An advection fog process during 4‒6 June 2019 over Zhejiang coastal area is simulated using the WRF model driven by the ERA5 data,and the effects of different boundary layer parameterization schemes(YSU,QNSE,ACM2,MYJ)on the simulation results are studied in comparison with observed meteorological data.The results show that all boundary layer parameterization schemes perform well on simulating surface air temperature and relative humidity,but all of them overestimate surface wind speed.The ACM2 and YSU schemes overall perform better than the QNSE and MYJ schemes on the relative humidity and surface wind speed simulation.The ACM2schemeexhibitsabettercapacitythantheYSUschemeontherelativehumiditypredictionstabilityand error test.The ACM2 scheme generates reasonable temporal evolution and spatial distribution of the fog.The YSU scheme overestimates the fog zone and generates occasional fail forecast.The QNSE and MYJ schemes produce insufficient accuracy in the spatial distribution of sea‒land fog zone.Boundary layer height directly reflects the intensity of turbulent flows,and thus plays a critical role in the evolution of advection fogs.Compared with the YSU and ACM2 schemes,the QNSE and MYJ schemes underestimate the turbulent mixing intensity in coastal areas,resulting in occasional fail forecast of coastal fog zone.