Research On Soil Moisture Spatial Distribution Analysis And Simulation

Taking Dongying, Dezhou, and Binzhou as the study area, based on the measured to obtain the ground soil moisture data, the spatial variability characteristics of GS + software analysis, select the most appropriate half variance function parameters. Using statistical analysis, GIS interpolation, buffer analysis methods, analysis soil moisture condition and spatial variation law of the study area. Use common interpolation model and HASM interpolation model, finding out its advantages and disadvantages of the models and the applied range. To improve the accuracy of HASM model interpolation results, let me increase the number of sampling points. Field hyperspectral data and sample point data is adopted to establish the high spectral estimation model of soil moisture, realizing quickly get soil moisture of the key point. Soil moisture value interpolated as a "space expansion" after the point data,establish the relationship with the remote sensing satellite band reflectance model, realizing large soil humidity value. Soil moisture belongs to medium spatial variability.The field hyperspectral soil forecast model is established. Decision coefficient of the model is 0.85. The relationship between soil moisture value in the room and its value in the field model.The Decision coefficient of the model is 0.97. The soil moisture inversion model based on a variety of spectral paramenters is established.Decision coefficient of the model is 0.65.This paper studies the main contents include:(1) According to the measured the soil moisture data, use statistical analysis and GIS interpolation, buffer analysis methods, by half the variation function parameters selection experiments, analysis soil moisture spatial distribution Dezhou, Binzhou and Dongying in shandong area. Selecting the optimal spatial variation parameter, finding the most suitable interpolation model, through cross validation, finding out the interpolation methods advantages and disadvantages, to improve the interpolation precision of the results and analysis the spatial distribution characteristics of soil moisture in the study area.(2) To solve the problem of insufficient number of sampling points, adopting HASM model interpolation method to get high precision, in the study area have higher precision of soil moisture spatial distribution simulation with the six sets of data interpolation experiments. Increasing the number of sampling points, to simulate the soil moisture spatial distribution with the bigger spatial variability in the study area.(3) Using the correlation analysis and multivariate stepwise regression analysis, the measured soil moisture data in combination with field hyperspectral data to establish models to predict the soil moisture, obtain the key soil moisture.(4)Set up the relationship between hyperspectral data and soil moisture model. Choosing soil moisture sensitive wavelengths, using multiple linear regression method, set up the interpolation as a "space expansion" after point data of hyperspectral data with the soil moisture model, for large area soil moisture.