Study On Radiative Transfer Mode And The Remote Sensing Estimation Of Offshore Oil Thickness Information

Ocean hydrocarbon remote sensing is a great significance in monitoring of hydrocarbon seepage in sea surface and estimates the total hydrocarbon volume which distributed in sea. The empirical approach often applies to retrieval of hydrocarbon thickness. However, the disadvantages of empirical algorithms are regional or seasonal effects, indeterminate of information mechanism. The complex optical characteristics of water background material and hydrocarbon itself arouse their great regional and temporal variations of detection methods. Therefore, the optical characteristics of water background material and their correlations have to be comprehend well when conducting hydrocarbon remote sensing retrieval model. In order to accurately retrieve hydrocarbon thickness in sea, we need know something more about optics properties of ocean hydrocarbon and the components of water and also the law of light radiative transfer in water, and establish the radiative transfer model and the retrieval model about hydrocarbon thickness according to the area of Liaodongwang sea. Based on these reasons, we discuss about inverse problems by experimental analysis indoor and in site optics investigation in the area of Liaodongwang sea.Based on Aqua-color Radiative transfer theory, we builds parameterization model about hydrocarbon in the water with the characteristical of hydrocarbon and water's absorption and scattering behavior. And through the optical parametric studies of oil and water color components,the studies of radiative transfer model and the semi-analytical algorithm of oil to analyze. The main conclusions of the research are as follows:(1) The thin film radiative transfer parametric model is established according to the water-color radiative transfer theory, and simulating the reflectance variation of different oil thickness. Reflectivity in the visible band would decline by 75%and 50% respectively in pure water and high turbidity water environment, Through simulate the change of oil thickness from 0 to 120nm. The oil attenuation effect is greater than the scattering effect, indicating that the short-wave blue band are more favorable for exploring oil. The main scope of the model is under the conditions of chlorophyll and suspended matter in low concentration. and the water optics parameters are difficult to obtain with high chlorophyll and suspended matter concentrations.(2) Based on the measured data in the area of Liaodongwang sea and experimental data indoor, we analyzed the remote sensing reflectivity date of pure water indoor and the seawater of Liaodongwang which covered with hydrocarbon, and the absorption and scattering properties of water body. Summarized the information of different spectral characteristics of the hydrocarbon film thickness and the spectral response of the sensitive bands. Indoor experiment of hydrocarbon spectral shows that the spectral in the near-infrared spectral range (450-800 nm) is the oil film sensitive response spectral range band with film thickness less than 1μm range, water surface spectral reflectance covered by oil film is slightly lower than normal water, And with the increase of oil thickness decreased rapidly in the visible band.Using the spectrum data of water and oil measured in laboratory, the correlation relationships between oil thickness and corresponding bands reflectance were analyzed. The bands with high correlation coefficients were selected as sensitive bands to conduct statistical regression model and semi-analytical model, And Simulation the hydrocarbon remote sensing reflectance on water surface, the inversion results of the statistics and Semi-analytical model showed highest responsibility for the hydrocarbon thickness in the bands of visible(500nm and 580nm). The R2(mulplex correlation coefficient) of sigle-band and Dual-band ratio exponential regression can be achieved as 0.94 and 0.96; while the R2 of Levenberg-Marquardt Semi-analytical algorithm regression is 0.98 and it can better to keep the measured spectrum shape.(3) Based on the radiative-transfer theory models and the remote sensing retrieval Levenberg-Marquardt algorithm, the remote sensing retrieval of hydrocarbon inherent optical characteristics and concentrations in the Liangdong Bay was achieved by this model. The results showed that the simulation reflectivity very close to the measured spectra. The retrieval precision of oil and suspended sediment concentration is high, while the retrieval precision of Chlorophyll concentration is low with high Chlorophyll concentration area. The retrieval results show the average error of suspended sediment concentration is 21.8%; the average error of Chlorophyll concentration is 42.5%;and the average error of oil concentration is 17.8%. With 579nm as main sensitive band, semi-empirical modes among apparent optical parameters and hydrocarbon concentrations were simulated to obtain the remote sensing retrieval of hydrocarbon apparent optical parameters and their concentrations.