侵蚀性降雨识别的模糊隶属度模型建立及应用

Development and application of a fuzzy membership model for identifying erosive rainfall events

  • 摘要: 根据降雨及其引发土壤侵蚀是一个连续变化过程的特点,基于次降雨雨量(P)和最大30 min雨强(I30)建立了侵蚀性降雨识别的模糊隶属度模型。模型分为3个部分:①推求次降雨各物理参数的模糊隶属度;②将各物理参数的模糊隶属度有机结合起来,获得次降雨的模糊隶属度;③依据次降雨模糊隶属度对自然降雨事件进行判别,筛选出其中的侵蚀性降雨事件,以提高土壤侵蚀预报精度,选用错选度和漏选度两个评价指标对识别结果进行检验。模型在三峡库区陈家沟小流域进行了率定,并在3个空间尺度上进行了应用和检验,错选度均在20%以下。模型结构简单、参数物理意义明确,所需数据直接来自水文、气象观测。

     

    Abstract: Soil erosion caused by rains is a continuous process of changes.Such a soil erosion character is used in the development of a fuzzy membership model for identifying erosive rainfall events.The precipitation amount (P)and the maximum 30 min rainfall intensity (I30)of a rainfall event are used as two physical parameters in the model.The model involves three parts.The first part calculates the membership values of P and I30,and constructs the fuzzy membership functions.Secondly,the membership values of P and I30 are combined to form a new membership value of individual rainfall event.And finally,an erosive rainfall event is identified using the new membership values.The identified events can be validated through the mistaking degrees and the missing degree.The model is applied over three small-sized watersheds in the Three Gorges Reservoir region of southwest China,including Chenjiagou (8.32 km2),Daijiagou (0.67 km2),and Qiaojia (0.02 km2).The mistaking degrees on the three watersheds are less than 20%.The model has a simple structure,and its parameters have clear physical meaning and can be derived directly from the hydro-meteorological observations.

     

/

返回文章
返回