The wind field in the atmospheric boundary layer(PBL)is the most important and fundamental meteorological element in atmospheric science.The accuracy of its simulation directly affects many fields such as climate prediction,weather forecasting,pollution forecasting,wind energy utilization,and wind engineering early warning.On the complex underlying surface,there is a non-uniformity of topographic and thermal forcing,resulting in an unusually complex atmospheric boundary layer process.Currently,there are still great uncertainties in the elaborate simulation of the wind field structure in the atmospheric boundary layer.The selection and optimization of boundary layer parameterization schemes directly affect the simulation results of the boundary layer wind field.Using high-resolution three-dimensional Doppler wind lidar(DWL)data obtained by the research team(observation stations located from east to west in Rizhao,Shandong,on the west coast of the Yellow Sea,Beijing,a megacity,and Baotou,Inner Mongolia,on the Mongolian mainland),this study focused on the three-dimensional wind field structure of the typical underlying surface in northern China,including the prevailing coastal land and sea breeze,the Beijing-Tianjin-Hebei mountainous plain wind,and sandstorms with obvious regional characteristics;We systematically evaluated the differences in dynamic,thermal,and wind fields generated by three types of typical wind fields in different boundary layer parameterization schemes;Based on the PBL turbulent momentum equation and assumptions,the reasons for the uncertainty of simulation results caused by various parameters in the simulation of vertical wind field structures in the boundary layer are systematically analyzed.The study found that the accuracy of the three-dimensional wind field simulation of the mesoscale meteorological model WRF is highly sensitive to the atmospheric boundary layer scheme in different regions,three type locality are illustrated as examples.In the west bank of the Yellow Sea,the three schemes GBM,MYNN2,and YSU are suitable for simulating the sea land breeze,which shows that the simulation biases at different heights in the vertical direction are small,and can better simulate the occurrence time,duration,and development height of the sea land breeze at this observation point.Among them,the GBM scheme considers the entrainment effect,improves the radiation process between the ground and the atmosphere,makes the turbulent diffusion generated by buoyancy more accurate,and ensures the accuracy of the calculation of turbulent diffusion caused by changes in temperature and humidity in the sea and land in coastal areas;The YSU scheme introduces a local gradient correction term,which enables non local geothermal convection to bring hot air from the surface to the upper middle of the boundary layer,increasing the turbulent mixing induced by thermal forcing during the simulation of sea and land breezes;The MYNN2 scheme clarifies the relationship between the length scale and buoyancy,turbulence scale,and near surface in the turbulent kinetic energy equation,and introduces a set of closure parameters representing buoyancy and shear effects.However,the disadvantage is that the impact on the underlying surface is not sufficiently considered,and the entrainment heat flux at the top of the convective boundary layer is underestimated.In Beijing,the simulation of eight boundary layer schemes found that the vertical mixing intensity of mountain plain wind was insufficient,the simulated virtual potential temperature and vertical wind field were not large enough,the model could not capture the position of plain wind vortices,and the wind direction deviation in the vortex layer was large.The wind speed simulation schemes with a height of more than 500 meters include ACM2,YSU,and Shin-Hong.These three schemes have nonlocal upward and local downward mixing schemes,which have certain advantages in vertical mixing strength simulation.Among them,ACM2 can not only simulate the transport process generated by large-scale turbulent vortices under convective conditions,but also reflect the sub grid small-scale turbulent mixing process;The Shin-Hong scheme calculates the nonlocal transport of the updraft and the local transport of the remaining small-scale vortices.At the altitude of 200~500 m,the wind field is stable,and the MYJ scheme has the advantage of stable boundary layer simulation at this altitude.The simulated wind speed and direction are both good.In the Baotou area of Inner Mongolia,through the simulation of dust storms,we found that non adiabatic heating plays an important role in the maintenance and development of dust storms.It is difficult for numerical simulation to capture the sanding process with rapidly developing and changing sandstorms.The wind shear index is very important because it represents the dynamic factors in the model.PBL schemes such as MYNN2,ACM2,and GBM have good wind shear simulation results for dust storms due to the consideration of buoyancy.In addition,the ACM2 scheme can better reproduce the basic change characteristics of meteorological elements during the dust storm process.The mesoscale meteorological model WRF has strong sensitivity to different atmospheric boundary layer schemes,largely due to its different vertical mixing intensities.The local thermodynamic mechanism scheme of the boundary layer is of great significance in the simulation of the sea land breeze circulation;The best performing schemes in mountain plain wind simulation are local and non local mixing schemes that can improve vertical mixing;The boundary layer scheme considering buoyancy mechanism plays an important role in sand storm simulation. |