The Application of the Adjoint Modeling System and Nonlinear Optimization Method in the Study of Predictability of the REM with Observational Data
伴随系统及非线性优化方法在REM模式可预报性研究中的实际个例应用

For a meso-scale numerical model,how to evaluate the error in the numerical model and how to evaluate the error in the initial data? In order to perform the best prediction to a given synoptic system,what kind of initial data is suited to the numerical model? Those are the key problems in the study of the predictability.Recent studies indicate that the nonlinear optimization method,which uses the numerical model's adjoint modeling system,is a useful way to evaluate the error in the model and the initial data.With the nonlinear optimization method and the adjoint modeling system of the Regional Eta-coordinate Model(REM),a study of the predictability of REM is conducted using the daily observational data in this paper.Three numerical tests are performed,the results suggest that the REM can give an acceptable forecast with the allowed forecast errors in the three synoptic tests.The first test suggests that the REM can give a satisfactory simulation just using the initial data by interpolation(such as,the optimal interpolation method) from daily station data.The second and the third tests suggest that the REM can give a satisfactory simulation by using improved initial data(via the four-dimendional variational data assimilation method).