Study On Hyperspectral Remote Sensing Of Marine Oil Emulsions

In the process of weathering and migration of marine spilled oils,different types and concentrations of oil-water emulsions will be formed,which will cause significant harm to the environment.Accurate identification and quantification of different types of oil-water emulsions show great importance in emergency treatment and damage assessment of marine oil spill pollution.Crude oil and water are incompatible two-phase media,however,under the action of oceanic dynamic,oil-in-water or water-in-oil emulsion can be formed.In recent years,great progress has been made in the study of optical remote sensing of marine oil spill.The application characteristics and advantages of optical remote sensing technology in this field are becoming clear,gradually,while there is still no systematic study of oil-water emulsions.Based on the indoor simulation experiments of oil-water emulsions,this paper studies the spectral response characteristics and variation rules of oil-water emulsions with different volume concentrations,and analyses the mechanism of their spectral absorption characteristics;The quantitative estimation of oil-water emulsions concentration is carried out based on reflectance and absorption depth of spectra;The studies of hyperspectral remote sensing identification characteristics and recognition of oil-spilled emulsions are carried out.The main research contents include:(1)Laboratory simulation of oil-water emulsions was carried out to obtain samples and their reflectance spectra of different types of oil-water emulsions with different volume concentrations;The spectral response characteristics and variation rules of different types of oil-water emulsions are clarified.(2)Based on the reflectance and absorption depth obtained from spectral reflectance data of different volume concentration,the numerical statistical analysis between spectral reflectance and volume concentration of oil-water emulsions was carried out,and the quantitative estimation of volume concentration of oil-water emulsions was realized.(3)Using Airborne Hyperspectral AVIRIS data obtained from BP oil spill in the Gulf of Mexico in 2010,the spectral characteristics of different types of oil-water emulsions were verified.A hyperspectral remote sensing recognition algorithm of different oil-water emulsions based on spectral groups feature detection was developed to realize the recognition and extraction of oil-water emulsions in AVIRIS data.Based on the above research,the main conclusions are as follows:(1)Different types of oil-water emulsions have siginificant differences in spectral response.The water-in-oil emulsion has high reflectance in the range of near infrared and short wave infrared spectra,and has obvious absorption characteristics of-C-H.The spectral response characteristics of oil-in-water emulsions are mainly concentrated in the range of 400 nm to 1400 nm,and siginificant spectral characteristics are formed at the range of～970 nm and～1130 nm due to the absorption of-O-H;(2)In the water-in-oil emulsion,the concentration can be effectively estimated by using the reflectance at～1295 nm,～1655 nm,～2200 nm and the absorption depth at～1445 nm;In the oil-in-water emulsion,within the acceptable error range,the concentration can be estimated by using the reflectance at～1075 nm,～1295 nm;(3)The hyperspectral oil-spilled emulsions recognition algorithm based on spectral group feature detection can recognize and identify oil-water emulsions in AVIRIS data.