新庙泡叶绿素a浓度高光谱定量模型研究

利用吉林省新庙泡的高光谱实测数据和水质采样分析数据,尝试通过单波段、波段比值、一阶微分和峰谷间距法建立叶绿素a反演模型。结果表明:单波段光谱反射率与叶绿素a浓度的相关性较差,不宜用于该区域的叶绿素a浓度估算;680 nm和700 nm波段反射率之比、700 nm处光谱一阶微分值和两波段峰谷间距反演模型都具有较高的决定系数,分别为0.783 4、0.792 7、0.796 9,验证模型的决定系数为0.651 3、0.431 7、0.756 4,均方根误差分别为8.69μg·L-1、14.50μg·L-1、10.04μg·L-1,显著水平P<0.01。这3种方法皆可以用于新庙泡叶绿素a浓度的定量遥感,其中又以峰谷间距法为最优。

Hyperspectral Remote Sensing of Chlorophyll-a in the Lake Xinmiao, China

In order to determine the chlorophyll-a concentration of inland water using remote sensing, field hyperspectral data and water samples were collected in the Lake Xinmiao of China from May to September in 2004. Through analyzing the correlation between data measured in laboratory and hyperspectral reflectance, it was found that single band reflectance could not denote the chlorophyll-a concentration well. The reflectance peak near 700 nm and absorption vale near 680 nm were the most important features.Three regressive linear fitting models were set up respectively between chlorophyll-a concentration and other factors as follows: the ratio of reflectance peak near 700 nm to absorbance vale near 680 nm, the first derivative of reflectance near 700 nm, the distance from reflectance peak to absorbance vale. All of the models gave good results with high determination coefficients 0.7834, 0.7927 and 0.7969 respectively.The determination coefficients of testing models were 0.6513, 0.4317, 0.7564 respectively and root mean square error (RMSE) were 8.69μg·L-1, 14.50μg·L-1, 10.04μg·L-1-1with significance levelP<0.01. By comparison of RMSE and two kinds of determination coefficients, the third model was the best to predict chlorophyll-a contents in the Lake Xinmiao, China.