Impacts of High-Resolution Land Surface Data on Meteorological Field Simulation over Hainan

Hainan province which located in the southernmost part of China is an important special economic zone and experimental zone of free trade.However， the studies of underlying surface impact on weather prediction in Hainan is fewer than other places.To understand the impact of underlying surface on numerical weather prediction， the WRF （Weather Research and Forecasting） model is used to simulate the meteorological field over Hainan in 2013.The comparison between default land surface data （land use， vegetation fraction， terrain， and soil type） and new high-resolution land surface data， i.e.Moderate Resolution Imaging Spectroradiometer （MODIS） land use data in 2013 processed based on IGBP （International geosphere biosphere program） classification standard， vegetation fraction data calculated based on MODIS normalized difference vegetation index （NDVI） in 2013， Shuttle Radar Topography Mission 3 （SRTM3） terrain data and Harmonized World Soil Database （HWSD） soil type data is investigated through numerical sensitivity tests to analyze the impact on meteorological field simulations.The study is divided into four parts.The first part is the introduction； The second part mainly introduces the research area， data sources， model design and evaluation methods； The third part compares the simulation results of using new high-resolution land surface data with the WRF default land surface data and evaluates the simulation results； Then the fourth part is about the result and discussion.The result indicates that using default or new high-resolution land surface data can both simulate the temporal-spatial distribution of regional meteorological fields.Using the new high-resolution land surface data can improve the simulated results of 2 m temperature and 2 m relative humidity， with the Root Mean Square Error （RMSE） reduced by 7.2% （6.5%） and 6.1%（7.7%） in spring （summer）， and the Hit Rate （HR） increased by 3.7%（2.8%） and 3.2%（2.9%） in spring （summer）.The impact of high-resolution land surface on 10 m wind speed simulation seems smaller than that on 2 m temperature and 2 m relative humidity.10 m wind speed error reduces slightly after using high-resolution land surface data， with the RMSE reduced by 3.8%（4.5%） and HR increased by 2.1%（2.9%） in spring （summer）.Compared to other underlying surface data， land use data has more significant impact on near-surface meteorological fields simulation.In general， using the high-resolution land surface data in Hainan can improve the accuracy of WRF model and reduce the simulation error.This study can deepen the understanding of numerical forecast products for forecasters， improve the accuracy of weather forecast， and provide theoretical basis for Seismic Prevention and Disaster Mitigation.