Method for estimating lake transparency from msi

Abstract

The application relates to a lake transparency estimation method, and belongs to the field of satellite remote sensing and application thereof. The application aims to solve the technical problem that the existing lake transparency test cannot realize continuous monitoring in a large-scale range. The method comprises the following steps: the normalized water-leaving reflectance combination model of the MSI sensor of the Sentinel-2 satellite is used to estimate the transparency of Chinese lakes; the normalized water-leaving reflectance data of the red edge band and the red band are substituted into the following formula to estimate the water transparency: SDD=(rhown5 / rhown4-0.799) / (0.134). The method can realize long-term high-precision monitoring of lake transparency, and can compensate for the defect that the existing lake transparency test cannot realize continuous monitoring in a large-scale range. The model effect of the lake transparency estimation model is good, and the lake transparency can be accurately estimated.

Description

Technical Field

[0001] This invention relates to a method for estimating lake transparency, belonging to the field of satellite remote sensing and its applications. Background Technology

[0002] In water quality monitoring, transparency (Secchi Disk Depth, SDD) is a relatively intuitive indicator and one of the fundamental parameters describing the optical properties of water bodies. It can also be used to assess the eutrophication status of lakes. Transparency measurement refers to the maximum depth that a secchi dart placed vertically in the water can see. It reflects the light transmittance of the water body; the higher the transparency, the clearer the water. Lake transparency exhibits significant spatiotemporal dynamics, influenced by many environmental and anthropogenic factors. Different lakes at the same time, or even the same lake in different seasons, show significant spatiotemporal heterogeneity. Existing lake and reservoir transparency tests cannot achieve continuous monitoring on a large scale. Transparency alters the optical radiation transmission of the underwater light field, severely impacting the growth of submerged vegetation and the life activities of other organisms that rely on visible light for survival. Therefore, obtaining more comprehensive water transparency data has positive practical application value for understanding changes in lake aquatic ecosystems, primary productivity of water bodies, and environmental monitoring and management. Summary of the Invention

[0003] The purpose of this invention is to solve the technical problem that existing lake and reservoir transparency tests cannot achieve continuous monitoring on a large scale, and to provide a lake transparency estimation method based on MSI.

[0004] MSI's lake transparency estimation method follows these steps:

[0005] The transparency of Chinese lakes is estimated using the normalized water reflectance combination model of the Sentinel-2 satellite MSI sensor, which is the ratio of the red edge band to the red band.

[0006] Substitute the normalized water reflectance data of the red-edge band and the red band into the following formula to estimate the water transparency:

[0007] SDD=(rhown5 / rhown4 -0.799)÷(-0.134)

[0008] Where rhown4 represents the normalized water reflectance data of the 4th band (red band) of the MSI sensor, rhown5 represents the normalized water reflectance data of the 5th band (red edge band) of the MSI sensor, and SDD represents the water transparency.

[0009] The water body is a lake or reservoir.

[0010] This invention estimates water transparency by acquiring the normalized water reflectance of the target lake's MSI sensor in bands 4 and 5, and then calculating the water transparency using a formula. This invention provides a simple MSI-based method for estimating lake transparency, with the calculated transparency value showing good agreement with measured values ​​and high reliability. It can be used in water quality monitoring and evaluation, ecological restoration performance assessment, and other fields. The accuracy of the water transparency calculated by this method meets the internationally accepted remote sensing inversion accuracy requirement (<45%), demonstrating good accuracy and strong practicality.

[0011] The method of this invention enables long-term, high-precision monitoring of lake transparency, overcoming the limitation of existing lake and reservoir transparency tests that cannot achieve continuous monitoring over large scales. Verification using a measured sample point model shows that the model of this invention for estimating lake and reservoir transparency performs well and can accurately estimate lake transparency. Attached Figure Description

[0012] Figure 1 This is a sampling distribution map of 45 lakes and reservoirs across the country in Experiment 1;

[0013] Figure 2 This is a correlation analysis diagram between the estimated water transparency and the measured water transparency in Experiment 1. Detailed Implementation

[0014] The technical solution of the present invention is not limited to the specific embodiments listed below, but also includes any combination of the specific embodiments.

[0015] Specific Implementation Method 1: The lake transparency estimation method of MSI in this implementation method is carried out according to the following steps:

[0016] The transparency of Chinese lakes is estimated using the normalized water reflectance combination model of the Sentinel-2 satellite MSI sensor, which is the ratio of the red edge band to the red band.

[0017] Substitute the normalized water reflectance data of the red-edge band and the red band into the following formula to estimate the water transparency:

[0018] SDD=(rhown5 / rhown4 -0.799)÷(-0.134)

[0019] Where rhown4 represents the normalized water reflectance data of the 4th band (red band) of the MSI sensor, rhown5 represents the normalized water reflectance data of the 5th band (red edge band) of the MSI sensor, and SDD represents the water transparency.

[0020] Specific Implementation Method Two: This implementation method differs from Specific Implementation Method One in that the water body is a lake or reservoir. Everything else is the same as in Specific Implementation Method One.

[0021] The following experiments were used to verify the effectiveness of the invention:

[0022] Experiment 1:

[0023] MSI's lake transparency estimation method follows these steps:

[0024] From 2017 to 2019, water transparency measurements were conducted at 45 lakes and reservoirs across the country. The distribution map of the sampling points is shown below. Figure 1 As shown, 2-4 sampling points were set up for each water body, for a total of 431 sampling points;

[0025] The normalized ionized water reflectance of the sampled lake MSI sensor in bands 5 and 4 was obtained. The sampled lake MSI data was filtered out to remove data values ​​affected by clouds, mountain shadows, land and aerosols to avoid interference with the normalized ionized water reflectance.

[0026] Substitute the normalized reflectance data from band 5 and band 4 into the following formula to estimate water transparency:

[0027] SDD=(rhown5 / rhown4 -0.799)÷(-0.134)

[0028] Where rhown4 represents the normalized water reflectance data of the 4th band (red band) of the MSI sensor, rhown5 represents the normalized water reflectance data of the 5th band (red edge band) of the MSI sensor, and SDD represents the water transparency.

[0029] The estimated transparency of each lake is obtained, and the estimated transparency is linearly fitted to the measured transparency to plot the results. Figure 2 .

[0030] Microsoft Excel 2017 software was used to perform a correlation analysis between the water transparency estimated according to Experiment 1 and the measured water transparency. The measured water transparency was plotted on the x-axis, and the water transparency calculated according to Experiment 1 was plotted on the y-axis. The results show that the measured data points are evenly distributed on both sides of the line y = 1.0186x + 0.0487, and R... 2 =0.79, N=431, p<0.001, RMSE and MAE are 0.52m and 0.41m respectively. The water transparency calculated by the method of this invention has a fitting accuracy with the measured transparency, which meets the internationally accepted remote sensing inversion accuracy requirements (<45%), and has good accuracy and strong practicality.

Claims

1. A method of estimating lake transparency of MSI, characterized by The lake transparency estimation method of the MSI is performed according to the following steps: The lake transparency estimation in China is performed by using a normalized difference water-leaving reflectance combination model of the MSI sensor of the Sentinel-2 satellite, and the normalized difference water-leaving reflectance combination model is a red edge band and red band ratio; The red edge band and red band normalized water-leaving reflectance data are substituted into the following formula to estimate the water transparency: SDD = (rhown5 / rhown4-0.799) / (-0.134) where rhown4 represents the MSI sensor 4th band normalized water-leaving reflectance data, rhown5 represents the MSI sensor 5th band normalized water-leaving reflectance data, and SDD represents the water transparency.

2. The method of claim 1, wherein The water body is a lake or a reservoir.

Images