Theory And Application Of Multi-scale Spatial Model

Multi-scale geographically weighted regression(MGWR)model is a relatively new s-patial analysis method.By running the regression relationship on different spatial scales,it can accurately capture the spatial heterogeneity of the explanatory variable’s influence on the response variable at different spatial scales.Therefore,more and more researches and applications are obtained.This article mainly conducts research from the following two aspects:First,based on spatial autocorrelation and the basic theory of the MGWR model,taking the urban air quality monitoring data of the Yellow River Basin in 2017 as the research object,firstly,using spatial autocorrelation analysis,it is concluded that the air quality in most areas of the upper reaches of the Yellow River Basin is "low-low" aggregation,and in the middle and lower reaches of the Yellow River basin in many parts of the air quality as"high-high" aggregation.Secondly,the characteristics of the temporal and spatial changes of air quality are given,and then the multi-scale geographic weighted regression(MGWR)model is used to analyze the air quality.The spatial heterogeneity of the effects of factors on air quality at different scales.The results show:(1)In 2017,the air quality index(AQI)of most areas in the middle and lower reaches of the Yellow River Basin was at a light pollution level,and the seasonal AQI was high in winter and low in summer.The monthly AQI has a "U"-shaped distribution,and its turning point appears in August;(2)The proportion of the secondary industry,the proportion of the urban population,the total population,rainfall and temperature have a significant impact on air quality.Among them,the proportion of the secondary industry and the proportion of urban population affect air quality globally.The total population affects air quality at a moderately heterogeneous level,and rainfall and temperature affect air quality at a higher heterogeneous level;(3)The proportion of the secondary industry,the proportion of urban population and the total population have positive effects on air quality,while the rainfall and temperature have negative effects on air quality.The total population has a greater impact on the middle reaches of the Yellow River,the rainfall has a greater impact on the upper reaches of the Yellow River,and the temperature has a greater impact on the middle and lower reaches of the Yellow River.Second,in order to make up for the lack of a single bandwidth for all variable coef-ficients of the mixed GWR model,a mixed multi-scale geographically weighted regression model was constructed,and based on the 2012 hand-foot-mouth disease data of 102 banner counties in the Inner Mongolia Autonomous Region,using the bootstrap method identifies the constant and variable coefficients in the model.The results show that the monthly aver-age temperature(AT),the monthly average relative humidity(AH),and the monthly average air pressure(AP)is a local variable,while monthly average wind speed((AW))is a global variable,and cumulative rainfall((AR))is an irrelevant variable.Further construct mixed multi-scale geographic weighting regression model to explore the spatial heterogeneity of the incidence of hand-foot-mouth disease(HFMD)on different scales of meteorological fac-tors.The results show that AT has a positive correlation with the incidence of HFMD in all regions of Inner Mongolia.AH has a positive correlation with the incidence of HFMD in most areas of Inner Mongolia.AP has both a positive correlation and a negative correlation with the incidence of HFMD in Inner Mongolia.