【作者】 徐翠玲；

【导师】 钱壮志；

【作者基本信息】 长安大学 ， 地球探测与信息技术， 2007， 博士

【摘要】 面对21世纪对矿产资源需求的严峻形势,矿产预测研究将面临新的机遇与挑战,随着我国矿产预测工作的发展和各种地学数据的急剧增长,GIS在矿产预测工作中的应用研究将随着地学信息科学的发展日益普及和深化。本论文以《新疆中天山沙垄以西金、银、铜、铅、锌等金属矿产成矿规律研究与靶区优选》项目(新疆地质矿产勘查开发局,项目编号XGMB2004001,2004年—2006年)为依托,选取基于GIS的矿产预测为主要研究内容,从GIS功能出发,结合矿产预测过程中的实际问题,探讨GIS在矿产预测中的作用和需要解决的关键问题,并以新疆沙垄以西地区为例,展开对多源地学信息的综合分析与解译,建立矿产预测空间数据库,探索GIS在矿产预测中的作用。取得的主要成果和结论如下:1、从GIS功能与矿产预测的目的任务出发,揭示了GIS在矿产预测中的地位和作用。一方面,GIS已经成为矿产预测强有力的工具,其先进的空间数据库管理大大提高了预测的效率,强大的可视化技术为地物化遥数据的处理创造了条件,有力地空间分析方法更是使矿产预测诸多模型的实施成为可能。另一方面,GIS在预测中也有其局限性,矿产预测的本质仍依赖于成矿规律的研究,数据质量及完整性才是GIS成功应用于矿产预测的根本。2、在分析GIS组成、功能及发展趋势的基础上,结合矿产预测工作的性质和流程,论证了将GIS应用于矿产预测的必然性,并提出应用GIS进行矿产预测的基本工作流程。3、全面系统地分析了GIS应用于矿产预测中需要解决的关键问题,研究成果显示,信息的完整和数据的质量是根本,数据库的设计是关键,成矿信息的提取分析是基础,空间分析方法模型的选择是保障,多学科专家的共同参与是原则。因此,GIS能否成功地应用于矿产预测的关键在于上述条件的成功实施。4、在总结GIG数据模型的重要性、研究现状以及模型类型的基础上,分析了我国在地学数据组织方式上的不足,指出面向对象的数据模型应当引起地学专家的关注。5、全面、系统地收集和分析了沙垄以西地区地质、地球物理、地球化学、遥感、矿产地信息,并基于Geodatabase在该地区建立了矿产预测空间数据库,从而为矿产预测和数据共享奠定了基础。6、对各种成矿信息进行了处理、分析和提取,获得沙垄以西地区矿产预测的变量信息,在此基础上应用证据权重模型对金矿产资源进行定位预测,优选出成矿有利靶区共107个,其中A类靶区20个,B类靶区37个,C类靶区50个。7、应用属于数据挖掘技术的决策树模型对沙垄以西地区铅锌矿产资源进行预测,优选出成矿有利靶区共31个,其中A类靶区3个,B类靶区12个,C类靶区17个。该探索也为矿产预测提供了一种新的思路和方法。8、通过实例,对比分析证据权重模型和决策树模型在矿产预测中的优缺点,得出,证据权重模型比较适合于地质工作程度高且能最大限度地获取找矿信息的地区;而决策树模型的优点是不需要大量的矿床,对于研究程度低的地区无疑是一种好方法,但也存在计算过程繁杂等缺点。开展基于GIS的矿产预测研究,不仅对于推进GIS在矿产预测中的应用进程具有重要的现实意义,而且还有助于地质、构造、矿产、物探、化探、航磁、遥感等信息的综合研究,对进一步完善矿产预测方法与理论体系具有一定的科学意义。同时,在实践方面,可探索开展矿产资源潜力评价工作的新方法。更多还原

【Abstract】 In the face of the grim situation of the demand for mineral resources in the 21st century, mineral resources prediction will face new opportunities and challenges. With the development of mineral resources prediction and the rapid growth of the various geological data at home, the applied research of GIS in mineral resources prediction will be getting popular and deepening along with the development of geological information science. Based on the project of "Metallogenic Law Research and Target Selection on Gold, Silver, Copper, Lead, Zinc etc to the West of shalong in Middle Tianshan in Xinjiang"(Xinjiang Mineral Exploration and Development Bureau, Serial number of the project, XGMB2004001, 2004-2006), the paper chose mineral resources prediction based on GIS as major studying content, started from GIS function, and discussed role and key issues to be resolved of GIS in mineral resources prediction combining practical problems in its process. And the paper has taken the area to the west of shalong in Xinjiang as an example, gotten along with to the comprehensive analysis and interpretation to Multi-source geological information in this area, established spatial database of mineral resources prediction, and discussed the role of GIS in mineral exploration prediction. The main results and conclusions are as follows:1. From GIS function and the purpose of mineral resources prediction, the paper has revealed the status and action of GIS in the mineral prediction. On the one hand, GIS has become a powerful forecast tool, whose advanced spatial database management has greatly enhanced the efficiency of mineral resources prediction, and powerful visualization technology has created the conditions for the conduct of the geologic, geophysical, geochemical, and remote sensing data. On the other hand, GIS has its limitations in the forecast, the inbeing of mineral resources prediction is still dependent on the law of mineralization, the quality and integrity of GIS data is fundamental for it to successfully apply to the mineral resources prediction.2. Based on the analysis of structure, function of GIS and its developing trend, combining the nature and processes of the mineral resources prediction work, this paper has discussed the inevitability that GIS will be used in mineral resources prediction work, and proposed the basic work flow of application of GIS in mineral resources prediction.3. This paper has comprehensively and systematically analyzed key issues to be addressed of the application of GIS in mineral prediction. The research result shows that the quality and integrity of GIS data is fundamental, database design is the key, the extraction of metallogenic information is the basis, the choice of spatial analysis model is the guarantee, and the multidisciplinary experts’ attendance is principle. So, the key that GIS can succeed in applying to mineral resources prediction is successful implementation of above conditions.4. Based on summarizing the importance, research present situation and type of GIS data model, this paper has analyzed the shortage of the organization model of geological data in our country, and pointed out the object-oriented data model should arouse geologists’ interest.5. This paper has comprehensively, systematically collected and analyzed information of Geology, geophysics, geochemistry, remote sensing and ore origin place to the west of shalong in Xinjiang, established a spatial database of mineral resources prediction in this area based on Geodatabase, and laid the foundation for mineral resources prediction and data sharing.6. The paper has dealt with, analyzed and extracted several of mineralization information, and obtained variable information of mineral resources prediction to the West of shalong. Based on the above work, the author has made a positioning prediction on gold resources by using weights of evidence model, and selected 107 available targets in all, in which there are 20 A-type targets, 37 B-type targets, and 50 C-type targets.7. The paper has made a prediction on Lead, Zinc resources to the west of shalong by using decision tree model belonging to data mining technology, and selected 31 available targets in all, in which there are 3 A-type targets, 12 B-type targets, and 17 C-type targets. The exploration has also provided a new thinking method for mineral resources prediction.8. By the cases, this paper has contrasted and analyzed advantages and disadvantages of weights of evidence model and decision tree model, and made a conclusion that weights of evidence model is suitable for the areas where geological work degree is high and the maximum prospecting information can be obtained; and decision tree model which don’t need a lot of known deposits is suitable for the areas where the degree of geological research is low, but its calculations are complicated.Developing mineral resources prediction on GIS not only has great practical significance for promoting the applied process of GIS in mineral prediction, but also contributes to the comprehensive study of information on geology, tectonics, mineral resources, geophysical, geochemical, aeromagnetic, and remote sensing etc. It also has a certain degree of scientific significance for further perfecting Mineral resources prediction methods and theoretical system. And at the same time, it contributes to searching after new method developing mineral resource potential evaluation.更多还原