Comparison Of Cultivated Land Digital Mapping Methods Of Soil Fertility

In recent years, as a low-power, high-efficiency, along with advances in technology, modern precision agriculture gradually replaces the existing backward agricultural production ways. Modern precision agriculture proposed new requirements to accurately soil fertility prediction, accelerating the establishment of soil fertility and digital atlas database has become a new research direction on digital soil mapping. Development of computer and3S technology brings new opportunities and challenges for digital mapping of soil fertility, the development of Proximal soil sensing accelerate the obtaining speed of soil fertility, remote sensing and DEM images improve the accuracy of mapping soil fertility. Soil fertility digital mapping provides information support for the protection of soil fertility, which has a very important practical significance for the protection of cultivated land’s quality.In this paper, Fuyang County which in Zhejiang Province as the study area, firstly, selecting the arable soil fertility major impact indicators of Fuyang County, using common statistical methods and statistical methods analyze indicators, and using ordinary kriging for digital mapping of each indicator; then using the master component analysis and other methods of data processing, obtained soil fertility minimum data set. Through determining the index weight, find out integrated soil fertility index by sampling points according to the common factor method; finally using ordinary kriging, fuzzy C-means clustering and regression kriging for digital mapping of integrated soil fertility. Specific content and conclusions are as follows:1) Determine soil fertility factors set in Fuyang County, including nitrogen, total phosphorus, total potassium, cation exchange capacity, potassium, phosphorus, alkaline hydrolysis, nitrogen, topsoil thickness, pH, organic matter content and bulk density. Through principal component analysis, according to the characteristic values, the correlation coefficient, Norm values and variation coefficient selecting, getting Minimum Data Set of soil fertility evaluation in Fuyang County, which are total nitrogen, total potassium, nitrogen, phosphorus, potassium, pH, CEC. Based on the weight coefficients obtained by principal component analysis, derived soil integrated fertility index.2) Based on integrated soil fertility index, Ordinary Kriging prediction map of soil fertility has a reasonable resulting distribution, prediction accuracy RMSE was0.096, indicating the result has an important practical significance. From soil classification chart, mainly arable land in the middle of the three levels, accounting for81.53%of the total arable land, of which34.67%total area is class3. This method requires a large amount of field data, which means need to support the higher cost of mapping, time-consuming and labor-intensive. Ordinary Kriging rely solely on the spatial variability of soil characteristics, without considering other factors, the prediction accuracy can be improved.3) Based on soil landscape model, select the elevation, slope, curvature plane, profile curvature, topographic wetness index, normalized difference vegetation index, the ratio vegetation index, soil type, land use types, using Fuzzy C-means clustering forecast soil fertility of Fuyang County. The results show relatively low prediction accuracy (RMSE0.126), each grade distribution differences, concentrated in the middle of the two levels of81.70%of the total area, the overall fertility distribution has a reasonable trend. Relying solely on environmental factors to predict fertility, which has low accuracy, at the same time, this method requires only a typical environment combined sample data, which means greatly reduce the workload, if continue to study, it may be one of the main methods for forecasting soil fertility distribution in the future.4) By analyzing environmental factors, Regression Kriging selects the higher correlation indicators concluding the slope and topographic wetness index as auxiliary variables, combined with integrated soil fertility index for predicting cultivated fields’ fertility of Fuyang County. Regression Kriging adding some environmental factors, can effectively increase the prediction accuracy (RMSE0.081), the classification mapping predicted a slightly difference compare to ordinary kriging,81.70%of the total area are the middle of three levels, in which class3area is the largest, contains34.16%of the total area.