Research On The Method Of Data Process For Soil Moisture Ground Observation

Soil moisture is a kind of important surface environmental variable,which regulates the balance of water and energy in the landing surface.It’s important for scientific research and production practice to obtain high quality regional soil moisture data.The traditional soil moisture acquisition is based on ground station monitoring,which has high precision and be continuous in time but its spatial continuity is poor.In order to obtain high precision,spatially continuous soil moisture data,the paper choice the middle reaches of Heihe’s full irrigation area as experimentation area,studied the spatial heterogeneity and interpolation method of the time-series soil moisture data which obtained by ground observation.The main research content is as follows:1、Spatial heterogeneity analysis on soil moistureIt mainly includes three aspects: statistical analysis,variance analysis and autocorrelation analysis.We understand the data layout and degree of variation through numerical statistical analysis,then taking the model fitting degree and the residual size as the test indexes,select the appropriate variation model analysis the spatial structure characteristics of regionalized variables.The third part of spatial heterogeneity is calculate the global autocorrelation Moran ’I coefficient of soil moisture,and examine the calculated results by Z test to check whether the soil moisture observation points with spatial location were significantly correlated with the observed values of adjacent observation points.2 、 Established a soil moisture interpolation method considering spatial-temporal characteristics is established based on the superposition theoryDecompose the soil moisture temporal data by wavelet decomposition method,in this way,we can get the low-frequency and residual parts.Then interpolate the two part by high accuracy surface model(HASM)and variogram function model separately.At last,fit the two part after interpolation,and obtained the high accuracy fitted results.3 、 Put forward an interpolation method in large area based on remote sensing classification and Markov field.In order to improve the accuracy of interpolation,we introduced remote sensing images classified them.We take Markov field theory into soil moisture interpolation,use spherical coordinate system to divide the region into grid,then establish Bayesian network.Get any point of soil moisture value by calculating the interpolation point using the method of distance square inverse and maximum likelihood estimation.