Remote Sensing Survey of Chlorophyll-a Concentration in River-Type Reservoirs Based on GF-1 Image Data: Taking Feilaixia–Changhu Reservoir Area as an Example
By taking the Feilaixia–Changhu Reservoir area in Guangdong Province, China as an example, the wide field view (WFV) multispectral image data of China's Gaofen-1 (spatial resolution of 16 m) was used, and the remote sensing physical analysis model for water quality (Deng model) based on radiati...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Editorial Office of Pearl River
2025-01-01
|
| Series: | Renmin Zhujiang |
| Subjects: | |
| Online Access: | http://www.renminzhujiang.cn/thesisDetails?columnId=122607476&Fpath=home&index=0 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | By taking the Feilaixia–Changhu Reservoir area in Guangdong Province, China as an example, the wide field view (WFV) multispectral image data of China's Gaofen-1 (spatial resolution of 16 m) was used, and the remote sensing physical analysis model for water quality (Deng model) based on radiation transfer theory was used to perform remote sensing inversion of the chlorophyll-a concentration in the river-type reservoir. The original image was subjected to a series of image preprocessing steps such as the improved dark-pixel atmospheric correction method for inland water and the water and land separation based on logical discrimination. From the preprocessed water-leaving reflectance result map, three types of single-type special pixels, namely clean water, organically polluted water, and eutrophic water, were selected to first calculate the optical parameters of the water components required by the Deng model. Finally, the remote sensing result map of the chlorophyll-a concentration in the study area was obtained by solving the equation group by combining remote sensing bands. The correlation analysis of the chlorophyll-a remote sensing inversion value and the measured value was performed, and the determination coefficient was R<sup>2 </sup>= 0.835 8; the result verification was reasonable. The chlorophyll-a remote sensing results reveal that there are several places with serious eutrophication in this area at the time of satellite image imaging, mainly distributed in two reservoir centers and one reservoir bay of Feilaixia Reservoir; two river bays of Lianjiang River; one in the upper reaches of Wengjiang River, and one in Yingde City section of the main stream of Beijiang River. It is inferred that it is caused by the discharge of sewage from cage aquaculture and the catering industry to nearby residents. The high-concentration pollution areas in the reservoir center of Feilaixia Reservoir and the main stream are temporary, because the hydrodynamic conditions are formed with the migration of river diffusion; while the pollutants located in the reservoir bay and river bay are gathered into stable pollution areas due to their poor flow conditions, indicating that the Deng model has also achieved good results under dynamic hydrological conditions. Further statistics show that the water area with chlorophyll-a concentration of 0–3 μg/L accounts for 92.05% (56.21 km<sup>2</sup>), and the pollution area with >3 μg/L accounts for 7.95% (4.85 km<sup>2</sup>), indicating that the chlorophyll-a concentration in the study area is generally at a low level, and the statistical results are consistent with the laws reflected by the image inversion results. This study not only expands the application of the Deng model in river-type reservoirs but also provides a scientific basis for regional water environment management. |
|---|---|
| ISSN: | 1001-9235 |