| 降低矿产资源定量预测不确定性的双向预测方法 |
| |
| 引用本文： | 孔维豪,肖克炎,陈建平,孙莉,李楠.降低矿产资源定量预测不确定性的双向预测方法[J].地学前缘,2021,28(3):128-138. |
| |
| 作者姓名： | 孔维豪 肖克炎 陈建平 孙莉 李楠 |
| |
| 作者单位： | 1.中国地质科学院 矿产资源研究所 自然资源部成矿作用与资源评价重点实验室, 北京 1000372.中国地质大学(北京) 地球科学与资源学院, 北京 1000833.北京市国土资源信息研究开发重点实验室, 北京 100083 |
| |
| 基金项目： | 国家自然科学基金项目(41802250);国家重点研发计划项目(2017YFC0601501);国家重点研发计划项目(2017YFC0601500) |
| |
| 摘 要： | 成矿过程是一个复杂的物理化学过程,由于地质自身的不确定性、原始数据采集和处理的不当、预测方法中经验参数的不确定性等多重因素的叠加,造成矿产资源定量预测结果中潜在大量不确定性。在科学认识这些不确定性的基础上,如何降低不确定性是预测评价研究的一个重要方向。以地质异常理论和成矿动力学为指导,双向预测评价方法是降低地质异常分析中不确定性的有效途径,该方法具体包括基于矿床成因模型的成矿模拟和基于找矿模型与勘查数据相结合的模型驱动预测。前者作为研究成矿地质演化过程、探讨成矿动力学机制的定量化方法之一,可以直观展示成矿过程内部物理化学变化。作为对致矿地质异常分析的有效手段,通过将成矿过程抽象为不受时空间限制的可迭代计算的偏微分方程组,可实现定量化描述复杂成矿动态过程并预测成矿有利部位。通过挖掘成矿有利信息,分析地质变量并赋值,为预测模型提供大量的定量化预测变量和特征值,是矿产资源定量预测评价的一个最具潜力的发展方向。后者以勘查学为指导的矿化异常分析,从矿致地质异常的角度开展定量预测,减少了单一成矿有利信息的多解性并降低了预测结果的不确定性。该技术手段是依托空间数据库、地质统计学和地理信息系统空间分析技术支撑,以三维地质体模型的建立为基础,以分析成矿规律并建立找矿模型为核心工作内容,以证据权、找矿信息量等数学方法为工具,统计分析研究区内各地质要素单元的分布情况来探讨各地质要素对矿产预测的影响,最终实现基于“立方体预测模型”的定位、定量和定概率的隐伏矿体三维预测目标。以上方法从两种不同的地质角度和定量化理念创新性地实现了双向联合预测评价,两种技术手段的融合作为综合圈定“5P”找矿地段的数学地质方法,其作用和价值是相互补充并有机结合的。通过文中方法介绍和应用实例的研究成果,可以明确该方法确实提高了矿产资源定量预测评价的预测精度,一定程度上降低了预测的不确定性,整体上推动了地球科学研究由定性描述向定量化自然科学的转变。
|
| 关 键 词： | 地质异常 定量预测 不确定性 双向预测 |
| 收稿时间： | 2021-01-02 |
A combined prediction method for reducing prediction uncertainty in the quantitative mineral resources prediction |
| |
| KONG Weihao,XIAO Keyan,CHEN Jianping,SUN Li,LI Nan.A combined prediction method for reducing prediction uncertainty in the quantitative mineral resources prediction[J].Earth Science Frontiers,2021,28(3):128-138. |
| |
| Authors: | KONG Weihao XIAO Keyan CHEN Jianping SUN Li LI Nan |
| |
| Institution: | 1. MNR Key Laboratory of Metallogeny and Mineral Assessment, Institute of Mineral Resources, Chinese Academy of Geological Sciences, Beijing 100037, China2. School of Earth Sciences and Resources, China University of Geosciences(Beijing), Beijing 100083, China3. Beijing Key Laboratory of Development and Research for Land Resources Information, Beijing 100083, China |
| |
| Abstract: | Mineralization is a complex physical and chemical process. In the quantitive prediction and evaluation of mineral resources, prediction uncertainties are common in the final predication results due to geologic factors, improper data collection and processing, use of empirical parameters, etc. On the basis of recognizing the sources of uncertainty, uncertainty reduction becomes a major research direction in mineral prospecting. A combined metallogenic prediction method using the forward inverse technique has proven to be an effective solution for reducing uncertainties in the geological anomaly analysis. The combined method is composed of numerical simulation of metallogenic processes based on the genetic model of deposit, and model-driven prediction and evaluation technique based on prospecting model and data. As the combined method takes into account both mineralization processes and model-driven prediction/evaluation in the geo-anomaly analysis, it reduces multiple interpretations of a single metallogenic information therefore reducing prediction uncertainty. In its core contents of analyzing metallogenic patterns and establishing prospecting models, the combined method, mainly using spatial database and geostatistics and GIS analysis techniques as well as 3D geological model building, performs weight of evidence analysis and information value evaluation of the distribution of geological variables to investigate its influence on the prediction result. The aim is to build a “3D cube prediction model” for quantitive prediction, fulfilling the goal of Location, Quantity, Probability quantification in blind orebody prospecting. In conclusion, the combined prediction method, as an important basis for delineating the “5P” prospecting areas, was born by innovatively combining two prediction methods that have complementary predictive functions and values. By doing so, a complete forward inverse combined prediction scheme is realized. We show by example that the combined method can improve accuracy and reduce uncertainties in the quantitative prediction and evaluation of mineral resources, which helps to promote the transformation of geoscience research from qualitative to quantitative. |
| |
| Keywords: | geological anomalies quantitative prediction uncertainty combined prediction |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《地学前缘》浏览原始摘要信息 |
| 点击此处可从《地学前缘》下载免费的PDF全文 |
|
