Standarlising Spatial Data: A Kriging Interpolation Based Approach

Spatial data analysis is one of the key methods in geophysical analysis. Thespatial data model is the foundation of spatial analysis. he model construction of spatialdata has a direct impact on the breadth and depth spatial analysis. Spatial data such asstress, strength, and the reservoir of the relevant data are often scattered or in needpoint no data distribution is extremely unreasonable for the observation points in areas,which need to be anchored to a known point interpolation normalized. The methodusing Kriging interpolation of spatial data can often achieve the desired results.This dissertation first describes the reason of choosing the Kriging interpolationmethod. It introduces the basic theories of the Kriging method, discusses thecharacteristics of different Kriging interpolations, and improves the Kriginginterpolation algorithm. The improved algorithm has been implemented and thecorrectness and advantage of the algorithm has been evaluated.The major contribution of this dissertation is on applying theory to practicalproblems.Its main contents include: grid partition theory, and estimate the searchstrategy, experiment point variograms deriving and fitting, ordinary Kriginginterpolation method combing the calculation process and ultimate theory variationfunctions optimal inspection. This dissertation introduces how to obtain the primarydata pretreatment, converts Kriging interpolation needed after the3d data, thecalculation flow diagram, and simulated Kriging interpolation results in use of Matlabsoftware. In this process, we improve the methods to calculate variograms parameters;And according to the theory of linear equations are solved the variation functions of akind of parameter optimization method, removed the man-made factors, there are rulesto follow, and make its calculation the automation on computer.Experiential results show that, the Kriging interpolation research method studiedin this dissertation, can be applied to the calculation of spatial correlation data. Bydoing this, we can achieve reasonable results. All in all, this proves that the studycarried out in this dissertation will contribute to the data normalization problems inindustry, agriculture and other domains..