论文部分内容阅读
悬浮泥沙定量研究对于调查长江的水质、地貌、生态环境等起着至关重要的作用。以长江中游武汉地区2012~2013年14幅不同时相的Landsat ETM+遥感影像为主要数据源,结合野外采样悬浮泥沙浓度数据,分析了悬浮泥沙遥感定量反演方法,数据处理中针对ETM+SLC-OFF影像缝隙问题,采用自适应局部回归匹配算法(ALR)进行影像自动恢复处理,在波段选择中对悬浮泥沙浓度和光谱反射率数据进行相关性分析,并运用传统关系建模方法和高斯模型方法对比,比较悬浮泥沙定量反演模型,利用实测验证数据对反演模型精度进行评估。研究结果表明:(1)ALR可以有效的获取悬浮泥沙敏感波段的遥感光谱反射率;(2)ETM+Band3悬浮泥沙浓度的高斯模型相关系数最高,通过对比得到模型反演的验证精度较高,研究证明遥感定量反演适合于长江流域武汉段泥沙含量大范围监测。
The quantitative study of suspended sediment plays a crucial role in the investigation of the water quality, topography and ecological environment of the Yangtze River. Taking the 14 different phases of Landsat ETM + remote sensing images from Wuhan in the middle reaches of the Yangtze River valley from 2012 to 2013 as the main data sources and the data of suspended sediment concentrations in the field, the quantitative inversion method of suspended sediment remote sensing was analyzed. SLC-OFF image gap, using adaptive local regression matching algorithm (ALR) for image restoration, correlation analysis of suspended sediment concentration and spectral reflectance data in the band selection, and the use of traditional relational modeling methods and Gaussian model method to compare the suspended sediment quantitative inversion model, the use of measured data to evaluate the accuracy of the inversion model. The results show that: (1) ALR can effectively obtain remote sensing spectral reflectance of suspended sediment sensitive band; (2) The correlation coefficient of Gaussian model of suspended sediment concentration of ETM + Band3 is the highest. High, the research proves that remote sensing quantitative inversion is suitable for a large-scale monitoring of sediment concentration in the Wuhan section of the Yangtze River.