The Establishment Of East Asian Regional Atmospheric Reanalysis System And Its Preliminary Analysis

Based on a national key research and development project,this paper firstly investigated the main data assimilation systems worldwide and the corresponding processes in data assimilation and data processing.In the view of the current lack of regional high-resolution reanalysis over East Asia,the East Asian Reanalysis System(EARS)was set up.Then,a one-year experimental reanalysis was generated and evaluated.Finally,this paper made a summary and proposed some suggestions for further improvement.The high-resolution regional reanalysis system,EARS,was developed based on the data assimilation system,GSI(Gridpoint Statistical Interpolation),from NOAA(National Oceanic and Atmospheric Administration)and mesoscale forecasting model,WRF(Weather Research and Forecasting Model).The EARS also utilized UPP(Unified Post Processor)for data post-processing and MET(Model Evaluation Tools)after postprocessing.The EARS used ERA-Interim from ECMWF(European Centre for MediumRange Weather Forecasts)as the initial and boundary conditions,and the conventional/retrieval and satellite observational data came from NCEP's(National Centers for Environmental Prediction of the NOAA)GDAS(Global Data Assimilation System).These satellite observations include the ones from AIRS(Atmospheric Infrared Sounder),AMSU-A(Advanced Microwave Sounding Unit-A),MHS(Microwave Humidity Sounder)and HIRS4(High-resolution Infrared Radiation Sounder 4.)Other data involved in the following evaluation were MICAPS from China Meteorological Administration,ERA5 reanalysis from ECMWF and precipitation data TRMM(Tropical Rainfall Measuring Mission.)The there-dimensional variational data assimilation and partial cycling scheme were implemented in the EARS.In the data assimilation system,conventional observations had passed quality control(QC)in the GDAS system,but the QC of satellite observations was applied within the GSI.The QC of satellite data for specific instruments was implemented based on the instrument errors,simulation errors due to cloud and precipitation,simulation errors caused by surface emissivity and errors introduced during the data processing.The evaluation of the one-year experimental data showed that the EARS could represent the atmospheric condition well.Below 250 h Pa,the performance of EARS was closed to the ERA5 or even better.In the estimation of 6-hour precipitation forecasts,the precipitation distribution and the centers of intensive rain of EARS fitted the ones in TRMM and its precipitation series was reasonable.However,above 250 h Pa,the simulation of EARS declined significantly.Besides,there was a disagreement between EARS and ERA5 over the Tibetan Plateau and Indian monsoon areas.On account of the problems above,based on the original system,this paper finally proposed to improve the data assimilation scheme,increase more observational data and introduce large-scale circulation information for further improvement.