Abstract
The method of generating Agricultural Statistics Mesh Data (ASMD) in the northern part of Tochigi Prefecture was examined in order to create a database for a model to estimate nitrogen load. Two available data sets were used for generating ASMD. One was the land use data, which was prepared as 100 m mesh data (1/10 subdivision of standard mesh of the Japanese Standard Mesh System) in Digital National Land Information (DNLI), and the other was the agricultural census data of rural communities and their attached maps. The procedure is as follows : (1) Categorize each mesh into 9 land use classes according to its land use and the adjoining mesh data. (2) Count the total mesh number of each land use class of each rural community. (3) Calculate the acreages of paddy fields and upland fields from the total mesh number. (4) Adjust the differences between the calculated acreage and the statistic acreage. (5) Determine the acreage of crops and livestock number of each mesh. The following matters were clarified as a result of examining the accuracy of the mesh data obtained by this method. 1) When the acreage was estimated using this method rather than directly from the land use data, the contribution ratio of regression analysis between the statistic acreage and the estimated acreage increased. 2) The acreage adjustment between paddy and upland fields within the rural community improved the accuracy of acreage estimation. 3) When the spatial accuracy of the method was examined in comparison with the land use data from aerial photography, the acreage estimated using 1 km mesh data (standard mesh of the Japanese Standard Mesh System) showed strong correlation while the acreage estimated using 100 m mesh data showed no correlation to the acreage from aerial photography, which suggested that this method can be adaptable to estimation using 1 km mesh data. 4) When ASMD is created by this method, it is possible to assess livestock numbers or livestock intensity within not only administrative districts but also catchments.
