青藏高原沱沱河地区冻土融化深度预测的概率分析

A Probability Analysis for Forecasting the Thaw Depth in Tuotuohe,Tibetan Plateau

多年冻土空间异质性和边界条件及土性参数的可变性,往往会导致了冻土变化预测的不确定性,全球气候变暖更加剧这一过程.以青藏高原沱沱河地区试验段冻土融化深度预测为例,提出了在全球气候变暖条件下,基于参数服从某一概率分布的确定性模型的概率分析方法,基于此方法进行了融化深度的概率预测.由含水量、干容重的概率分布和20组ATI(空气融化指数)可能值,经蒙特卡罗技术同Stefan公式应用模拟得到了未来100a内天然冻土融化深度,并得到失效概率曲线.结果表明:土体性质的不确定性和温度的可变性导致了冻土变化预测很大的不确定性,概率分析相比于传统确定性分析更加符合实际.在冻土区工程建设中设定冻土融深时,运用概率分析的方法预测融深相比于确定性分析更加安全.同时,气温持续升高对多年冻土区的基础可靠性有强烈影响,设计时应予以考虑.

Due to the spatial heterogeneity of permafrost and the variability of boundary conditions and soil parameters,forecasting the change in frozen soil has an uncertainty and is exacerbated by global warming.In this paper,probability analysis method is proposed on the basis of parameters in deterministic model submiting to some probability distribution under global warming,with a case to forecast the thaw depth for frozen soil in Tuotuohe on the Tibetan Plateau,and then a probabilistic prediction is conducted.The thaw depths and failure probability curves of frozen soil for the next 100 years are obtained by using both Monte-Carlo simulation and Stefan equation based on the distributions of water content and dry unit weight,together with 20 air thawing indexes.It is revealed that the uncertainty of soil properties and variability of temperature make the prediction of frozen soil change exhibits great uncertainty.Compared with the conventional deterministic analysis,the probability analysis method for the thaw depth design in engineering construction in permafrost regions is more realistic and safer.The reliability of foundation in permafrost regions is susceptible to temperature,which must be taken into account in design.