The Study And Application Of Fuzzy Soil Mapping

With the rapid development of precision agriculture, environmental management, land management, eco-hydrological modeling and so on, there has being put forward higher and higher requirements on soil mapping. The traditional method of soil mapping depends on the census of experts’field work and their accumulating experiences, which spends much manpower and time. The traditional soil map is established on discontinuous conceptual model, which cannot express the soil’s gradient feature. While the smallest units of the traditional soil maps are limited by the scale size, we often merge it into other soil type or omit it directly when dealing with the smallest units. With the limits the soil map’s accuracy is low. And the traditional soil mapping method has become increasingly difficult to meet the high-efficiency requirement. Therefore, exploring the soil mapping methods with high accuracy and timeliness has attached much more attention.Accompanying with the rapid development and application of3S and other advanced technologies, it has provided some new technologies and the data support for the soil mapping. Meanwhile soil scientists around the world have also carried out extensive soil researching and mapping surveys, which has gradually developed a soil survey and mapping method—the digital soil mapping. It has the quantitative soil—landscape model as the theoretical basis, has the spatial analysis and mathematical methods as the technical means. Among the methods of digital soil mapping, the method of fuzzy mathematics is mostly widely-used, and the most successful. The method of fuzzy mathematics can well describe the fuzziness of the real world, which can objectively reflect the real world, so the method based on fuzzy clustering can more accurately describe the spatial variation of soil attribute in the digital soil mapping.In this study, we take Hong’an County as an example. The five main types of terrain factors (dem, slope, profile curvature, plan curvature, Topographic Wetness factor,) determine the water movement and redistribution, which basically are the main factors influencing the formation and development of soil. After dealing with the data of the terrain factors, we use three methods:the fuzzy performance index (FPI), the separation distance method (S) and the optimal index method to do fuzzy cluster analysis. The optimum cluster index for the study area numbers. By the fuzzy clustering maps we choose the typical points which membership are bigger than0.9. Each type chooses three points. Then we get the typical points With GPS in the wild to obtain the corresponding relationship between environmental and soil type. At last we apply the SoLIM model to mapping the soil type in the study area. To verifying the soil mapping’s accuracy we use three methods: subjective sampling, grid sampling, profile sampling. The results show that the overall accuracy is61%. Further more, forest sand soil and sand soil which accounted for more than50%in the study area have much influence on other soil types.All in all, the FCM method is effective to obtain the corresponding relationship between the environmental combining categories and the soil type in Hong’an county, this method not only needs less field samplings which could achieve the typical environmental knowledge effectively, but also improves the efficiency of taking samplings outdoor.