Spatial Prediction Effects And Appropriate Sampling Quantity Of Soil Organic Matter Based On Soil-Terrain Relationship

The precondition of precision agriculture and environmental monitoring management is the accurate basic data of soil attributes.But the traditional soil mapping contains limited soil information,which mainly comes from the collection and measurement of field soil data by experienced soil experts.Through the collection of soil data including soil nutrients,soil types,soil forming factors,etc.,the soil terrain model is constructed by using GIS and terrain factors,and applied to soil information management and the acquisition of soil information is expressed by geospatial analysis.On the basis of geostatistical analyst,soil terrain relationship and geographical weighted regression Kriging model,this paper explores the extraction and expression of soil collaborative environmental elements,the reasonable sampling of soil in the system grid,the expression of terrain units,the determination of the best statistical unit of terrain fluctuation,the modeling of soil terrain,the prediction and accuracy of soil organic matter attribute information.Taking Wanrong County of Shanxi Province as an example,based on the relationship between soil and topography,this paper studies the reasonable sampling layout and optimization scheme of sampling points,determines the reasonable sampling points and spatial layout and predicts the spatial distribution of soil organic matter on the basis of terrain unit division and different grid scale sampling.The main conclusions are as follows:(1)Taking DEM as data source and 10 topographic factors,such as elevation,slope,aspect,vegetation coverage,relief amplitude,profile curvature,topographic wetness index,stream power index,topographic position index,relative position index,are calculated and extracted.Firstly,taking the relief amplitude as an example,using the Arc GIS neighborhood analysis method and the average point variation analysis method to extract the relief amplitude of the study area,the optimal statistical unit size of the relief amplitude is 10×10 pixels.(2)Through the correlation analysis of organic matter and terrain factors,the matrix of soil terrain correlation coefficient is obtained,and it is found that there is a synergistic relationship between soil and terrain in the study area.The Pearson correlation coefficient analysis and CCA sorting analysis were carried out to select the relevant factors under different sampling layout,so as to carry out the terrain unit according to the constructed soil terrain relationship and relevant factor classification standards The research shows that the accuracy of the sampling points based on the terrain unit zoning determined by a large number of terrain factors is high.(3)In this paper,among the four sampling schemes based on grid division which are close to the geometric center of grid,the layout scheme with 1720 sampling points distributed on500×500m grid scale is similar to the layout range of 1656 sampling points in the division of terrain statistical unit,so the G4 scale is more accurate than other grid scale sampling methods.(4)By analyzing the residual and interpolation results of each terrain unit and combining with the weighted Kriging model of geographical regression,the spatial distribution of organic matter containing the influence of terrain factors and the distribution of predicted influence are obtained.Through comparative analysis,it is shown that the prediction accuracy of soil organic matter in the terrain unit with large topographic relief is roughly consistent with that of the global prediction,while that in the flat terrain unit(such as wet flat land)is higher than that of the global prediction,and the content of organic matter is relatively high,and the prediction value range is roughly consistent with the measured value range(the prediction value is 8.79-12.66g/kg,the measured value is 8.67-14.82g/kg)The spatial prediction based on terrain unit can accurately and effectively express the spatial structure and heterogeneity of soil organic matter in each terrain area.Based on the system grid multi-scale sampling,the best statistical unit sampling of terrain undulation,the statistical characteristics of soil organic matter content and terrain factors of different point sets under the terrain unit sampling mode are similar,but the spatial information expressed by different point sets under each sampling method is more significant,and the expression of local spatial information is also more obvious.Under different sampling methods,the spatial structure of soil organic matter content is irregular with the change of the number of sample subset.Considering the accuracy of spatial variability evaluation of soil nutrients and the results of cultivated land quality evaluation,the number and spatial prediction distribution of different points were discussed,which laid the foundation for obtaining spatial information with high spatial-temporal resolution and updating soil attributes.