Evaluation Of Forward Models For GNSS Radio Occultation Data Processing And Assimilation

The forward model(FM)is important in Radio occultation(RO)data processing and RO data assimilation,during which FM is performed to obtain bending angle(BA)from refractivity(N).Therefore,the accuracy and precision of FM has significant impact on the accuracy and precision of two processes.The accuracy of FM is related to not only the refractivity profiles,but also FM methods,which includes Abel integral algorithms(direct,exp,exp＿T)and methods of interpolating refractivity during the integral(log-cubic spline and log-linear).Experiment 1 evaluated those forward models using the experimental value(the forward modeled BA from ECMWF analysis)and the true value(BA of FY3 D RO data).Results demonstrated that the exp with log-cubic spline(log-cubic)interpolation is the most accurate FM.By contrast,the error of direct reaches 10%,and the improvement of exp＿T to exp is limited.Experiment 2 studies the exact errors of an FM on different model levels,by simulating inputs(refractivity profiles)on model levels of three widely used analyses,including ECMWF 4Dvar analysis,final operational global analysis data(FNL),and ERA5.Results suggests that based on FM methods of exp plus log-cubic interpolation,BA on model level of ECMWF 4Dvar has the highest accuracy(RD are0.5%,4% and 1.8% between 0–35 km,35–58 km and 58–80 km.This paper established a way to better understand the FM during RO data processing and assimilation.The simulated errors on model levels of three analyses can be a helpful FM error reference.