Research On The Impact Of Soil Organic Matter Of Sampling Points Allocation And Spatial Interpolation Method On Agricultural Land Classification Achievements

The classifications of agricultural land is an important way to protect the quality ofcultivated land resource security and achieve scientific management and sustainableutilization of cultivated land resource. How to enhance the precision of agricultural landclassification result has became one of the important issues in academic research nowadays.In order to save the cost of land classification and improve the accuracy at the same time,they are the important aspect to carry out efficient sample point layout and optimization ofspatial interpolation methods to get value of classification unit. The topsoil’s organic matteris one of the important soil attributes, and it’s also the necessary factor of agricultural landclassification. The study about its high efficiency of the samples Allocation methods andspatial interpolation model has important theoretical and practical value.This study chooses the hilly area which locates in south of Longhai City, ZhangzhouCity, Fujian Province as a typical area and designs eight different sample point griddensities（0.5×0.5km、1×1km、1.5×1.5km、2×2km、2.5×2.5km、3×3km、3.5×3.5km、4×4km）by using grid method which can efficiently characterize spatial variability of soilproperty. Applying the most commonly used ordinary kriging spatial interpolation methodupon the Land Statistics Method at present，we design6methods which are KDM、KDL、KTR、KDMTR、KDLTR and KYJZ, and we acquire Agricultural Land Classificationachievements according to the method from《Agricultural Land Classification Regulation》and compare classification achievement accuracy. The article is to study the impact ofCounty-level soil organic matter sampling point grid density and spatial interpolationmethod to the agricultural land classification achievements.The results are as follows:1、The spatial distribution maps predicted by using the6kinds of designed Krigingspatial interpolation methods are basically in line with the actual situation of soil organicmatter geospatial distribution in Longhai City with different soil sampling points griddensity. The research results show that the accuracy of the organic matter spatial predictionwill gradually reduce with the increase of the sample point grid density and the decrease ofsoil sampling points. At the same time, the performance of the various interpolationmethods has obvious difference. And the outcome of the agricultural land classificationaccuracy by Kriging(KTR、KDLTR、KDMTR) which combines with soil type information is significantly better than the unbound ones（KYJZ、KDL、KDM）. It has significantdifference between4×4km grid density and others when using KTR、KDLTR、KDMTRon spatial interpolation of organic matter prediction. Therefore, we could laid the3.5×3.5km sample point grid density when it needs higher soil organic matter prediction accuracy,and it’s better to predict by ordinary Kriging method (KDLTR).2、They are not significantly affected to agricultural land classification achievementsbetween eight kinds of samples grids density and six kinds of organic matter spatialinterpolation methods. Therefore, upon carrying out agricultural land classification work, ifonly to obtain the spatial information of soil organic matter content, we could laid the4×4km sample point grid density. At the same time, we could straightly predict by ordinaryKriging method from the data of soil organic matter content, and it will meet the accuracyrequirements of the agricultural Land Classification achievements. But the precondition isto ensure the soil sampling points combine the geomorphic types, the status of land usetypes and soil types scientifically, reasonably and evenly in the entire study area.3、 Even though soil sampling point grid density and optimization of spatialinterpolation methods significantly improve the accuracy of organic matter spatialprediction, the accuracy of agricultural Land Classification is not improved. The mainreason is that the simple indicator factors graded assignment method weakens correlation oforganic matter in the Agricultural Land Classification, which neutralizes the improvedprediction accuracy of organic matter at a certain degree. It puts forward further improvedthe way of soil organic matter conten’s score classification when carrying out theagricultural land classification work. According to local conditions, building piecewisefunction assignment model to determine the dimensionless score of organic matter contentat different stages. It will largely reduce the influence of human factors and effectivelyimprove the correlation between organic matter content and quality of the land, which willimprove the accuracy and precision of the agricultural land classification.