Hyperspectral remote sensing of chlorophyll-a in the Xuanwu Lake
Yan Li, Zhaofu Li
Abstract
Yan Li, Zhaofu Li
Abstract
The concentration of chlorophyll-a can reflect water quality to some extent. An approach for the determination of chlorophyll-a from field reflectance spectra was presented in the Xuanwu Lake. The reflectance spectra was measured from April to May in 2009 with FieldSpec 3 portable spectrometer, and water samples were collected directly from the lake to be analyzed in lab. Using correlations between the ground-truth data and combinations of spectral bands from the field spectral data, spectral indices including single band, band ratios, and the first-derivative models were developed which could be used to estimate chlorophyll-a. The results show that single band, band ratios and the first-derivative models have a better correlation with chlorophyll-a. But the model validation results show that coefficient of determination of the band ratios and the first-derivative models are higher (R2=0.7835 and R2=0.5913 with significance level P<0.01), and the single band model is the lowest (R2=0.4748). Therefore, band ratios and the first-derivative models can be used to indicate chlorophyll-a concentration of Xuanwu Lake, and the band ratios model is better than first-derivative model.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The concentration of chlorophyll-a can reflect water quality to some extent. An approach for the determination of chlorophyll-a from field reflectance spectra was presented in the Xuanwu Lake. The reflectance spectra was measured from April to May in 2009 with FieldSpec 3 portable spectrometer, and water samples were collected directly from the lake to be analyzed in lab. Using correlations between the ground-truth data and combinations of spectral bands from the field spectral data, spectral indices including single band, band ratios, and the first-derivative models were developed which could be used to estimate chlorophyll-a. The results show that single band, band ratios and the first-derivative models have a better correlation with chlorophyll-a. But the model validation results show that coefficient of determination of the band ratios and the first-derivative models are higher (R2=0.7835 and R2=0.5913 with significance level P<0.01), and the single band model is the lowest (R2=0.4748). Therefore, band ratios and the first-derivative models can be used to indicate chlorophyll-a concentration of Xuanwu Lake, and the band ratios model is better than first-derivative model.
Key concepts: Hyperspectral imaging, Derivative (finance), Remote sensing, Chlorophyll, Chlorophyll a, Spectral line, Spectrometer, Spectral bands