柴达木盆地地表酸解烃含量的高光谱遥感估算

Estimating acidolysis hydrocarbon content in surface of Qaidam basin through hyperspectral remote sensing

地表烃类物质的探测对遥感油气勘探和燃料烃污染监测都具有重要的实际意义。在有天然气分布的柴达木盆地东部三湖地区获取了一景Hyperion高光谱遥感影像,贯穿影像覆盖区进行了土样和光谱采集并测定了酸解烃含量。用单变量预测模型来探讨烃类光谱特征参数与酸解烃总含量之间的线性和非线性关系,结果表明:位于烃的特征吸收波段内1699.4nm处反射率和反射率一阶微分与酸解烃总含量之间的相关性最好,该波段反射率一阶微分的乘幂函数为酸解烃总含量的最佳单变量估算模型。进一步用逐步回归分析技术探索反射率和反射率一阶微分光谱与酸解烃总含量的关系,将估算模型应用在Hyperion影像上并用实测值对估算结果进行验证,结果发现:一阶微分估算模型虽然能改善估算精度但相对误差较大。结论认为,基于反射率光谱的回归方程是酸解烃总含量的最佳预测模型。

Exploring hydrocarbon content in surface is of important practical significance for exploring oil and gas with remote sensing and for monitoring fuel hydrocarbon pollution.A Hyperion hyperspectral image was obtained in the Sanhu region in the eastern part of the Qaidam basin where gas reservoirs occur.Soil samples were collected throughout the area covered by the image,their spectra were measured in the field and content of acidolysis hydrocarbon was determined in the lab.Linear and non-linear correlations between spectral characteristic parameters of hydrocarbons and total acidolysis hydrocarbon content were investigated by using univariate regressive models.Results show that the reflectance and first-order derivative reflectance at 1699.4 nm,which is within the characteristic absorption wave band of hydrocarbon,have the best correlation with total acidolysis hydrocarbon content.The power model of first-order derivative reflectance at 1699.4 nm is the optimal univariate estimation model for total acidolysis hydrocarbon content.Stepwise regressive analysis techniques were further used to study the relationship between the original and first-order derivative reflectance data and total acidolysis hydrocarbon content,estimation equations were applied to the Hyperion image,and the measured values were used to verify the estimated results.The results show that the estimation accuracy was significantly improved with the first-order derivative spectra but still has quite large relative error.It is believed that the regressive equation of reflectance spectra is the best model for estimating the total acidolysis hydrocarbon content.