Performance Verification of Medium-Range Forecasts bzeny T639, ECMWF and Japan Models from December 2017 to February 2018
The performance of medium-range forecasts by the models of T639, ECMWF and Japan (JP) from December 2017 to February 2018 are verified and compared. The results show that all the three models can predict the variation and adjustment of the atmospheric circulation over Asian middle and high latitudes well, of which ECMWF model performs the best. The three models perform also well in predicting the transitions of temperature at 850 hPa, and they all have smaller biases in temperature forecasts for southern China than for northern China. The T639 model has an overall low biases in the temperature forecast of northern China while the JP model has higher temperature forecasts for southern China. The ECMWF forecast has the best effect comprehensively. The T639 model forecasts the intensity of cold high-pressure center better than the ECMWF and JP models. The ECMWF model predicts the location of the high-pressure center weakly and by east, T639 and JP models also have obvious biases in the location forecast of high-pressure center.