A Study On The Relationships Between Landscape Patterns And Water Quality Across Multiple-scales In Basin Based On Remote Sensing Technology

The issue of water environment in basin has always been the focus of ecological environmental research.Taking full advantage of the remote sensing technology(multi spatial resolution,multi spectral resolution and multi time resolution),the parameters of the heterogeneous underlying surface in the basin are obtained.Combined with fixed point observation and experimental analysis,quantitative analysis theory and technical support can be provided for the optimal allocation of landscape pattern of non-point source pollution in watershed.Taking the Jiulong River Basin as the research area,this paper calculated landscape pattern indices on the basis of source and sink landscape information,which reflect the features of landscape shape,connectivity and degree of polymerization.The landscape index system with low redundancy is constructed by the Spearman rank analysis method.This paper developed a landscape spatial load comparison index based on minimum hydrological response unit(HRULI)with ecological significance and multi-scale adaptability.Finally,by selecting the buffers with different distances on both sides of the riparian zone and catchment areas as the complementary research scales,this article constructs multiple linear regression model of landscape pattern index and water quality index in different time and space scales,and then quantificationally calculates and analyses the relationship between landscape pattern indices and water quality in the basin.The main conclusions and innovations of this paper are as follows:1?The concept of HRULCI and its calculation method are put forward.Compared with the traditional landscape pattern index,the landscape spatial load comparison index based on the minimum hydrological response unit can effectively reflect the difference in the effect of different source-sink types on non-point source pollutants in water quality.The significance of model results of different landscape pattern indices for multivariate linear regression models also indicate that HRULCI is an important indicator that influences the relationship between the source-sink landscape pattern and water quality indicators.2?The scale effect of landscape source sink pattern and hydrological index of watershed non-point source pollution is analyzed.The multiple linear regression model of the same water quality index between different spatial scales and different landscape pattern indices was constructed,and the significant degree of the water quality effect of different landscape pattern indices under different spatial scales was quantitatively analyzed.3?Taking pH and DO as an example,by analyzing the change trend of the model's significance,this paper put forward the different suitable scales in the study of the relationship between the source-sink landscape pattern of the non-point source pollution and the water quality indexes.4?The regression model of CODMn and landscape pattern index shows that,compared to the buffer scale,the catchment is also an important research scale when study on the relationship between landscape pattern and water quality in basin.The research results deepen the understanding of the impact of the landscape pattern on the water quality indicators under the multi space-time scale,and it also have important guiding significance for the treatment of non-point source pollution,the improvement and the optimization of the landscape pattern spatial structure.