评述了自20世纪80年代末我国开展泥石流危险度评价研究以来，两个主要不同时期的区域泥石流危险度评价方法及其发展过程。详述了我国现行的区域泥石流危险度评价技术及其改进的评价因子转换赋值新方法。现行的区域泥石流危险度评价方法于1995年提出，是以不同的行政区划为基本单元来进行的。评价因子包括主要因子泥石流分布密度，次要因子岩石风化程度系数、断裂带密度、≥25°坡地面积面分比、洪灾发生频率、月降雨量变差系数、年雨量平均≥25mm大雨日数和坡度≥25°坡耕地面积百分比。区域泥石流危险表达为以上8个因子极差变换后的赋值与其相应的权重乘积之和。这一评价公式，加强了主要因子的作用。主要因子的权重系数上升到33%时则使区域泥石流危险度评价可用一个简单的数学公式来表达。得出的危险度数值已经标准化，取值范围介于0-1之间。这一评价方法的不足之处是得出的危险度数值只有相对意义而不具有绝对意义。在最新的区域泥石流危险度评价研究中，通过分段函数赋值的新方法克服了上述不足。分段函数赋值是基于以下的假定：当评价因子超过和等于某一上限值时赋值为1；等于下限时赋值0；处于中间值的赋值为0.5。以这3个数值为控制点，假定在每2个点之间，赋值呈线性变化。当中间点为上、下2点之平均值时为双线性模型；当中间点不为上、下限2点之平均值时为三线性模型。根据我国西南地区已有资料的统计，得出了区域泥石流危险度8项评价因子的分段赋值函数。通过2种不同赋值方法在四川凉山州和阿坝州区域泥石流危险评价中计算结果的比较，从绝对数值来看，分段函数赋值得出的危险度高于极差变换得出的危险度。对于泥石流危险度分析来说，相对偏大的保守估计更为可取。从相对数值来看，两者都具有相同的变化趋势。说明分段函数赋值计算出的区域泥石流危险度既能在同一区域内进行比较，也能在不同区域间进行比较，达到了改进的目的。现阶段我国正在使用的区域泥石流危险度多因子综合模型仍然属于经验模型的范畴。它是在经验的基础上采用推理和统计的方法建立的一种评价模型。我们的最终目标是在科学假定和合理简化的基础上，用代表泥石流危险度的2个本质的特征参数泥石流规模和发生频率来建立起区域泥石流危险度的理论模型，并用量化的数学公式来表达。因此，泥石流规模和发生频率相互关系的研究是实现这一目标的突破点。也是今后区域泥石流危险度评价研究的重点所在。

This paper has reviewed two assessing methods for regional debris flow hazard developed in two different stages since the end of the 1980's in China, and the technique currently used and its newest development are presented in detail. Present assessing technique was proposed in 1995, which is based on the administrative regions as the basic assessing units. Assessing factors include debris flow spatial density, degree of weathering of rock, active fault density, percentage of slope greater than 25 degrees of the total, frequency of flooding hazards, average covariance of monthly precipitation, average days with rainfall greater than and equal to 25mm by ten years, and percentage of cultivated slope greater than 25 degrees of the total. Regional debris flow hazard assessment is expressed as a sum of the products of the normalized eight factors and their weights. This assessing formula has emphasized the leading factor, debris flow spatial density, whose weight has been raised to 33%. Meanwhile the assessment has been simply expressed as a mathematical equation. Moreover the values of the hazard degree have been raised to 33%. Meanwhile the assessment has been simply expressed as a mathematical equation. Moreover the values of the hazard degree have been standardized between 0 and 1, which is consistent with the degrees of the vulnerability and risk of debris flow. The inadequacy of the present assessing method is that their hazard values are only relative values which cannot be compared each other from one region to another. The newest research on regional debris flow risk overcomes this problem. The transformation functions are based on the following regulation. Having determined the upper limit value and the mean value for each factor, the factor can be transformed by taking the upper limit value, the mean value, and the zero as the control points, and the three points are transformed to 1, 0.5, and 0, respectively. It is assumed that the transformed factors vary linearly between the control points. Therefore, each factor can be represented by linear or bilinear model. Based on the statistical data in southwestern China, we estimated the upper limit value, and the mean value for each of the eight factors. Then the transformation functions for these factors may be obtained. Through the comparison of the two methods applied to the prefectures of Liangshan and Aba of Sichuan province in southwestern China, the results show that, from the view of the absolute values, the hazard degrees calculated by transformation functions are greater than those by the normalized functions, while the conservative assessment with a reasonable higher hazard values are probably more desirable for the analysis of debris flow hazard; from the view of the relative values, the variation tendency of the two methods are generally similar, which indicates that the hazard degree computed by the transformation functions are absolute values, and can be compared each other not only in the assessing region but also in the other regions. At present, the Multiple Factor Composite Assessment Model (MFCAM) for regional debris flow hazard is still an empirical model, which is established on a basis of the experiences and inferences derived from the nature of debris flow that is not yet fully understood now. Our ultimate objective is to construct a theoretic model for regional debris flow hazard as sessment based on the reasonable assumption and simplification of debris flow hazard with two intrinsic variables: magnitude of debris flow and its frequency of occurrence, and to use a standardized and quantified formula to express the regional debris flow hazard. Therefore, the relationship between magnitude of debris flow and its frequency of occurrence is substantially the research emphasis for regional debris flow hazard in the near future.