Research on Temperature and Precipitation Correction Algorithms Based on Multi-Model Ensemble Methods for Qianxinan Prefecture of Guizhou Province
This study investigates localized objective forecasting algorithms for intelligent grid systems aiming at improving the accuracy of temperature and precipitation forecasts from numerical models in the Qianxinan Prefecture of Guizhou Province.Made corrections to temperature and precipitation forecasts from numerical models(ECMWF,GERMANY,and JAPAN)at both station and grid levels for the Qianxinan Prefecture and the adjacent areas by utilizing two multi-model ensemble methods(bias correction ensemble averaging and weighted bias correction),dynamic error weighting coefficient,separate modeling for high and low temperatures and precipitation classification statistics.Moreover,a comparative analysis of forecast errors,accuracy and skills is conducted before and after the corrections using various evaluation metrics.The results indicate that:(1)The bias correction forecasts can significantly reduce the errors of numerical model forecasts,with 2 m temperature and 12 h accumulated precipitation errors decreasing to below 2.5℃ and 2.5 mm,respectively.(2)The weighted bias correction based on dynamic errors can notably enhance the accuracy of temperature and precipitation forecasts.The separate modeling for high and low temperatures,along with precipitation classification,further can improve forecast effects.Specifically,relative to the best-performing model ECMWF,the accuracy of 2 m temperature forecasts at grid and station levels is improved by 41.48％and 12.47％,respectively.The accuracy of light rain forecasts at stations is 23.9％higher than that of the best-scoring model JAPAN.Moreover,the accuracy of minimum temperature and light rain forecasts exceeds the forecasts of local forecasters by 1.44％and 27.8％,respectively.In summary,the multi-model ensemble bias correction forecasting model by means of separate modeling for high and low temperatures and graded precipitation statistics can effectively reduce forecast errors of 2 m temperature and 12 h accumulated precipitation,significantly improving the forecast accuracy.The improvement in 2 m temperature grid forecasts and precipitation station corrections demonstrate considerable effectiveness,with light rain forecasts outperforming historical forecasts made by local forecasters.These findings are of great reference value for operational forecasting.
multi-model ensembleBREMWBREMseparate modeling for high and low temperaturesdynamic error weighting
NETL
NSTL
万方数据
