GRAPES Typhoon Track Ensemble Forecast Method Research

Typhoon is a very destructive weather system on the tropical ocean.Its high intensity and sudden characteristics make the typhoon numerical forecast very uncertain.The typhoon ensemble forecast is an important forecast technology to describe the uncertainty of the typhoon numerical forecast.In order to improve the ability of typhoon ensemble prediction,this paper analyzes the uncertainty characteristics of Typhoon Structure and center location,improves the method of typhoon center location of members of ensemble prediction,carries out the sensitivity test of typhoon ensemble prediction,and provides scientific basis for improving the probability prediction of typhoon path.Firstly,taking typhoon 1617 "Silurus" as an example,the structural characteristics of the typhoon were diagnosed and analyzed by the method of synoptic diagnosis and analysis.The results show that the development of barotropic instability is the dynamic condition for the formation and development of the typhoon,and the positive SST anomaly is conducive to the formation and development of the typhoon.The diagnostic analysis of latent heat flux shows that CISK positive feedback mechanism is the key process of typhoon development and movement.The diagnosis of the evolution of typhoon vertical structure shows that the strengthening of vertical wind shear leads to the development of vertical convection,which is conducive to the strengthening of the high-level warm center,and the meridional gradient anomaly is conducive to the development of the typhoon along the meridional movement.The convergence of the low-level is the key to maintain the vertical warm center structure.Then,based on the data of typhoon best path from 2009 to 2018 from China Meteorological Administration and Japan Meteorological Agency,the uncertainty characteristics of typhoon best path vortex center location are analyzed,and the conditional typhoon vortex relocation method is designed Relocation,ctvr)is used to construct the threshold conditions of typhoon vortex center relocation,the mathematical processing of typhoon vortex separation and vortex relocation,etc.,and to use the global / regional assurance and prediction system regional ensemble system(graphsreps)of China Meteorological Administration,In this paper,the conditional typhoon vortex relocation ensemble prediction experiment and test evaluation are carried out for three typhoons(Maria No.1808,Tarmi No.1824 and Kong No.1825)in the Northwest Pacific in 2018.The research shows that the conditional typhoon vortex relocation method can effectively improve the probability prediction effect of the regional ensemble prediction typhoon path in grapes-reps,such as the ensemble prediction average of typhoon path The error is reduced,the consistency of ensemble prediction(the ratio of path dispersion to path root mean square error)is increased,especially the improvement of probability prediction effect in the early stage of prediction is more significant,but the improvement in the middle and later stage of prediction is limited.Finally,the sensitivity test is carried out on the relocation threshold condition of "Maria" typhoon and the change of each factor field before and after the specific gravity location.The results show that when the relocation threshold of conditional typhoon vortex is set to 15 km,the consistency of the initial prediction is slightly higher than that of other test results,but with the extension of the prediction time,the consistency difference of different threshold tests is significantly smaller than that of the initial prediction.The results of the comparison of the elements before and after the relocation show that after the conditional typhoon vortex relocation,the typhoon path error of the members of the assembly is obviously reduced in the early stage of the prediction,and the typhoon vortex circulation and the large-scale environment field are still relatively continuous and coordinated,the large-scale environment field outside the vortex circulation has the characteristics of consistency,the minimum pressure error,the maximum wind speed error and the precipitation prediction skills Basically unchanged.