using minimum geopotential height at 500 hPa (H500). The results show that using H500, VT500 and MSLP methods to determine TC center location, can generate some degrees of zigzag shaped irregular swing paths, especially the VT500 method. The VT500 method has the largest irregular swing amplitude and the positioning result is the worst. GVT method can better correct the zigzag shaped swing path and make the model prediction path closer to the observed one, so its positioning result is the best. However, the positioning results of H500, MSLP and GVT are relatively close when the typhoon location is far away from the coastline. The errors of the minimum pressure of TC center by these four methods show that the VT500 method has the largest error, the GVT method has the smallest error, and the errors by MSLP and H500 methods are between those of VT500 and GVT. The errors of maximum wind speed near the TC center by four methods show the errors of the four methods are not much different. The operational application assessment in 2016-2018 show that the Zhejiang Typhoon Vortex Tracker System based on GVT technology can improve the prediction quality of Zhejiang WRF ADAS Regional Model System and Zhejiang WRF ADAS Rapid Refresh System.