Models of regional agro-climatic resources survey and system assessment based on geo-spatial data mining

The research is to develop models for regional agro-climatic resources survey and system assessment based on geo-spatial data mining in Fujian Province of China. A set of average values of 30-years data in 8 agro-climatic indicators associated with geographical coordinates (collected from 71 weather stations in the province) was first processed using Statistical Analysis System (SAS) to develop a Trend-surface Analysis model. The resulting 8 equations would, then, produce estimated values for the indicators when supplying with geographical coordinates. Using the 8 equations with Digital Elevation Model (DEM) of Arc-GIS, the model would obtain estimated values for no data areas, i.e. integrated predicted grid values of Trend-surface with Residual (TSR), so as to form a distribution grid system of the provincial agro-climatic resources. At the end, the research introduced a fuzzy concept of “agro-climatic suitability” by a model which applied Fuzzy Mathematics and Factor Analysis over the indicators for assessing the agro-climatic system. The result of the model includes the agro-climatic suitability corresponding to the 8 indicators and 7 agro-climatic resources coefficients, which reveal the organic relationship between agro-climate and plant ecology. Using these models together, a series of assessments can be conducted at the grid level accuracy against a single as well as the general quantity and quality of the agro-climatic resources. In the sense of agro-climatology and physical geography, the output of the models has been well explained and will deliver backings to the regional agricultural planning and arrangement.