基于Hyperion数据的油砂分布遥感研究  

Application of Hyperion Hyperspectral Image for Studying on the Distribution of Oil Sands

作　　者：尤金凤[1,2] 邢立新[1] 潘军[1] 单玄龙[3] 樊瑞雪[1] 曹会[4] 

机构地区：[1]吉林大学地球探测科学与技术学院,长春130026 [2]空军航空大学,长春130022 [3]吉林大学地球科学学院,长春130061 [4]中国人民武装警察部队黄金第一支队,黑龙江牡丹江157021

出　　处：《吉林大学学报（地球科学版）》2016年第5期1589-1597,共9页Journal of Jilin University：Earth Science Edition

基　　金：国家科技重大专项(2011ZX05028-002);中国石油天然气股份有限公司科学研究与技术开发项目(2013E-050102);中国地质调查局项目(1212010761502)~~

摘　　要：依据油砂中烃类的微渗漏和油砂组分光谱特征响应原理,利用Hyperion高光谱影像提取和识别与油砂分布相关的波谱信息,进行非常规油气能源——油砂分布的有利区预测。根据油砂所致烃类微渗漏的地表特征可知,低植被覆盖区的异常以矿物异常为主,中、高植被覆盖区的异常以植被异常为主。利用归一化植被指数表征地表植被的不同覆盖程度:当其值为[0.0,0.4)时,采用SAM(spectral angle method)提取矿物异常信息;当值为[0.4,0.7]和(0.7,1.0]时,分别采用LIC(lichenthaler index)和CTR(carter indices)方法提取植被异常信息。同时,为确保提取的矿物和植被异常信息的产生是由油砂中烃类的微渗漏所导致,以野外油砂反射光谱为端元,运用光谱角分类方法提取油砂信息,并将其与获取的矿物和植被异常信息进一步应用空间叠置分析确定油砂分布有利区。结果表明,综合运用野外实测高光谱数据和高光谱影像数据能够较准确地预测出研究区中油砂的分布位置。因此,应用高光谱影像进行油砂分布的有利区预测,可为未来利用遥感技术深入研究油砂可采储量评价提供参考依据。The research was mainly based on the principles of hydrocarbon microseepage and spectral response of oil sands composition characteristics. The spectral information related to oil sands distribution was extracted and identified from the hyperspectral image to predict the favorable reservoir for oil sands. Based on the analysis of ground characteristics of hydrocarbon microseepage caused by oil sands, the anomalous feature from low plantation coverage was primarily in mineralization anomaly, the main identification features of medium and high vegetation covering areas were vegetation anomalies. Normalized difference vegetation index （NDVI） was used to represent the different vegetation coverage degree. When NDVI is [0.0,0.4）, SAM （spectral angle method） was used to extract mineralization anomaly information. When NDVI are [0.4,0.7] and （0.7,1.0], the vegetation anomaly information were taken by using LIC（lichenthaler index） and CTR（carter indices）respectively. Meanwhile, in order to ensure the extraction of mineralization and vegetation abnormal information caused by the leakage of hydrocarbons from oil sands, the spectral reflectance of oil sands was encouraged to be the endmember to get oil sands spatial information by using SAM. Finally, spatial superimposed analysis was applied to integrate oil sands composition spatial information with mineralization and vegetation exception information for delineating the prospective areas of oil sands distribution. The results showed that a combination of field measurement hyperspectral data and hyperspectral image could predict the distribution of oil sands reservoir. So hyperspectral image plays an important role in prediction of the oil sands bearing reservoir prospective areas, it could also provide some useful information for researching into recoverable reserves evaluation of oil sands by using remote sensing technology.

关 键 词：油砂 Hyperion影像数据 烃类微渗漏 含油率 短波红外光谱 

分 类 号：P627[天文地球—地质矿产勘探]