Ensemble-Based Adaptive Observations for Improving Sea Fog Prediction in Coastal Regions around the Bohai Sea: Case Study with Cold-Front Synoptic Pattern

This study explored the observation strategy and effectiveness of synoptic-scale adaptive observations for improving sea fog prediction in coastal regions around the Bohai Sea based on a poorly predicted fog event with cold-front synoptic pattern (CFSP). An ensemble Kalman filter data assimilation system for the Weather Research and Forecasting model was adopted with ensemble sensitivity analysis (ESA). By comparing observation impacts (estimated from a 40-member ensemble with ESA) among different meteorological observation variables and pressure levels, the temperature at 850 hPa and surface layer (850hPa-and-surface temperature) was selected as the target observation type. Additionally, the area with large observation impacts for this observation type was predicted in the transition region of the surface low–high system. This area developed southward with the low and moved eastward with the low–high system, which could be explained by the main features of CFSP. Moreover, both experiments assimilating synthetic and real observations showed that assimilating 850hPa-and-surface temperature observations generally yielded better fog coverage forecasts in areas with greater observation impacts than areas with smaller impacts. However, the effectiveness of adaptive observations was reduced when real observations rather than synthetic observations were assimilated, which is possibly due to factors such as observation and model errors. The main conclusions above were verified by another typical fog event with CFSP characteristics. Results of this study highlight the importance of improved initial conditions in the transition region of the low–high system for improving fog prediction and provide scientific guidance for implementing an observation network for fog forecasting over the Bohai Sea.