论文部分内容阅读
本文根据黄海现场离水辐亮度及遥感反射比，采用４种统计算法估算海水透明度（Ｓｅｃｃｈｉ Ｄｉｓｃ Ｄｅｐｔｈ）。前２种算法中海水透明度作为因变量，两不同波段辐亮度之比及两辐亮度之和与第三者之比分别作为自变量，进行一元回归分析，前者相关系数ｒ２为０．９３３５。后者相关系数ｒ２为０．９４８３。第３种算法中，所有波段的遥感反射比做因子分析形成的主因子作为自变量，透明度作为因变量进行多元回归分析，子样全相关系数ｒ２为０．９６４９。第４种算法中，遥感反射比作为自变量，透明度作为因变量进行逐步回归分析，复相关系数ｒ２为０．９８１５。计算结果表明，多元统计分析的估算精度高于一元回归分析。
In this paper, four kinds of statistical methods are used to estimate the Secchi Disc Depth according to the radiance of the water from the Yellow Sea and the reflectance of remote sensing. For the first two algorithms, the transparency of seawater was taken as the dependent variable, the ratio of the radiances of two different bands and the ratio of the sum of the two radiances to the third party were taken as independent variables respectively. The correlation coefficient r2 was 0.9335. The latter correlation coefficient r2 is 0.9483. In the third algorithm, the remote sensing reflectance of all the bands was taken as the independent variable and the transparency was taken as the dependent variable. The total correlation coefficient r2 was 0.9649. The fourth algorithm, remote sensing reflectance as an independent variable, transparency as a dependent variable stepwise regression analysis, the complex correlation coefficient r2 is 0.9815. The calculation results show that the accuracy of multivariate statistical analysis is higher than that of one-dimensional regression analysis.