Spatial Autocorrelation Analysis Of Land Use Pattern Based On Different Weight Matrices

The land use pattern is a natural economic complex formed by the combination of natural factors and social and economic factors,and its change is a fairly complex dynamic evolution process.Existing conventional models for studying the spatial and temporal evolution of land use patterns are basically based on the data itself being normally distributed and independent,ignoring the spatial correlation of the data in the spatial distribution.Therefore,spatial autocorrelation analysis is introduced into the spatial and temporal evolution of land use pattern,which has become a hot issue for scholars.Before conducting spatial autocorrelation analysis of land use patterns,the first task is to define the adjacency relationship between various spatial phenomena,d etermine the spatial weight matrix,and then conduct spatial self-phase analysis on this basis,so that the setting of the spatial weight matrix becomes the spatial autocorrelation A key point of research.At present,the spatial autocorrelation analysis of land use patterns and the establishment of regression models are mostly based on a spatial weight matrix,ignoring the spatial heterogeneity of land use patterns on different spatial weight matrices.Therefore,it is necessary to study the spatial autocorrelation of land use patterns with different weight matrices.This article takes the central Yunnan urban agglomeration as the research area,based on the basic land use data and socioeconomic data of the central Yunnan urban agglomeration,and comprehens ively uses multidisciplinary theories such as ecological environment,geography,geographic information science,geostatistics,and spatial autocorrelation theory,Extract and classify the land use change data in 2015,establish a land use database for th e central Yunnan urban agglomeration,build a spatial weight matrix based on the proximity and distance rules,and analyze the land use degree,land use types and driving factors of the central Yunnan urban agglomeration Spatial autocorrelation on the weig ht matrix reveals the coupling characteristics and spatial heterogeneity of its land use spatial pattern and natural social economy;construct spatial autoregressive models of different land use types on different weight matrices to quantitatively evaluate each spatial self The ability of related models to explain the land use change reveals the quantitative relationship between the land use pattern and its driving factors;optimize the land use pattern of the central Yunnan urban agglomeration to provide a theoretical basis for later land use policy reform,structural adjustment,and regional planning.After comprehensive research and analysis,the following conclusions are drawn:(1)The land use degree,land use types and driving factors of the central Yu nnan urban agglomeration have spatial autocorrelation,with certain spatial differentiation rules,and the spatial autocorrelation is affected by different weight matrices.Moran's obtained on different weight matrices I index is different,showing a certa in degree of convergence.The spatial autocorrelation with the weight order increases,the threshold distance and K value increase,its spatial autocorrelation is weakening,but because of its different spatial distribution,the sensitivity to different we ight matrices is different,so the rate of decrease Certain differences.(2)Due to the spatial autocorrelation of model linear regression model residuals,the spatial autoregressive model is introduced.The study found that the spatial error model is more sensitive to the dynamic changes of the driving factors.There is no spatial autocorrelation of the model residuals.The model has the best fitting degree and the strongest explanatory ability,which can better characterize the central Yunnan urban agglomeration.The spatial and temporal evolution of land use pattern.(3)Select the spatial error model to simulate the spatial and temporal evolution of the land use pattern in the central Yunnan urban agglomeration,and find that the spatial and temporal evo lution of different land use types has different driving factors.The main driving factors affecting the spatial and temporal evolution of the cultivated land pattern in the study area are slope,aspect,topography,distance from town,distance from villag e,distance from highway,distance from rural road,distance from river,distance from lake,total population,agriculture Population,GDP,GDP per capita,net income per farmer per capita,cultivated land per capita and rainfall;the main driving factors of grassland are slope,surface roughness,terrain relief,distance from railway,distance from highway,population density and average annual temperature;forest land the main driving factors are slope,aspect,terrain relief,distance from town,distance from village,distance from highway,distance from lake,total population,agricultural population,population density,GDP,GDP per capita,net income per farmer,rainfall Volume and annual average temperature;the main driving factors for construction l and are slope,surface roughness,topography,distance from town,distance from highway,distance from river,agricultural population,population density,GDP,GDP per capita,net income per farmer,Rainfall and annual average temperature;the main drivin g factors of water use are elevation,slope,slope,Surface roughness,terrain relief,distance from town,distance from river,total population,net income per capita of farmers,rainfall and average annual temperature;the main driving factors of other land use are elevation,surface roughness,distance from lake,distance Rural road distance,GDP per capita,net income per farmer,rainfall and average annual temperature.It can be seen that the topographical factors and socioeconomic factors have a grea ter influence on the spatial and temporal evolution of the land use pattern of the central Yunnan urban agglomeration.The innovation of this paper: The land use pattern has spatial autocorrelation.In the past,the spatial autocorrelation analysis of the land use pattern was mostly based on a weight matrix,ignoring the spatial heterogeneity of the land use pattern on different spatial weight matrices.Based on a variety of weight matrices,this paper carried out spatial autocorrelation analysis of the lan d use pattern in the central Yunnan urban agglomeration,and linked the spatial autocorrelation analysis with the spatial weight matrix,taking into account the impact of the spatial weight matrix on spatial autocorrelation.The spatial autoregressive mode l was introduced.After simulation verification,it was found that the spatial autocorrelation of the residuals of the spatial error model is not significant,and the model accuracy is high.It can provide method reference and research ideas for land use p attern simulation research of similar land use spatial layout areas.