Multi-model Ensemble Forecast Based On Atmospheric Numerical Simulation Of China's Greater Bay Area And WRF Model

People's daily travel,agricultural and sideline harvests and industrial production are closely related to the weather forecast,but there are uncertainties in the forecast of meteorological elements such as precipitation and temperature.With the continuous development of forecasting systems and forecasting techniques,the uncertainty of weather forecasting has decreased.The collection system,which combines the uncertainty of multi-mode,has attracted more and more attention and is also used in actual business.Based on the European Medium-Range Weather Forecast Center(ECMWF),the National Center for Environmental Prediction(NCEP),the Japan Meteorological Agency(JMA)and the China Meteorological Administration(CMA),the four numerical forecasting centers report daily 2m temperature data at 12 o'clock.Using a variety of interpolation methods,and performing downscaling and ensemble prediction,a downscaled ensemble prediction model is established.The forecasting effect of the model was evaluated through a case study of Guangdong,Hong Kong and Macau Bay Area.This article mainly completes the following work:(1)Firstly,the interpolation test of the single center mode is performed on four different interpolation methods.After comparison,it shows that Kriging's performance is the best of the four interpolation methods.And with the extension of the forecasting time,although the error of the four interpolation methods will increase,the advantage of Kriging interpolation is still obvious.(2)The effect of direct interpolation prediction of each single center mode is effectively improved by statistical downscaling.As the prediction time increases,even within the longest prediction time,the root mean square error after downscaling is lower than the root mean square error of direct interpolation.(3)Using the superset method,multi-mode integration of the single-center mode downscaling prediction results is performed.The results show that multi-mode integration significantly improves the prediction effect.(4)Taking Guangdong,Hong Kong and Macau Bay Area as an example,using WRF(weather research and forecasting)mode combined with dynamic downscaling and statistical downscaling to simulate the temperature of Guangdong,Hong Kong and Macau Bay Area,multi-mode integration has improved the Guangdong,Hong Kong and Macau Bay Area The accuracy of the ground temperature forecast.