古岩溶系统训练图像制作

摘要

训练图像的制作是多点地质统计学方法用于储层建模的必要步骤。目前，古岩溶系统建模研究中训练图像的制作方法相对较为缺乏。为此，提出了一种以GoogleEarth影像数据为基础的古岩溶系统训练图像制作方法。综合GoogleEarth影像和野外露头资料对比研究，可得到溶洞、溶蚀沟道分布位置及大小，溶蚀洼地分布、断裂—裂缝系统的位置及规模等，将采集到的定量化GoogleEarth影像原始数据，用相关算法分析处理；其中，趋势面分析得到地貌面，椭球体方程得到溶洞特征，从而可得到用于多点地质统计学建模的古岩溶系统训练图像。%Making training image is a necessary step in multiple-point geostatistic reservoir modeling.Up to now, there is lack of proper training image making methods for fracture-care system modeling.Therefore,it was proposed that a method making training image based on GoogleEarth satellite image.Combined with GoogleEarth satellite image and out-crop survey data,we can get the position and size of karst channel,dissolution lowlands,and fracture system.As a result, we can get quantitative geological data based on the collected GoogleEarth image.Through the analysis of original data,we can get characteristics of geomorphology and karst cave.A training image for multiple-point geostatistics fracture-cave system modeling was made by using all the above obtained data.