Research On The Methods Of Uncertainty Analysis Of Three-dimensional Geological Model

A 3D geological model is an approximation of an actual geological phenomenon.Various uncertainty factors in modeling reduce the accuracy of the model.Hence,it is necessary to assess the uncertainty of the geological model.In the process of modeling,uncertainties from various sources are propagating and accumulating,which leads to the decline of model quality.In order to ensure the availability of the model,the comprehensive effect of multi-source uncertainties on the quality of the model should be considered.Since the dependence between stratigraphic variables will affect the uncertainty estimation of multi-strata scene,it is necessary to consider the influence of spatial dependence and interdependence of strata on the transmission of information.For these problems,we study the comprehensive effect of multi-source uncertainties on the model to improve the accuracy of uncertainty evaluation,expand the object of uncertainty analysis from isolated point to the spatial neighborhood across multi-strata by considering the dependences among strata,and improve the modeling precision of the model by using the dependence structure as a constraint.In this dissertation,the propagation and accumulation of multi-source uncertainties in the modeling process and the interaction between stratigraphic variables are studied in depth and applied to the uncertainty evaluation and model update of geological models.Specific research contents and innovations are as follows:(1)Uncertainty assessment of a 3D geological model by integrating data errors,spatial variations and cognition bias.For the problem that the model quality is affected by multi-source uncertainties in the process of modeling,this dissertation proposes a method to assess the comprehensive uncertainty of a 3D geological model affected by data errors,modeling method and cognition bias.Based on Bayesian inference,the proposed method utilizes the established model and geostatistics algorithm to construct a likelihood function of modeler's empirical knowledge.The uncertainties of data error and modeling method are integrated into the probability distribution of geological interface with Bayesian Maximum Entropy(BME)method and updated with the likelihood function.According to the contact relationships of the strata,the comprehensive uncertainty of the geological model is calculated using the probability distribution of each geological interface.Using this approach,we analyze the comprehensive uncertainty of a 3D geological model of the Huangtupo slope in Badong,Yichang,China.The change in the uncertainty of the model during the integration process and the structure of the spatial distribution of the uncertainty in the geological model are visualized.The application shows the ability of this approach to assess the comprehensive uncertainty of 3D geological models.(2)Joint uncertainty modeling of multi-strata structure based on spatial R-Vine Copula.To overcome the limitation of uncertainty analysis of geological model for a single location,this dissertation proposes a joint uncertainty modeling method of multi-strata structure based on spatial R-Vine Copula.This method can be used to calculate the joint distribution and conditional distribution of multiple stratigraphic variables,analyze the joint uncertainty of multi-strata,and improve the prediction accuracy of geological structure.By analyzing the spatial dependence and interdependence of strata,the Copula method is used to describe the dependence structure of stratigraphic variables.According to geological rules,a Vine-Copula was constructed to describe the complex dependence structure of multiple stratigraphic variables in the spatial neighborhood.Combined with spatial Copula method,a spatial R-Vine Copula model is established.Finally,the joint distribution and conditional distribution of multi-strata variables are calculated,the joint uncertainty of multi-strata structure is analyzed,and the geological structure constrained by the data from multi-strata is predicted.In this chapter,the Quaternary sedimentary strata in Shanghai area are taken as an example to test the feasibility of the spatial R-vine Copula method by predicting the structure of multi-strata and analyzing the joint uncertainty of multi-strata scene.(3)Improving the accuracy of geological models based on the interdependency of adjacent strata.For the low modeling accuracy of single stratum caused by sparse observations,this dissertation proposes a method of model update based on the interdependency of adjacent strata.This method does not need to add additional auxiliary data,and makes full use of the mutual information hidden in the adjacent strata to eliminate the uncertainty of the geological model and improve the modeling accuracy.Based on the interdependence between adjacent strata,the Copula function of adjacent interfaces and the marginal distribution function of the elevation of each interface are fitted using the sample data.The joint distribution model of adjacent interfaces and the likelihood function of the interface to be updated are constructed.In the Bayesian framework,the interface elevation calculated by the data from a single stratum is used as the prior value of the elevation parameter.The posterior distribution of the interface elevation is obtained by Bayesian update with a likelihood function.With the conditional expectation of the interface elevation,the interface model is updated.Using this approach,we take the borehole data and subsurface model of Shanghai,China to test the validity of the proposed method.The experimental results show that after the model is updated,the geological interface model accuracy is improved.To sum up,the uncertainty analysis and model update of 3D geological model are studied and explored in this dissertation.In the aspect of uncertainty evaluation of geological model,the influence of multi-source uncertainty on the model quality is studied.An integration method of multi-source uncertainty of 3D geological model based on Bayes framework is proposed to evaluate the comprehensive uncertainty of geological model.With the the probability distribution of geological interface elevation,the dependence structure of geological multivariable is discussed.Using Copula theory,we expand the uncertainty evaluation from a single stratum to the multi-strata scene.This dissertation proposes a joint uncertainty modeling method of multi-strata structure based on spatial R-Vine Copula to predict the multiple-strata structure and analyze the joint uncertainty of the scene.Based on dependence structure modeling and Bayesian inference,a geological model update method with the interdependence constraint of adjacent strata is proposed.Without adding additional auxiliary data,the geological interface model is updated to improve the model accuracy.The above research work provides new methods support for the uncertainty analysis of 3D geological model to improve the accuracy of uncertainty analysis and reduce model error.