青海湟水流域土壤-景观关系的空间推理研究

Huangshui river basin is the most important farming region of the Tibetan Plateau, the existing soil map can't meet the urgent needs of detailed soil informations in the field of environmental simulation and resource management, which usually required both space continuity and attributes accuracy. Soil-landscape space reasoning model is based on 3S technology and modern mathematical method, rooted in the soil forming factors theory, and formulated diagnosis and prediction system on soil-landscape relationship, which can be used to space reasoning of soil properties on watershed scale. Based on above reasons, this project is designed as following:Firstly, diagnose systemly on the environmental factors of soil formation in Huangshui watershed, identify collaborative relationship between soil properties and environmental factors, and characterize spatial pattern of soil properties's variation; Secondly, adopt 3S technique and fuzzy reasoning method, build a soil-landscape space reasoning model, and using it to calculateenvironmental membership degree curves,of major soil properties, and predict their spatial distribution patern as well; And mapping the fuzzy membership degree maps and uncertainty maps of major soil properties such as soil type, soil texture, soil thickness, soil organic matter etc. This research result can not afford some refined informantion of soil properties, which can be used for land suitability evaluation, soil erosion risk assessment and refined utilization of farming land.

湟水流域是青藏高原重要的农耕区域，现有的土壤图已不能满足资源环境和生态管理领域对空间上连续、属性上精确土壤信息的迫切需求。定量化土壤-景观模型是基于3S技术和现代数学方法，建立在成土因素学说之上的土壤-景观关系诊断和预测系统，可用于流域尺度的土壤属性空间推理。基于以上认识，本项目拟通过对湟水流域成土因素的系统诊断，识别土壤属性与环境因子的协同关系，刻画土壤属性的空间变异性规律；在此基础上，采用3S技术和模糊推理方法，构建土壤-景观空间推理模型，计算主要土壤属性的环境隶属度曲线，预测主要土壤属性的空间分布；最后绘制土壤类型、土壤质地、土层厚度、土壤有机质等主要土壤属性的模糊隶属度图和不确定性图。研究结果不仅可为湟水流域的土地适宜性评价、土壤侵蚀风险评估和农业土地利用提供精细化土壤信息，还可为类似地区的土壤-景观模型构建和预测制图提供方法借鉴。