Impact of the Sequential Bias Correction Scheme on the CMA-MESO Numerical Weather Prediction Model

Systematic bias is a type of model errors that can affect the accuracy of data assimilation and forecasting, and it must be addressed. An online bias correction scheme, called the sequential bias correction scheme (SBCS), was developed using the 6-hour average bias to correct the systematic bias during model integration. The primary purpose of this study is to investigate the impact of the SBCS in the high-resolution China Meteorological Administration Meso-scale (CMA-MESO) numerical weather prediction (NWP) model to reduce the systematic bias, and to improve the data assimilation and forecast results via this method. The SBCS is improved and applied to the CMA-MESO 3 km model in this study. Four-week sequential data assimilation and forecast experiments driven by rapid update and cycling (RUC) were conducted for the period from May 2 to 29, 2022. In terms of the characteristics of systematic bias, both the background and analysis show diurnal bias, and these large biases are affected by complex underlying surfaces (e.g., oceans, coasts, and mountains). After the application of the SBCS, the results of the data assimilation show that the SBCS can reduce the systematic bias of the background and has a neutral to slightly positive result for the analysis fields. In addition, the SBCS can reduce forecast errors and improve forecast results, especially for surface variables. The above results indicate that this scheme has good prospects for high-resolution regional NWP models.