Water surface chlorophyll concentration inversion method fused with atmospheric correction optimization result

By using the QA Score model to perform cell-by-cell atmospheric correction optimization in the inversion process of chlorophyll a concentration, the problem of poor adaptability of atmospheric correction algorithms under different sea areas and water bodies types is solved, the inversion accuracy and data coverage are improved, and the improvement is provided. Accurate chlorophyll a concentration estimate.

CN118980646APending Publication Date: 2024-11-19OCEAN UNIV OF CHINA +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411054110.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-02
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

The existing atmospheric correction algorithms have poor adaptability in different sea areas and water bodies types, resulting in large errors in inversion of chlorophyll a concentration on the surface of the water body, and lack effective methods for evaluating the atmospheric correction effect of optical satellite data.

Method used

The QA Score model is used to perform cell-by-cell atmospheric correction, and the satellite data is processed through multiple atmospheric correction algorithms, and the optimal spectral data is selected according to the scoring results for inversion, thereby improving the inversion accuracy of chlorophyll a concentration.

Benefits of technology

Effectively eliminate the inversion error introduced due to poor adaptability of different atmospheric correction algorithms, improve the accuracy of chlorophyll a concentration inversion and the coverage of effective data, and provide a band ratio result closer to the actual measured spectrum.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118980646A_ABST
    Figure CN118980646A_ABST
Patent Text Reader

Abstract

The invention belongs to the technical field of chlorophyll a concentration inversion, and particularly relates to a water surface chlorophyll a concentration inversion method fused with atmospheric correction optimal results. According to the method, a pixel-by-pixel atmospheric correction optimization result based on a QA Score model is fused in the process of inverting the concentration of chlorophyll a on the surface layer of a water body by using satellite data, so that an inversion error introduced due to the adaptability difference of different atmospheric correction algorithms in different sea areas is eliminated, and a band ratio result closer to an actually measured spectrum of the water body on each pixel is obtained; therefore, the water surface chlorophyll a concentration inversion precision is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of chlorophyll a concentration inversion, and in particular relates to a method for inverting chlorophyll a concentration on the surface of a water body by integrating an atmospheric correction optimization result. Background Art

[0002] Among the numerous water quality parameters, chlorophyll a concentration is the main indicator of phytoplankton biomass. It can characterize the degree of eutrophication of water bodies and provide important information for understanding the status of aquatic ecosystems and primary productivity. Therefore, it has been widely studied. Optical remote sensing satellite data has become an effective data source for estimating chlorophyll a concentration in the surface layer of water bodies due to its advantages such as large spatial coverage, many revisits, and long historical archives. The ratio of the water-leaving radiance to the downward irradiance just above the water surface is defined as the remote sensing reflectance of the water spectrum (R rs ), high-precision inversion of chlorophyll a concentration in the water surface requires accurate R rs The water-off-water radiance contributed by the ocean color satellite sensor is less than 10% of the total signal received, so the received signal needs to be corrected for the atmosphere, which is the key to obtaining high-precision R rs Data is the premise for achieving high-precision chlorophyll a concentration inversion. At present, there are many atmospheric correction algorithms that are open and commercially operated internationally. However, studies have shown that different atmospheric correction algorithms are suitable for different sea areas and water body types. The atmospheric correction algorithm with good correction performance in sea area A may not perform well in sea area B. The research results show that when inverting the surface chlorophyll a concentration of water bodies in a large range of sea areas, using only one atmospheric correction algorithm for a scene of satellite data will introduce large errors. However, there was a lack of effective methods for evaluating the atmospheric correction effect of optical satellite data in the past. In 2016, Wei et al. proposed a QA Score model that can evaluate normalized spectral data of water bodies. Since the model cannot effectively evaluate the original spectrum, it has not been widely used so far.

[0003] Most of the chlorophyll a concentration inversion models used in business operations internationally are based on the band ratio of water body spectra. Studies have shown that the QA Score model can well judge the results of the water body spectrum band ratio, that is, the higher the QA Score value, the closer the spectrum band ratio result is to the measured data. This model was proposed by Wei et al. in 2016. However, since this model cannot effectively evaluate the original remote sensing reflectance spectrum, it has not been well applied. Summary of the invention

[0004] In view of the above technical problems, the present invention provides a method for inverting the chlorophyll a concentration of the water surface by integrating the optimal results of atmospheric correction. The method incorporates a pixel-by-pixel atmospheric correction optimization step based on the QAScore model in the process of inverting the chlorophyll a concentration of the water surface using satellite data, so as to obtain a band ratio result for each pixel that is closer to the actual measured spectrum of the water body, thereby improving the inversion accuracy of the chlorophyll a concentration of the water surface.

[0005] The present invention is achieved through the following technical solutions:

[0006] The invention discloses a method for inverting chlorophyll a concentration of a water body surface by integrating the optimal result of atmospheric correction. The method incorporates the optimal result of atmospheric correction for each pixel based on the QA Score model in the process of inverting chlorophyll a concentration of a water body surface by using satellite data, so as to improve the inversion accuracy of chlorophyll a concentration of a water body surface.

[0007] Furthermore, the method comprises the following steps:

[0008] (1) m kinds of atmospheric correction algorithms are used to perform atmospheric correction processing on satellite data respectively, and the spectral data obtained by the m kinds of atmospheric correction algorithms are scored pixel by pixel through the QA Score model; the spectral data corresponding to the maximum score of each pixel point is selected as the optimal spectral data of the pixel point; wherein the number of types of atmospheric correction algorithms used m ≥ 2;

[0009] (2) Using the optimal spectral data obtained in step (1) to invert the chlorophyll a concentration in the surface layer of the water body to improve the inversion accuracy of the chlorophyll a concentration.

[0010] Furthermore, in step (1), the atmospheric correction algorithm used includes two or more of the GDPS2.0 atmospheric correction algorithm, the MUMM atmospheric correction algorithm, the NASA near-infrared atmospheric correction algorithm, the Acolite DSF atmospheric correction algorithm, the FLASSH atmospheric correction algorithm, and the Polymer atmospheric correction algorithm.

[0011] Furthermore, step (1) is specifically as follows:

[0012] Apply the selected m kinds of atmospheric correction algorithms to perform atmospheric correction processing on the optical satellite data, and obtain m spectral data corresponding to each pixel point;

[0013] The m spectral data are scored using the QA Score model to obtain corresponding m Score values, and the spectral data corresponding to the maximum Score value is taken as the optimal spectral data of the pixel point.

[0014] Furthermore, if there are multiple maximum values ​​Score, the average of the spectral data corresponding to the multiple maximum values ​​is taken as the optimal spectral data of the pixel point.

[0015] Furthermore, the m spectral data are scored by applying the QA Score model to obtain corresponding m Score values, which specifically includes the following steps:

[0016] S1: The remote sensing reflectance R measured on site rs The (λ) spectrum was normalized to obtain the reference spectrum nR of the normalized remote sensing reflectance of 23 optical water bodies in nine bands. rs (λ);

[0017]

[0018] R rs (λ i ) indicates the wavelength is λ i The corresponding remote sensing reflectance when i is 1-9; the wavelengths corresponding to λ1 to λ9 are 412, 443, 488, 510, 531, 547, 555, 667, and 678 respectively; N represents the total number of bands, N = 9; nR rs The value range of (λ) is between 0 and 1.

[0019] S2: By reference spectrum nR rs (λ) Get the upper boundary nR of each spectral band i of each water body rs U (λ i ) and the lower boundary nR rs L (λ i )value;

[0020] S3: Each of the m spectral data corresponding to each pixel point is taken as the target spectrum R to be scored. rs *(λ'), use the QA Score model for scoring:

[0021] S31: Determine the target spectrum R rs *(λ') band and reference spectrum nR rs Whether the band of (λ) matches;

[0022] S32: If the target spectrum R rs *(λ') has a spectral band that is smaller than the reference spectrum nR rs (λ) has many spectral bands, that is, the target spectrum R rs *(λ') has a spectral band greater than 9; only select the target spectrum R rs *(λ') with reference spectrum nR rs (λ) The bands with the same wavelength proceed to step S33;

[0023] If the target spectrum R rs*(λ') has a spectral band that is smaller than the reference spectrum nR rs (λ) has few spectral bands, that is, the total number of bands is less than 9, and all target spectra R rs *(λ') all proceed to step S33;

[0024] From the reference spectrum nR rs (λ) Extract the nR corresponding to the wavelength λ' rs (λ') subset and the nR rs The upper boundary of the spectrum corresponding to the (λ') subset nR rs U (λ') subset and lower bound nR rs L (λ') subset;

[0025] S33: Target spectrum R rs *(λ') is normalized to obtain the normalized target spectrum nR rs *(λ'):

[0026]

[0027] According to the SAM spectrum similarity formula, the normalized target spectrum nR rs *(λ') Assign water body type:

[0028]

[0029] α is the reference spectrum nR rs (λ) and the normalized target spectrum nR rs *(λ), SAM is calculated by taking the reference spectrum nR rs (λ) and the normalized target spectrum nR rs *(λ) is used as a spatial vector with a dimension equal to the total number of bands N of the target spectrum to determine the spectral similarity, and the spectral water type corresponding to the largest cosine value is determined as the normalized target spectrum nR rs *(λ) water body type;

[0030] S4: By normalizing the target spectrum nR rs *(λ) is compared with the upper and lower boundary values ​​of the water type spectrum to estimate the QA score:

[0031]

[0032] C(λ i ) is the score for a specific wavelength, N is nR rs *(λ) total number of bands;

[0033] If nR rs *(λ i) value in nR rs (λ i )’s upper boundary nR rs U (λ i )) or lower boundary nR rs L (λ i ), the corresponding wavelength score will be assigned 0, that is, C(λ i )=0; if nR rs *(λ i ) value in nR rs (λ i )’s upper boundary nR rs U (λ i )) and the lower boundary nR rs L (λ i ), the corresponding wavelength score will be assigned a value of 1, that is, C(λ i )=1;

[0034] S5: Repeat steps S3-S4 to obtain m spectral data R for each pixel point. rs_1 (λ),R rs_2 (λ),...,R rs_i (λ),...,R rs_m (λ) Apply the QA Score model to score and get m Score values: Score(R rs_1 (λ)),Score(R rs_1 (λ)),...,Score(R rs_i (λ)),...,Score(R rs_m (λ)).

[0035] Furthermore, in step (2), the OC2 algorithm, the YOC algorithm, the OC3G algorithm, or the OC2M-HI algorithm is used to invert the chlorophyll a concentration in the surface layer of the water body.

[0036] Beneficial technical effects of the present invention:

[0037] The present invention proposes for the first time to incorporate a pixel-by-pixel atmospheric correction optimization step into the process of inverting the chlorophyll a concentration in the water surface layer using satellite data, so as to eliminate the inversion error introduced by different atmospheric correction algorithms due to the differences in adaptability in different sea areas, thereby obtaining a more accurate chlorophyll a concentration inversion result. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a flow chart of a method for inverting chlorophyll a concentration on the surface of a water body incorporating the optimal results of atmospheric correction in an embodiment of the present invention;

[0039] Figure 2a The spatial distribution result of chlorophyll a concentration corresponding to GDPS2.0 atmospheric correction algorithm;

[0040] Figure 2b The spatial distribution result of chlorophyll a concentration corresponding to the MUMM atmospheric correction algorithm;

[0041] Figure 2c The spatial distribution results of chlorophyll a concentration obtained after QA Score optimization of two atmospheric correction algorithms (GDPS2.0 atmospheric correction algorithm and MUMM atmospheric correction algorithm) in the embodiment of the present invention.

[0042] Figure 3a The comparison results of chlorophyll a concentration around Shandong Peninsula obtained by GOCI inversion using GDPS2.0 atmospheric correction algorithm and chlorophyll a concentration at measured stations;

[0043] Figure 3b The comparison results of chlorophyll a concentration around Shandong Peninsula obtained by GOCI inversion using the MUMM atmospheric correction algorithm and the chlorophyll a concentration at the measured stations;

[0044] Figure 3c The comparison results of the chlorophyll a concentration around the Shandong Peninsula obtained by GOCI inversion using two atmospheric correction algorithms (GDPS2.0 atmospheric correction algorithm and MUMM atmospheric correction algorithm) after QA Score optimization in the embodiment of the present invention and the chlorophyll a concentration at the measured stations. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0046] On the contrary, the present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention as defined by the claims. Further, in order to make the public have a better understanding of the present invention, some specific details are described in detail in the detailed description of the present invention below. Those skilled in the art can fully understand the present invention without the description of these details.

[0047] The present invention provides an embodiment of a method for inverting chlorophyll a concentration on the surface of a water body by integrating an optimal result of atmospheric correction. The method incorporates the optimal result of atmospheric correction for each pixel based on a QA Score model in the process of inverting chlorophyll a concentration on the surface of a water body using satellite data, so as to improve the inversion accuracy of chlorophyll a concentration on the surface of the water body.

[0048] In this embodiment, if Figure 1As shown, the method comprises the following steps:

[0049] (1) m kinds of atmospheric correction algorithms are used to perform atmospheric correction processing on satellite data respectively, and the spectral data obtained by the m kinds of atmospheric correction algorithms are scored pixel by pixel through the QA Score model; the spectral data corresponding to the maximum score of each pixel point is selected as the optimal spectral data of the pixel point; wherein the number of types of atmospheric correction algorithms used m ≥ 2;

[0050] (2) Using the optimal spectral data obtained in step (1) to invert the chlorophyll a concentration in the surface layer of the water body to improve the inversion accuracy of the chlorophyll a concentration.

[0051] In step (1) of this embodiment, the atmospheric correction algorithm used includes two or more of the GDPS2.0 atmospheric correction algorithm, the MUMM atmospheric correction algorithm, the NASA near-infrared atmospheric correction algorithm, the Acolite DSF atmospheric correction algorithm, the FLASSH atmospheric correction algorithm, and the Polymer atmospheric correction algorithm.

[0052] In this embodiment, step (1) is specifically as follows:

[0053] Apply the selected m kinds of atmospheric correction algorithms to perform atmospheric correction processing on the optical satellite data, and obtain m spectral data corresponding to each pixel point;

[0054] The m spectral data are scored using the QA Score model to obtain corresponding m Score values, and the spectral data corresponding to the maximum Score value is taken as the optimal spectral data of the pixel point.

[0055] In this embodiment, if there are multiple maximum values ​​Score, the average of the spectral data corresponding to the multiple maximum values ​​is taken as the optimal spectral data of the pixel point.

[0056] In this embodiment, the m spectral data are scored by applying the QA Score model to obtain the corresponding m Score values, which specifically includes the following steps:

[0057] S1: The remote sensing reflectance R measured on site rs The (λ) spectrum was normalized to obtain the reference spectrum nR of the normalized remote sensing reflectance of 23 optical water bodies in nine bands. rs (λ);

[0058]

[0059] R rs (λ i ) indicates the wavelength is λ iThe corresponding remote sensing reflectance when i is 1-9; the wavelengths corresponding to λ1 to λ9 are 412, 443, 488, 510, 531, 547, 555, 667, and 678 respectively; N represents the total number of bands, N = 9; nR rs The value range of (λ) is between 0 and 1.

[0060] nR rs (λ) retains the original R rs (λ) spectrum (i.e., the remote sensing reflectance R measured on site) rs The spectral shape, band ratio and original R rs Specifically, in this step, 958 sets of on-site R rs The (λ) spectral data were normalized, and the normalized remote sensing reflectance spectra of 23 optical water bodies in nine bands were obtained as reference spectra.

[0061] S2: By reference spectrum nR rs (λ) Get the upper boundary nR of each spectral band i of each water body rs U (λ i ) and the lower boundary nR rs L (λ i )value;

[0062] S3: Each of the m spectral data corresponding to each pixel point is taken as the target spectrum R to be scored. rs *(λ'), use the QA Score model for scoring:

[0063] S31: Determine the target spectrum R rs *(λ') band and reference spectrum nR rs Whether the band of (λ) matches;

[0064] S32: If the target spectrum R rs *(λ') has a spectral band that is smaller than the reference spectrum nR rs (λ) has many spectral bands, that is, the target spectrum R rs *(λ') has a spectral band greater than 9; only select the target spectrum R rs *(λ') with reference spectrum nR rs (λ) The bands with the same wavelength proceed to step S33;

[0065] If the target spectrum R rs *(λ') has a spectral band that is smaller than the reference spectrum nR rs (λ) has few spectral bands, that is, the total number of bands is less than 9, and all target spectra R rs*(λ') all proceed to step S33;

[0066] From the reference spectrum nR rs (λ) Extract the nR corresponding to the wavelength λ' rs (λ') subset and the nR rs The upper boundary of the spectrum corresponding to the (λ') subset nR rs U (λ') subset and lower bound nR rs L (λ') subset;

[0067] S33: Target spectrum R rs *(λ') is normalized to obtain the normalized target spectrum nR rs *(λ'):

[0068]

[0069] According to the SAM (Spectral Angle Mapper) spectrum similarity formula, the normalized target spectrum nR rs *(λ') Assign water body type:

[0070]

[0071] α is the reference spectrum nR rs (λ) and the normalized target spectrum nR rs *(λ), SAM is calculated by taking the reference spectrum nR rs (λ) and the normalized target spectrum nR rs *(λ) is used as a spatial vector with a dimension equal to the total number of bands N of the target spectrum to determine the spectral similarity. The spectral water type corresponding to the maximum cosine value is determined as the normalized target spectrum nR rs *(λ) water body type;

[0072] S4: By normalizing the target spectrum nR rs *(λ) is compared with the upper and lower boundary values ​​of the water type spectrum to estimate the QA score:

[0073]

[0074] C(λ i ) is the score for a specific wavelength, N is nR rs *(λ) total number of bands;

[0075] If nR rs *(λ i ) value in nR rs (λ i )’s upper boundary nRrs U (λ i )) or lower boundary nR rs L (λ i ), the corresponding wavelength score will be assigned 0, that is, C(λ i )=0; if nR rs *(λ i ) value in nR rs (λ i )’s upper boundary nR rs U (λ i )) and the lower boundary nR rs L (λ i ), the corresponding wavelength score will be assigned a value of 1, that is, C(λ i )=1;nR rs *(λ)’s total score varies in the range of [0,1];

[0076] S5: Repeat steps S3-S4 to obtain m spectral data R for each pixel point. rs_1 (λ),R rs_2 (λ),...,R rs_i (λ),...,R rs_m (λ) Apply the QA Score model to score and get m Score values: Score(R rs_1 (λ)),Score(R rs_1 (λ)),...,Score(R rs_i (λ)),...,Score(R rs_m (λ)).

[0077] In step (2) of this embodiment, the OC2 algorithm, YOC algorithm, OC3G algorithm, or OC2M-HI algorithm is used to invert the chlorophyll a concentration in the surface layer of the water body. Specifically, any method that uses band ratios to invert the chlorophyll a concentration can be applied to the method provided by the present invention to improve the inversion accuracy.

[0078] The present invention takes the sea area around Shandong Peninsula as an example, selects GOCI satellite data, applies two atmospheric correction algorithms (GDPS2.0 atmospheric correction algorithm and MUMM atmospheric correction algorithm) to perform atmospheric correction processing on it, and uses the QA Score model to perform pixel-by-pixel light scoring on the spectral data obtained by the two atmospheric correction algorithms. For each pixel point, the spectral data corresponding to the maximum score is selected as the preferred result. If the two scores are equal, R rs_score (λ) is its mean, and then the optimal R rs_score(λ) data were used to invert chlorophyll a concentration.

[0079] The spatial distribution of chlorophyll-a concentration around the Shandong Peninsula inverted by GOCI under three conditions at 2018030904:16 is taken as an example. Figure 2a is the spatial distribution result of chlorophyll a concentration corresponding to GDPS2.0 atmospheric correction algorithm, Figure 2b is the spatial distribution result of chlorophyll a concentration corresponding to the MUMM atmospheric correction algorithm, Figure 2c The spatial distribution results of chlorophyll a concentration obtained after two atmospheric correction algorithms were optimized through QA Score.

[0080] Figure 3a-3c Comparison results of chlorophyll a concentration around Shandong Peninsula inverted by GOCI and measured station chlorophyll a concentration under three conditions. Figure 2a , Figure 2b and Figure 2c It can be found that the new chlorophyll a concentration inversion results mentioned in the present invention are spatially distributed ( Figure 2c ) will combine the spatial distribution of multiple atmospheric correction algorithms, compared with the results of using a single atmospheric correction algorithm ( Figure 2a ), which improves the effective data coverage of the inversion results in space. The results compared with the field experimental data ( Figure 3a , Figure 3b and Figure 3c ) shows the new chlorophyll a concentration inversion result mentioned in the present invention ( Figure 3c , correlation coefficient 0.9, root mean square error 0.54, average percentage error 19%) is better than the result of using a single atmospheric correction algorithm ( Figure 3a , correlation coefficient 0.79, root mean square error 0.53, average percentage error 19%; Figure 3b , correlation coefficient 0.71, root mean square error 0.69, average percentage error 26%). Therefore, the technical invention proposes that incorporating the pixel-by-pixel atmospheric correction optimization step into the process of inverting chlorophyll a concentration can not only improve the inversion accuracy of chlorophyll a concentration, but also improve the coverage of effective data.

[0081] The algorithm of the present invention proposes for the first time to incorporate the pixel-by-pixel atmospheric correction optimization step into the process of inverting the chlorophyll a concentration of the water surface using optical satellite data, so as to remove the chlorophyll a concentration inversion error caused by the application of a single atmospheric correction algorithm to process optical satellite data. And the example of inverting the chlorophyll a concentration of the water surface using GOCI satellite data in the waters around Shandong Peninsula shows that the innovative introduction of this step has significantly improved the accuracy of chlorophyll a concentration inversion.

[0082] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for inverting chlorophyll a concentration in the water surface layer by integrating the optimal results of atmospheric correction, characterized in that: The method incorporates the pixel-by-pixel atmospheric correction optimization result based on the QA Score model in the process of inverting the chlorophyll a concentration of the water surface using satellite data, so as to improve the inversion accuracy of the chlorophyll a concentration of the water surface.

2. The method for inverting chlorophyll a concentration in the surface layer of water body by integrating the optimal result of atmospheric correction according to claim 1, characterized in that: The method comprises the following steps: (1) m kinds of atmospheric correction algorithms are used to perform atmospheric correction processing on satellite data respectively, and the spectral data obtained by the m kinds of atmospheric correction algorithms are scored pixel by pixel through the QA Score model; the spectral data corresponding to the maximum score of each pixel point is selected as the optimal spectral data of the pixel point; wherein the number of types of atmospheric correction algorithms used m ≥ 2; (2) Using the optimal spectral data obtained in step (1) to invert the chlorophyll a concentration in the surface layer of the water body to improve the inversion accuracy of the chlorophyll a concentration.

3. The method for inverting chlorophyll a concentration in the water surface layer by integrating the atmospheric correction optimization result according to claim 1, characterized in that: In step (1), the atmospheric correction algorithm used includes two or more of the GDPS2.0 atmospheric correction algorithm, the MUMM atmospheric correction algorithm, the NASA near-infrared atmospheric correction algorithm, the Acolite DSF atmospheric correction algorithm, the FLASSH atmospheric correction algorithm, and the Polymer atmospheric correction algorithm.

4. The method for inverting chlorophyll a concentration in the water surface layer by integrating the atmospheric correction optimization result according to claim 2, characterized in that: Step (1) is specifically as follows: Apply the selected m kinds of atmospheric correction algorithms to perform atmospheric correction processing on the optical satellite data, and obtain m spectral data corresponding to each pixel point; The m spectral data are scored using the QA Score model to obtain corresponding m Score values, and the spectral data corresponding to the maximum Score value is taken as the optimal spectral data of the pixel point.

5. The method for inverting chlorophyll a concentration in the water surface layer by integrating the atmospheric correction optimization result according to claim 4, characterized in that: If there are multiple maximum values ​​Score, the average of the spectral data corresponding to the multiple maximum values ​​is taken as the optimal spectral data of the pixel point.

6. The method for inverting chlorophyll a concentration in the surface layer of water body by integrating the optimal result of atmospheric correction according to claim 4, characterized in that: The m spectral data are scored by applying the QA Score model to obtain the corresponding m Score values, which specifically includes the following steps: S1: The remote sensing reflectance R measured on site rs The (λ) spectrum was normalized to obtain the reference spectrum nR of the normalized remote sensing reflectance of 23 optical water bodies in nine bands. rs (λ); R rs (λ i ) indicates the wavelength is λ i The corresponding remote sensing reflectance when i is 1-9; the wavelengths corresponding to λ1 to λ9 are 412, 443, 488, 510, 531, 547, 555, 667, and 678 respectively; N represents the total number of bands, N = 9; nR rs The value range of (λ) is between 0 and 1; S2: By reference spectrum nR rs (λ) Get the upper boundary nR of each spectral band i of each water body rs U (λ i ) and the lower boundary nR rs L (λ i )value; S3: Each of the m spectral data corresponding to each pixel point is taken as the target spectrum R to be scored. rs *(λ'), use the QA Score model for scoring: S31: Determine the target spectrum R rs *(λ') band and reference spectrum nR rs Whether the band of (λ) matches; S32: If the target spectrum R rs *(λ') has a spectral band that is smaller than the reference spectrum nR rs (λ) has many spectral bands, that is, the target spectrum R rs *(λ') has a spectral band greater than 9; only select the target spectrum R rs *(λ') with reference spectrum nR rs (λ) The bands with the same wavelength proceed to step S33; If the target spectrum R rs *(λ') has a spectral band that is smaller than the reference spectrum nR rs (λ) has few spectral bands, that is, the total number of bands is less than 9, and the total target spectrum R rs *(λ') all proceed to step S33; From the reference spectrum nR rs (λ) Extract the nR corresponding to the wavelength λ' rs (λ') subset and the nR rs The upper boundary of the spectrum corresponding to the (λ') subset nR rs U (λ') subset and lower bound nR rs L (λ') subset; S33: Target spectrum R rs *(λ') is normalized to obtain the normalized target spectrum nR rs *(λ'): According to the SAM spectrum similarity formula, the normalized target spectrum nR rs *(λ') Assign water body type: α is the reference spectrum nR rs (λ) and the normalized target spectrum nR rs *(λ), SAM is calculated by taking the reference spectrum nR rs (λ) and the normalized target spectrum nR rs *(λ) is used as a spatial vector with a dimension equal to the total number of bands N of the target spectrum to determine the spectral similarity, and the spectral water type corresponding to the largest cosine value is determined as the normalized target spectrum nR rs *(λ) water body type; S4: By normalizing the target spectrum nR rs *(λ) is compared with the upper and lower boundary values ​​of the water type spectrum to estimate the QA score: C(λ i ) is the score for a specific wavelength, N is nR rs *(λ) total number of bands; If nR rs *(λ i ) value in nR rs (λ i )’s upper boundary nR rs U (λ i )) or lower boundary nR rs L (λ i ), the corresponding wavelength score will be assigned 0, that is, C(λ i )=0; if nR rs *(λ i ) value in nR rs (λ i )’s upper boundary nR rs U (λ i )) and the lower boundary nR rs L (λ i ), the corresponding wavelength score will be assigned a value of 1, that is, C(λ i )=1; S5: Repeat steps S3-S4 to obtain m spectral data R for each pixel point. rs_1 (λ),R rs_2 (λ),...,R rs_i (λ),...,R rs_m (λ) Apply the QA Score model to score and get m Score values: Score(R rs_1 (λ)),Score(R rs_1 (λ)),...,Score(R rs_i (λ)),...,Score(R rs_m (λ)).

7. The method for inverting chlorophyll a concentration in the surface layer of water body by integrating the optimal result of atmospheric correction according to claim 1, characterized in that: In step (2), the OC2 algorithm, the YOC algorithm, the OC3G algorithm, or the OC2M-HI algorithm is used to invert the chlorophyll a concentration in the surface layer of the water body.