Numerical Simulation and Analysis of the Persistent Sea Fog in the Qiongzhou Strait in February 2018

During the Spring Festival in February 2018, persistent heavy fog was present in the Qiongzhou Strait, causing a large number of ships to stop sailing. Synoptic causes for the heavy fog event from February 18 to 20 in 2018 were analyzed by using Himawari-8-derived sea fog products, observed visibility data gathered from coastal stations over Qiongzhou Strait, and Final Operational Global Analysis (FNL) from the National Centers for Environment Prediction. Based on the high-resolution numerical model Global and Regional Assimilation and Prediction System (CMA-MESO), sensitive experiments were conducted based on a comparison of multiple boundary layer schemes, vertical resolutions of the model, and algorithms related to the visibility of sea fog. It was found that the offshore sea temperature in South China during the heavy fog event was lower than average. Affected by the supplemented weak cold air via the cold high pressure southward, the warm and humid air in the east flowed through the cold sea surface and condensed quickly. The results of contrast experiments demonstrated that the prediction accuracy of visibility could be significantly improved by employing the Yonsei University (YSU) boundary layer scheme and the sea fog diagnosis scheme of the National Oceanic and Atmospheric Administration Forecast Systems Laboratory (FSL) along with increasing the vertical levels of the boundary layer. Compared to the MRF (Medium Range Forecast Model) boundary layer scheme, the simulations of the spatial distribution of heavy fog and occurring time of minimum visibility are much better using the YSU boundary layer scheme. Moreover, expanding the levels of the lower layer of the numerical model was found to improve the simulation results regarding the evolution of low visibility. Through the visibility algorithms and the real comparison, based on the well-predicted moisture and temperature, the predictions of visibility using the FSL method are much more accurate.