A Study On The Temporal And Spatial Variation Of Environmental Heterogeneity In The Typical Middle And Lower Reaches Of The Yellow River

The change of environmental heterogeneity in space and time is a comprehensive reflection of the distribution of resources in space and time,which is related to the growth and development of plants,human life and production,animal protection and so on.This paper measures the change of environmental heterogeneity from two dimensions of time and space.In the space dimension,this paper compares and analyzes the environmental heterogeneity of different terrains,plains and mountains.In the same terrain,this paper establishes a regression model among environmental heterogeneity and different geographical elements.In the time dimension,this paper quantifies the texture index of remote sensing images in different periods,and takes the amount of precipitation as the time change index to measure the relationship between them.In this paper,two types of landforms are selected,each of which is selected for three different precipitation levels of remote sensing images,and each precipitation level is selected for the dry season and the rainy season,with a total amount of twelve remote sensing images.Based on these graphs,the principal component image is firstly changed,and the first principal component is used as the basis of the analysis.According to the DEM data,7 indexes are extracted as stable geographical elements,such as elevation,slope,slope direction,water flow direction,water influx,convergence index of the river basin,and topographic convergence index.Then,the correlation between them is analyzed and the variance inflation coefficient of each factor is tested.The geographical elements with high correlation are then excluded.Based on the DEM data and the first principal component remote sensing image,combined with the change of precipitation condition,the change of environmental heterogeneity in time dimension is analyzed.A linear regression model and a geographically weighted regression model are established to analyze the relationship between the stable geographical factors and the spatial distribution of environmental heterogeneity.The following conclusions are drawn through the analysis:On the change of time,the environmental heterogeneity of the plain is not related to precipitation and seasonal variation.Yet for the mountain area,the uniformity decreases with the increase of precipitation,the Entropy and Shannon index increase with the increase of precipitation,and the environmental heterogeneity of mountainous area increases with the increase of precipitation.However,under the same precipitation condition,the environmental heterogeneity of the mountain area is not related to the seasonal variation.On the spatial distribution,the heterogeneity of the mountain area is higher than that of the plain,which mainly shows that the evenness index of the mountain area is less than that of the plain,and the information entropy and Shannon index of the mountain area are higher than that of the plain.The geographically weighted regression model has a higher fitting degree than the ordinary linear regression model.After the geographically weighted regression,the amount of model interpretation of mountain area is 0.6-0.8,and the one of plain is 0.4-0.6.Obviously,the fitting degree of the geographically weighted regression model is better than that of ordinary linear regression.The P values in the model indicate a significant relationship between environmental heterogeneity and stable geographic elements(Usually P=0).