Global Vision Media Focus

The company was featured on CCTV a total of 15 times, and its reach through Party media, central media, and local official media reached a total of 50 million people.

0101

News Categories

Chlorophyll a concentration inversion in Hongze Lake

2026-04-18

Overview

Chlorophyll a is the main photosynthetic pigment of phytoplankton and is often used as an indicator of phytoplankton biomass, making it an important component of water environment management. Chlorophyll a concentration (Chla) is a key indicator for assessing the degree of eutrophication in water bodies and studying carbon cycling in aquatic environments; its rapid and accurate acquisition is crucial for lake ecological management.

Traditional methods for assessing lake trophic status primarily rely on in-situ sampling and analysis. However, this method is susceptible to local weather and environmental influences, and the sampling and testing process is time-consuming, labor-intensive, and costly, making it difficult to monitor lake eutrophication at fine spatiotemporal scales. In contrast, remote sensing technology offers advantages such as wide coverage, rapid data acquisition, and short processing times, overcoming many shortcomings of traditional water quality sampling. Therefore, it has been widely applied to the monitoring of lake water environment and aquatic ecology.

data

Data source:

(1) Sampling point data: In this study, a total of 19 valid sampling point data were collected and analyzed. These data points were strictly screened and quality controlled to ensure the accuracy and reliability of subsequent analysis.

(2) Remote sensing data: The Level-2A standard product provided by the Sentinel-2 multispectral instrument (MSI) was used as the main remote sensing data source. This data product covers 10 key spectral bands, including bands B1 to B9 and band B8A, which can comprehensively capture the spectral characteristics of surface water bodies. All remote sensing data were efficiently acquired and processed through the Google Earth Engine platform.

Data processing:

(1) Data cleaning: First, remove data points with chlorophyll a concentration (Chl-a) less than or equal to 0 and missing values to ensure the validity and accuracy of the data.

(2) Water body extraction: The Normalized Water Index (NDWI) is applied, and the calculation formula is (B3 - B8) / (B3 + B8). A water body mask is generated by setting a threshold greater than 0.1, thereby accurately extracting the water body area and eliminating the interference of non-water body pixels.

(3) Spatiotemporal matching processing: Spatially, the average reflectance value of the 3x3 pixel window around each sampling point is extracted to reduce the impact of pixel-level noise and spatial heterogeneity; temporally, precise matching is performed based on the actual sampling date to ensure that the remote sensing data and the field observation data are highly consistent in time, thereby improving the spatiotemporal accuracy of the analysis.

technical route

Figure 1. Technology Roadmap

method

In water quality parameter inversion, identifying sensitive bands closely related to these parameters and using them as input factors for the model leads to higher prediction accuracy. This study utilizes Pearson correlation analysis to identify sensitive bands closely related to chlorophyll a concentration. The Pearson correlation coefficient, a linear correlation coefficient, reveals the strength of the correlation between different bands or band combinations and chlorophyll a concentration, allowing for the elimination of weakly correlated bands that might interfere with the model's establishment. Since the reflectivity of water bodies is primarily located in the visible and near-infrared bands, and studies have shown that visible and near-infrared reflectivity can be successfully used to invert chlorophyll a concentration in water bodies.

Based on existing measured data and satellite imagery data, this study compares the correlation coefficients of different single bands and band combinations. The band combination with the highest correlation is used as the independent variable, and the water quality parameter chlorophyll a concentration is used as the dependent variable. A statistical regression model is constructed to retrieve water quality data. The accuracy of the model is tested using the coefficient of determination R2 as the evaluation standard, and the final retrieval model is obtained.

Results and Analysis

Based on the above inversion model, the distribution map of chlorophyll a concentration in Hongze Lake water at the sampling points on that day was obtained:

Figure 2.Chlorophyll a concentration in Hongze Lake (mg/L)

As shown in the figure, the overall chlorophyll a (Chl-a) concentration in Hongze Lake, retrieved based on Sentinel-2 inversion, is at a low trophic level, with a maximum concentration of only 0.0145 mg/L. This aligns with the seasonal characteristic of phytoplankton biomass being suppressed by low temperatures and weak light during winter. Spatially, a distinct heterogeneous pattern emerges, characterized by a distribution pattern of "low concentration in the lake center and high concentration at the edges": the open waters in the lake center generally exhibit low Chl-a concentrations, indicating relatively clean water; while along the shore, especially at river mouths and in shallow bays near nearby towns, localized high-value patches appear, indicating frequent human activity and a potential impact on eutrophication. Overall, the inversion occurred in January (winter), when the water temperature of Hongze Lake was low, which suppressed the overall growth rate of algae. As a result, the Chla concentration in the main lake area was at a low level throughout the year. However, the shoreline was affected by the "point source" influence of land-based input, which still formed a local high-value area. This is consistent with the seasonal distribution pattern of Chla concentration in Hongze Lake, which is "high near the shore in winter and spring and low in the core area".