An Ensemble Analysis Field Of Temperature And Humidity Wind And Precipitation Based On Error Perturbation

The high-precision near surface variable analysis still can not fully describe the real situation of the atmosphere.The precipitation analysis blending by QPE and observation supplements the precipitation information in the non station area,and the station observation ensures the accuracy of the data.Simulating the error between precipitation analysis and observation and carrying out random perturbation can reflect the uncertainty of analysis error and obtain a rainfall ensemble analysis that can better capture the true value.Similarly,the background error of near surface temperature,humidity and wind in the assimilation process can also be described by random disturbance,a high-resolution analysis ensemble of surface temperature,humidity and wind can be constructed.Combining the analysis error with Gaussian random vector can reduce the absolute error and root mean square error,and improve the TS score of each magnitude of precipitation.Its spread can well reflect the uncertainty of precipitation estimation error,that is,any member of precipitation ensemble analysis can replace the original analysis field.The ensemble also has high reliability,rationality and high resolution for precipitation of different orders of magnitude.Simulating the background error characteristics in the analysis' calculation process,the ensemble analysis of ground 2m temperature,relative humidity,specific humidity and ground10 m wind speed components is constructed.The spread of the four ensemble is close to the root mean square error between the ensemble mean and the verification data,which shows that the ensemble well reflects the uncertainty of the background error and has high reliability.These ensemble can distinguish the occurrence of each threshold event well.For the threshold with enough valid samples,the probability of the occurrence of the ensemble judgment event is very close to the actual frequency.The ensemble generated by properly perturbed the error has good resolution and reliability.Such ensemble can further improve the quality of ground variable analysis,and provide a valuable surface initial condition data set for short-term and immediate weather forecast.