Accuracy Loss Analysis in the Process of Cultivated Land Quality Data Gridding

Abstract : Spatial data gridding is one of the effective methods to solve the multi-source data fusion. In view of the current problems in the process of comprehensive analysis between the cultivated land quality data and other multi-source data, This paper, by adopting the method of the Rule of Maximum Area (RMA), converted the cultivated land quality data to the grid scale and analyzed the accuracy loss in the process of cultivated land quality data gridding in 6 grid scales (10M × 10M, 5M × 5M, 3M × 3M, 2M × 2M, 1M × 1M, 30S × 30S). Some conclusions have been reached. (1)The use of gridding methods will have assigned any analysis units to the specified data grid scales, and it provides a basis for spatial data integration, comprehensive analysis and spatial models construction;(2) Grid scale accuracy is higher, the original figure segmentation of cultivated land quality data is more serious, and grid results is more accurate, but grid computing time is increased step by step;(3) Through the study of the multistage of cultivated land quality data grid, the smaller the grid scale, the smaller the loss area of cultivated land quality, such as 10M × 10M gridding results lead to the most loss area, and Each grade area loss curve has a certain regularity;(4)From accuracy and computational efficiency, the most appropriate grid scale choice is 1M × 1M grid of 1:10000 cultivated land quality data of Daxing for the gridding processing.