Research On The Extrapolation Of Radar Echo Based On The Weather Radar Retrieved Variables And Optical Flow

The major difficulty of the radar nowcasting is the forecast of echo's generation and dissipation.The optical flow algorithm is mainly based on mathematical operation,without considering the influence of the physical field changing for the precipitation system,which causes errors.In this paper,considering the correspondence between the physical field and the evolution of the precipitation system,the radar echo extrapolation results can be revised by physical fields,to achieve a better forecast effect.Local temperature advection,storm relative helicity,divergence,vertical velocity and divergence vertical flux have a good relationship with two precipitation processes in Nanjing in 2016.Analysis shows that the temperature advection structure of‘up-cold and down-warm'in stratiform precipitation or‘down-cold and up-warm'in the back of convective storm is positive for precipitation enhancement.The variation of storm relative helicity is 2.5³.5 hours ahead of precipitation development;divergence distribution over 3km-layer affects directly the evolution of precipitation progress.Vertical ascending velocity over 4km-layer within the storm is a significant factor for precipitation enhancement.The negative divergence vertical flux over4.5km-layer within the storm is related to precipitation strengthening.Hence,these derived radar products can be used for precipitation nowcast.The LK-optical flow method is applied to three types of precipitation?large-area mixed precipitation,typhoon and thunderstorm?for forecast test,and the results show s:prediction effect in the large-area mixed precipitation?linear motion echo?,typhoon?rotating motion echo?and thunderstorm?fast changing echo?,is well.However,LK-optical flow method also has some shortcomings.It is difficult to forecast the gradual decrease of the stratified echo.Newburn strom cells can't be forecasted.The forecast of the extension and contraction of the echo edge is not accurate.As the forecast time increases,the forecast effect will also become worse.These errors are caused by the limitations of the optical flow,which does't have the ability to forecast the intensity change of precipitation system.In this paper,10 precipitation processes with similar background fields are selected to fit the empirical relationship between physical quantities such as temperature advection and dBZ₈0)?69?).Then the corresponding echo correction can be obtained.In three cases of large-scale precipitation,typhoon and thunderstorm,this empirical relationship improves the prediction results.The improvement of 60 min is better than 30 min,which can make up for the forecast effect reduction with extrapolation of time growing.The correction effects of large-scale precipitation and typhoon are better using this empirical relationship,but the correction effect of thunderstorm is poor.The reliability of empirical relation need to be improved by a large number of individual cases.