Research On Spectral Response Mechanisms Of Oil Sands Components And Ore Prospecting By Remote Sensing Technology

With the rapid development of economics, the demand for oil gas has increasedcontinuously, exploitation and utilization of non-conventional oil and gas astransitional replacement for energy and resource. Oil sands have the contagiousdistribution and huge storage, and its mining technique is most mature, which will bewidely used. The oil sand resources is rich in China, the detection of the distributionof oil sands not only improve tense situation of the soil and gas resource, but alsoenhance the effective supply of energy and resource, which guarantee resource safetyand have significantly economical and political meanings.Remote sensing identifies ground objects by reflection and radiation in differentelectromagnetic radiation information area, which has been widely used inconventional investigation of gas and oil storage. Oil sand s have common features ofoil and gas like hydrocarbon microseepage, also have its own features. The spectraanomaly caused by physicochemical property and ecological features resulting fromhydrocarbon microseepage, and the abnormal information can be delineated byremote sensing. Diagnostic absorbing feature of mineral and hydrocarbon collected inthe field can be analyzed by hyperspectral technology, and also background of surfacein occurrence region of oil sands can be determined. Preliminary research on therelationship between spectra of oil sands and oil content, porosity, permeability couldprovide a reference direction for using remote sensing to assess reservoir physicalproperty.Main research results are listed as follows, from the spectral responsemechanisms of oil sands components and the oil sands surface anomaly informationextraction method on the basis of multispectral image, to research on ore prospectingby remote sensing technology and the relationship between oil sands spectrum andoiliness, rervoir physical property: (1) Multi-spectra extraction mode for oil sands without prior knowledge wasestablished.In the geological background of oil sands may be occurrence, according to thelevel of vegetation coverage criteria, with the normalized difference vegetation indexobtaining by multispectral image computed vegetation coverage characterization ofthe surface vegetation coverage, when the vegetation coverage is less than0.3, thevisible near infrared anomaly information extraction method indicate thedistribution of oil sands areas, when the vegetation coverage is higher than0.3, it isextracted temperature anomaly information to delineate favorable area of oil sandsdeposits for remote sensing image from7to9months by adopting the method ofatmospheric radiative transfer equation. Taken Yusupu Aleke in Ruoqiang county ofXinjiang Province and Nianzi dam–Songbaoshi district in Guangyuan of SichuanProvince, this mode has been put into practice and used for determining distributionof oil sands, which shows that this mode could indicate the oil sands distributioneffectively.(2) The extraction mode for the main alteration types in the study area byabsorption depth characterization of oil sands spectral response strength was built.The result of scanning electron microscope and field investigation shows that itis scientific to suing absorbing depth that reflect spectra response radiation of oilsands. Alteration mineral extraction mode based mineral composition was applied toMiaowan-Cuijiagou in Ordos Basin, which filtered non-sand-oil distributioninformation and indicate its distribution selectively. Taken Nianzi dam-loose stone inGuangyuan of Sichuan Basin as an example, the spectra response features areanalyzed and, the results show that this model could extract oil sands distributioneffectively.(3) Aimed at core research of conventional gas and oil and based on spectra toreflect inner composition feature, relationship between spectra and oiliness, physicalproperties are established.Pearson and Partial correlation are calculated between oil content and absorption depth of hydrocarbon spectra band falling between1740～1780nm、2300～2340nm and2340～2360nm, and oil content are estimated by the method of ULR andPLSR. Results show that, URL using absorption depths around2350nm and PLSRusing1760nm,2310nm and2350nm have higher accuracy, which could offerscientific bias to rapid calculation of oil content.Analyzing the correlation betweenaverage absorption depth and porosity, permeability, the relationship betweenspectra of oil sands and physical properties of reservoir in the background are studied,the analysis shows that there exists a negative correlation between absorption depthsof groups that stands for clay mineral content and permeability, and there exists apositive correlation between hydrocarbon groups that stand for oiliness. Therefore theresearch on relationship between spectra of oil sands and physical properties couldprovide a reference study direction for reservoir assessment and recoverable reserves.Two innovation points are involved in this thesis, listed as following:(1)As for using remote sensing to find ore, multi-spectra extraction mode for oilsands without prior knowledge are established, which adopt different extractionmethods under different vegetation and carry out research on oil and distributionwithout the knowledge of background that play a good indicative role.(2)As for spectra parameters, spectra response radiation of oil sands groups arecompared and used for identifying main alteration mineral, which could layfoundation for ore research by remote sensing.