Research On Optimization Of Spatial Layout Of Surface Temperature Monitoring Station

Meteorological service is widely used in all aspects of the society.The quality of surface meteorological observation data directly affects the quality of meteorological service,and surface meteorological station is one of the main sources of surface meteorological observation data.The reasonable layout of the surface meteorological station can ensure that in the case of a certain interpolation accuracy,as far as possible to avoid wasting too much financial and material resources.Therefore,on the basis of the distribution of existing surface temperature monitoring stations,this paper conducted a series of optimization studies on the spatial layout of surface temperature monitoring stations from the perspective of using the least number of stations to retain the most temperature information.The main contents are as follows:This paper uses Moran index to analyze the spatial correlation of the surface temperature data,which lays a theoretical foundation for using Kriging interpolation method in geostatistics.Aiming at the distance between the surface temperature monitoring stations in the study area,this paper starts with the structure function and uses the structure function to estimate the relationship between the interpolation accuracy of the daily average temperature and the station spacing,and estimates the maximum allowable distance of the surface temperature monitoring stations in the study area;the distribution of existing sites leads to the problem of weak temperature monitoring ability at the edge of large-scale spatial regions,in the case of the maximum allowable distance,through geostatistical methods,the Kriging variance is used to find the area where the site is preferentially added,combined with the optimization ability of genetic algorithm based on genetic algorithm and the expansion method of Kriging’s surface temperature monitoring station;considering the increase of the density of the surface temperature monitoring stations in the small and medium-sized areas in the future,combined with the residual error caused by Kriging interpolation,a method of expanding the surface temperature monitoring stations based on the residual distribution is proposed by using the residual situation in the long-term scale of each station in the study area.Through the analysis,the fitting effect of quadratic polynomial on the structure function of Jiangsu Province and Hunan Province is the best.Finally,from the structure function,the maximum distance of surface temperature monitoring stations in Jiangsu Province and Hunan Province is calculated as 40 km and 35 km respectively.Based on genetic algorithm and Kriging’s expansion method of surface temperature monitoring stations,it can find suitable number of new stations and reasonable station network configuration for areas with weak monitoring ability,and improve the temperature monitoring ability of the whole area;based on the residual distribution,the expansion method of surface temperature monitoring station can assign the residual value to the simulation station.Through the experiment of Taizhou,Huai’an,Changsha and Yiyang,it is found that compared with the Kriging cross validation method to reduce the redundancy,the method can reserve as many simulation stations as possible in the area with high residual value,so as to improve the temperature monitoring in the area with high residual value Ability.This study will provide some suggestions for the future development of surface temperature monitoring stations.