基于Monte Carlo模拟的三维剖面地质界线不确定性分析

三维剖面地质界线是构建三维地质结构模型的重要基础数据,其不确定性会影响三维模型的几何形态和属性分布。以单一分布为假设前提的统计学不确定性分析方法掩盖了其他概率分布特征对模型的影响。突破单一误差分布条件的假设前提,本文使用Monte Carlo方法模拟了不同概率分布情况下地质剖面数据中地质界线的抽样采集,以及地质界线空间分布的不确定性;依托地质界线空间位置与地质属性的耦合关系,提出了用地质属性概率分布实现地质界线空间不确定性的定量可视化,并结合实际地质剖面探讨了多种概率分布条件下地质界线的空间不确定性。实例研究表明,基于Monte Carlo模拟的不确定性分析方法可以突破单一误差分布假设条件,结合地质属性概率可充分揭示出建模数据的内在不确定性与模型外在要素形态之间的耦合关系。

Uncertainty Analysis of Geological Boundaries in 3D Cross-Section Based on Monte Carlo Simulation

Geological boundaries in 3D cross-sections are important data for constructing 3D geological models, and their uncertainties will affect the geometry and attributes distribution in final models.The approaches based on the assumption of single error distribution for uncertainty analysis cover up other characteristics caused by other different error distributions.To break the single error distribution assumptions, based on the Monte Carlo simulation method, this paper simulated the sampling acquisition with different error probability distributions and spatial uncertainty of geological boundaries in 3D geological cross-section.Based on the coupling relationship between spatial position and geological attributes, a term of "geological attribute probability" was proposed to quantitatively visualize uncertainty of geological boundaries.Combined with concrete geological cross-section, this paper discussed the spatial uncertainty distribution of geological boundaries that followed multiple error probability distribution functions.The concrete example shows that the presented approach can overcome the defects of single error distribution assumption.Combined with the geological attribute probability, the approach can also reveal the coupling relationship between inner uncertainties existed in modeling input data and feature shape of models.