A water color remote sensing method and device for estimating the concentration of particulate organic carbon in the ocean surface

Abstract

The present invention discloses a water color remote sensing method and device for estimating the concentration of particulate organic carbon in the ocean surface layer, and relates to the technical field of remote sensing inversion of water quality parameters. The method extracts the absorption coefficients at different wavelengths of all matching points in the optical information image; constructs the relationship between the measured organic carbon concentration in the matching point data set and the absorption coefficient at each wavelength, and performs multivariate nonlinear regression fitting to obtain the band corresponding to the absorption coefficient with the highest determination coefficient; takes the absorption coefficient at the band as the independent variable and the measured particulate organic carbon concentration as the dependent variable, establishes a logarithmic model, and obtains a remote sensing inversion model of particulate organic carbon concentration based on the absorption coefficient; uses the remote sensing inversion model of particulate organic carbon concentration based on the absorption coefficient and the satellite remote sensing image to be measured to obtain the concentration of particulate organic carbon in the ocean surface layer at different geographical locations. The method solves the problem that the inversion value of the current particulate organic carbon concentration inversion algorithm is low in high productivity areas, and has the advantages of wide applicable areas and strong feasibility.

Description

Technical Field

[0001] The invention belongs to the field of remote sensing inversion of water quality parameters, and relates to a water color remote sensing method and device for estimating the concentration of particulate organic carbon in the ocean surface. Background Art

[0002] The ocean plays an important role in the global carbon cycle by regulating the concentration of carbon dioxide in the atmosphere. Carbon in the ocean can be divided into dissolved inorganic carbon, dissolved organic carbon, particulate inorganic carbon, and particulate organic carbon. Although particulate organic carbon only accounts for a small part of the ocean surface carbon pool, its importance lies in its high turnover rate. Knowledge of the distribution of particulate organic carbon concentrations in the ocean is crucial to understanding the growth rate of marine phytoplankton and the methods of carbon-based net primary productivity.

[0003] Traditional ship-based measurements of particulate organic carbon are time-consuming and laborious, and are constrained by time and space. Ocean color remote sensing technology has the advantage of long-term and large-area observations. With the continuous development of ocean color satellites, the temporal and spatial resolution of water color sensors has been significantly improved, providing better support for the observation of global ocean surface particulate organic carbon concentrations.

[0004] At present, a variety of water color remote sensing inversion methods for sea surface particulate organic carbon concentration have been established. Most of these methods are based on limited, regional measured particulate organic carbon data sets. Due to the diversity of global ocean water bodies, it is difficult to apply regional methods to global waters, i.e., they have poor applicability.

[0005] At the same time, the band ratio (BG) algorithm is currently widely used to obtain the global ocean surface particulate organic carbon concentration. This algorithm is also used by NASA to provide global particulate organic carbon concentration in the L2 product suite of sensors such as the Moderate Resolution Imaging Spectroradiometer (MODIS). Although the band ratio method gives relatively reliable results in the inversion of global open ocean surface particulate organic carbon concentration, in areas with high concentrations of particulate organic carbon, the inversion values ​​obtained by this method are significantly lower than the measured values, that is, its inversion accuracy is poor in areas of high productivity. Summary of the invention

[0006] In view of this, the present invention provides a water color remote sensing method for estimating the concentration of particulate organic carbon in the ocean surface, so as to solve the technical problems that the existing method of inverting particulate organic carbon by water color remote sensing technology has poor inversion accuracy and poor applicability in high productivity areas.

[0007] To this end, the technical solution adopted in the present invention is as follows:

[0008] In one aspect, the present invention provides a method for estimating the concentration of particulate organic carbon in the surface layer of the ocean by remote sensing, comprising the following steps:

[0009] Obtain a global dataset of measured organic carbon concentrations in the ocean surface;

[0010] Obtain optical information images representing the global ocean surface water body;

[0011] Matching the data of the measured organic carbon concentration data set with the optical information image in time and space to obtain a matching point data set, and dividing the matching point data set into a matching point training set and a matching point test set; the matching point training set is used to construct a remote sensing inversion model of particulate organic carbon concentration based on absorption coefficient; the matching point test set is used to verify the accuracy of the remote sensing inversion model of particulate organic carbon concentration based on absorption coefficient;

[0012] Extracting absorption coefficients at different wavelengths of all matching points in the optical information image;

[0013] Constructing the relationship between the measured organic carbon concentration in the matching point data set and the absorption coefficient at each wavelength, and performing multivariate nonlinear regression fitting to obtain the band corresponding to the absorption coefficient with the highest determination coefficient; taking the band corresponding to the absorption coefficient with the highest determination coefficient as the model band;

[0014] Taking the absorption coefficient at the model band as the independent variable and the measured particulate organic carbon concentration as the dependent variable, a logarithmic model is established, wherein the logarithmic model is a remote sensing inversion model of particulate organic carbon concentration based on the absorption coefficient;

[0015] The concentration of particulate organic carbon in the ocean surface at different geographical locations is obtained by using the absorption coefficient-based particulate organic carbon concentration remote sensing inversion model and the satellite remote sensing image to be measured.

[0016] Furthermore, the absorption coefficient with the highest determination coefficient corresponds to a wavelength band of 490 nm (a(490)).

[0017] Furthermore, the remote sensing inversion model of particulate organic carbon concentration based on the absorption coefficient is:

[0018]

[0019] Where X = log 10 a(490).

[0020] Furthermore, the data of the measured data set of particulate organic carbon concentration are matched with the optical information image in time and space, including: the sampling time of the measured data of particulate organic carbon concentration and the satellite transit time are on the same day, and the pixel corresponding to the sampling point of the measured data is used as the center pixel to extract 3×3 pixels around it and take the average value.

[0021] Furthermore, the optical information images characterizing the global ocean surface water body include: OC-CCI remote sensing images.

[0022] Furthermore, the satellite remote sensing image to be measured includes: OC-CCI remote sensing image or OLCI L1B remote sensing image.

[0023] Furthermore, when the satellite remote sensing image to be measured is an OLCI L1B remote sensing image, the particulate organic carbon concentration remote sensing inversion model based on the absorption coefficient and the satellite remote sensing image to be measured are used to obtain the concentration of particulate organic carbon in the ocean surface at different geographical locations, including:

[0024] Acquire OLCI L1B remote sensing images;

[0025] Performing geographic correction and atmospheric correction on the OLCI L1B remote sensing image;

[0026] The QAA_V5 semi-analytical algorithm is used to extract the absorption coefficient of all matching points in the OLCI L1B remote sensing image under the model band;

[0027] Based on the absorption coefficient and the remote sensing inversion model of particulate organic carbon concentration based on the absorption coefficient, the concentration of particulate organic carbon in the ocean surface at different geographical locations in the OLCI L1B remote sensing image is obtained.

[0028] Furthermore, when the satellite remote sensing image to be measured is an OC-CCI remote sensing image, the particulate organic carbon concentration remote sensing inversion model based on the absorption coefficient and the satellite remote sensing image to be measured are used to obtain the concentration of particulate organic carbon in the ocean surface at different geographical locations, including:

[0029] Obtain OC-CCI remote sensing images;

[0030] Extracting the absorption coefficient of all matching points in the OC-CCI remote sensing image under the model band;

[0031] Based on the absorption coefficient and the remote sensing inversion model of particulate organic carbon concentration based on the absorption coefficient, the concentration of particulate organic carbon in the ocean surface at different geographical locations in the OC-CCI remote sensing image is obtained.

[0032] On the other hand, the present invention also provides a water color remote sensing device for estimating the concentration of particulate organic carbon in the ocean surface, comprising:

[0033] The first acquisition module: obtain the measured data set of global ocean surface organic carbon concentration;

[0034] The second acquisition module: acquires optical information images representing the global ocean surface water body;

[0035] Matching module: performing temporal and spatial matching of the data of the measured organic carbon concentration data set with the optical information image to obtain a matching point data set, and dividing the matching point data set into a matching point training set and a matching point test set; the matching point training set is used to construct a remote sensing inversion model of particulate organic carbon concentration based on absorption coefficient; the matching point test set is used to verify the accuracy of the remote sensing inversion model of particulate organic carbon concentration based on absorption coefficient;

[0036] Extraction module: extracting the absorption coefficients at different wavelengths of all matching points in the optical information image;

[0037] Construction module: construct the relationship between the measured organic carbon concentration in the matching point data set and the absorption coefficient at each wavelength, and perform multivariate nonlinear regression fitting to obtain the band corresponding to the absorption coefficient with the highest determination coefficient; the band corresponding to the absorption coefficient with the highest determination coefficient is used as the model band;

[0038] A logarithmic model is established with the measured organic carbon concentration as the independent variable and the absorption coefficient at the model band as the dependent variable. The logarithmic model is a remote sensing inversion model of particulate organic carbon concentration based on the absorption coefficient.

[0039] Estimation module: using the remote sensing inversion model of particulate organic carbon concentration based on absorption coefficient and the satellite remote sensing image to be measured, the concentration of particulate organic carbon in the surface layer of the ocean at different geographical locations is obtained.

[0040] The present invention provides a water color remote sensing method for estimating the concentration of particulate organic carbon in the ocean surface layer. Combined with the global particulate organic carbon shared data set, a new method for particulate organic carbon in the ocean surface layer is established to achieve accurate characterization of the global particulate organic carbon concentration in the ocean surface layer, thereby enhancing the adaptability of the scheme. Compared with the band ratio (BG) algorithm, in areas with low concentrations of particulate organic carbon, the present invention obtains inversion results similar to the band ratio (BG) algorithm; however, in areas with high concentrations, the present invention obtains results that are more consistent with the measured values. This application constructs a remote sensing inversion model for particulate organic carbon concentration based on the inherent optical quantity of water bodies, and uses satellite data to obtain a global ocean surface particulate organic carbon concentration product. This can solve the problem of low inversion values ​​of the current particulate organic carbon concentration inversion algorithm in high productivity areas, and provide more accurate results for further estimating global ocean carbon reserves. It has the advantages of wide applicability and strong feasibility. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0042] Figure 1 It is the technical roadmap of the present invention;

[0043] Figure 2 is a construction fitting diagram of the method of the present invention;

[0044] Figure 3 (a) is a scatter plot between the POC concentration inversion value obtained by the BG algorithm and the measured value; (b) is a scatter plot between the POC concentration inversion value obtained by the algorithm proposed in the present invention and the measured value;

[0045] Figure 4 It is a global monthly average POC product map generated by the present invention using the OC-CCI monthly average image;

[0046] Figure 5 (a) is a map of the study area; (b) is a POC concentration map I inverted from an OLCI image in Bohai Sea using the algorithm proposed in the present invention; (c) is a POC concentration map II inverted from an OLCI image in Bohai Sea using the algorithm proposed in the present invention; (d) is a POC concentration map III inverted from an OLCI image in Bohai Sea using the algorithm proposed in the present invention. DETAILED DESCRIPTION

[0047] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0048] like Figure 1 As shown, a water color remote sensing method for estimating the concentration of particulate organic carbon in the ocean surface provided in an embodiment of the present invention comprises the following steps:

[0049] S1: Obtain measured data of particulate organic carbon (POC) concentration;

[0050] Among them, the measured data of POC concentration is the measured data set of POC concentration in the global ocean surface. In the specific implementation, the shared measured data set of POC concentration in the global ocean surface is obtained from SeaBASS (https: / / seabass.gsfc.nasa.gov / ), BCO-DMO (https: / / www.bco-dmo.org / ) and PANGAEA (https: / / www.pangaea.de / ) websites; in addition, field sampling and laboratory measurements are also carried out to obtain the measured data of POC concentration.

[0051] S2: Acquire optical information images representing the global ocean surface water body;

[0052] In specific implementation, you can download the OC-CCI (Ocean Color Climate Change Initiative) v5.0 global hybrid satellite remote sensing image all_products product from Ocean Colour (https: / / climate.esa.int / en / projects / ocean-colour / ).

[0053] S3: Match the data in the POC measured data set with the optical information image in time and space to obtain a number of matching points to form a matching point data set, and divide the matching point data set into a training set and a test set;

[0054] The matching requirements are: the measured data sampling time and the satellite transit time are on the same day, 3×3 pixels are extracted with the pixel corresponding to the measured data sampling point as the center pixel, and the average value is taken.

[0055] In a specific implementation, 2 / 3 of the points in the matching point data set can be used as a training set, and the rest can be used as matching points as a test set.

[0056] S4: Extract the absorption coefficient a(λ) at different wavelengths of all matching points in the optical information image of the global ocean surface water body.

[0057] S5: construct the relationship between the measured POC concentration in the training set and the absorption coefficient a(λ) at each wavelength, and perform multivariate nonlinear regression fitting to obtain the band corresponding to the absorption coefficient with the highest determination coefficient, and use the band corresponding to the absorption coefficient with the highest determination coefficient as the model band; use the absorption coefficient at the model band as the independent variable and the measured POC concentration as the dependent variable to establish a logarithmic model, which is a POC concentration remote sensing inversion model based on the absorption coefficient;

[0058] In the embodiment of the present invention, the determination coefficient R 2 The highest absorption coefficient a(λ) corresponds to the wavelength band λ of 490nm.

[0059] The absorption coefficient a(490) at 490nm of all matching points was extracted from the optical information image of the global ocean surface water body using SNAP (SeNtinel Applications Platform) software;

[0060] like Figure 2 As shown, the expression of POC concentration in the ocean surface at different geographical locations is as follows:

[0061]

[0062] Where X = log 10 a(490).

[0063] Formula (1) is named a-POC algorithm.

[0064] like Figure 3 As shown, the a-POC algorithm and NASA's official POC algorithm BG algorithm are applied to the accuracy verification matching point data. Figure 3 (a) and (b) are the accuracy relationships between the BG algorithm and the a-POC algorithm and the measured POC concentration, respectively. The black solid line represents the 1:1 line, and the dotted line represents the best fit line of the Model-II linear regression. The linear fitting relationship and the determination coefficient R are also given. 2 .

[0065] The a-POC algorithm has a slope closer to 1 and an intercept closer to 0 than the BG algorithm. The root mean square errors of the two algorithms are 128.49 mg / m 3 and 184.48 mg / m 3 , the deviations were -3.29mg / m 3 and -49.06mg / m 3 .

[0066] S6: Using the POC concentration remote sensing inversion model based on the absorption coefficient and the satellite remote sensing image to be measured, the POC concentration of the ocean surface at different geographical locations is obtained.

[0067] In specific applications, the satellite remote sensing image to be measured can be an OC-CCI image. Specifically, the OC-CCI image a(490) data that is synchronized with the measured data in time and space is extracted, and the a-POC algorithm is applied to obtain the POC concentration of the ocean surface at different geographical locations.

[0068] The a-POC algorithm was applied to the monthly average image of OC-CCI in May 2020, and the POC concentration of each pixel was estimated using Python tools, and a global POC concentration distribution map was generated, as shown in the figure. Figure 4 shown.

[0069] The satellite remote sensing image to be measured may also be an OLCI L1B remote sensing image, which specifically includes the following steps:

[0070] S61: Data collection, collecting water samples on site near the Bohai Sea and measuring the concentration of POC in the laboratory;

[0071] S62: Download satellite remote sensing images, download the OLCI (Ocean and Land Colour Instrument) study area L1B full-resolution (300m) remote sensing images corresponding to the data sampling time;

[0072] S63: Satellite remote sensing image preprocessing, L1B image is first subjected to atmospheric correction. Previous studies have shown that the POLYMER atmospheric correction algorithm is more suitable for the Bohai Sea coast. Therefore, POLYMER v4.13 is used to perform atmospheric correction on the L1B image to generate the L2 remote sensing reflectance Rrs(λ);

[0073] S64: Calculate a(490) using Rrs(λ) using the QAA_V5 semi-analytical algorithm. The calculation steps are as follows:

[0074]

[0075]

[0076]

[0077]

[0078]

[0079]

[0080]

[0081]

[0082] Among them, g0=0.089, g1=0.125.

[0083] S65: Extract OLCI L1B remote sensing image a(490) data that is synchronized with the measured data in time and space, and apply a-POC algorithm to obtain the POC concentration of the ocean surface at different geographical locations.

[0084] Compared with NASA's official POC algorithm BG algorithm, the median absolute percentage error of the a-POC algorithm was reduced from 40.77% to 19.13%, and the root mean square error was reduced from 66.26 mg / m 3 Reduced to 35.97 mg / m 3, the deviation is -297.80mg / m 3 Reduced to -89.04 mg / m 3 This shows that the a-POC algorithm significantly improves the low inversion accuracy of the BG algorithm in nearshore waters.

[0085] The a-POC algorithm was applied to OLCI images of the Bohai Sea on June 14, September 19 and 23, 2017, respectively. The POC concentration of each pixel was estimated using Python, and the POC concentration distribution map of the Qinhuangdao nearshore area was generated. (b) is the POC concentration map I inverted by the algorithm proposed in the present invention in the Bohai Sea; (c) is the POC concentration map II inverted by the algorithm proposed in the present invention in the Bohai Sea; (d) is the POC concentration map III inverted by the algorithm proposed in the present invention in the Bohai Sea. Figure 5 The scattered points in the middle are the locations of the measured sampling points, and the background color and the scattered point color correspond to the color bar below to show the POC concentration.

[0086] Corresponding to the water color remote sensing method for estimating the POC concentration of the ocean surface in the above embodiment, an embodiment of the present invention further provides a water color remote sensing device for estimating the POC concentration of the ocean surface, comprising:

[0087] The first acquisition module: obtain the measured data set of global ocean surface POC concentration;

[0088] The second acquisition module: acquires optical information images representing the global ocean surface water body;

[0089] Matching module: performing temporal and spatial matching of the data of the POC concentration measured data set with the optical information image to obtain a matching point data set, and dividing the matching point data set into a matching point training set and a matching point test set; the matching point training set is used to construct a POC concentration remote sensing inversion model based on the absorption coefficient; the matching point test set is used to verify the accuracy of the POC concentration remote sensing inversion model based on the absorption coefficient;

[0090] Extraction module: extracting the absorption coefficients at different wavelengths of all matching points in the optical information image;

[0091] Construction module: construct the relationship between the measured POC concentration in the matching point data set and the absorption coefficient at each wavelength, and perform multivariate nonlinear regression fitting to obtain the band corresponding to the absorption coefficient with the highest determination coefficient; use the band corresponding to the absorption coefficient with the highest determination coefficient as the model band; use the measured POC concentration as the independent variable and the absorption coefficient at the model band as the dependent variable to establish a logarithmic model, which is a POC concentration remote sensing inversion model based on the absorption coefficient;

[0092] Estimation module: using the POC concentration remote sensing inversion model based on the absorption coefficient and the satellite remote sensing image to be measured, the POC concentration of the ocean surface at different geographical locations is obtained.

[0093] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A water color remote sensing method for estimating the concentration of particulate organic carbon in the ocean surface, characterized by: The following steps are involved: Obtain a global dataset of measured concentrations of particulate organic carbon in the ocean surface; Obtain optical information images representing the global ocean surface water body; Matching the data of the measured organic carbon concentration data set with the optical information image in time and space to obtain a matching point data set, and dividing the matching point data set into a matching point training set and a matching point test set; the matching point training set is used to construct a remote sensing inversion model of particulate organic carbon concentration based on absorption coefficient; the matching point test set is used to verify the accuracy of the remote sensing inversion model of particulate organic carbon concentration based on absorption coefficient; Extracting absorption coefficients at different wavelengths of all matching points in the optical information image; Constructing the relationship between the measured organic carbon concentration in the matching point data set and the absorption coefficient at each wavelength, and performing multivariate nonlinear regression fitting to obtain the band corresponding to the absorption coefficient with the highest determination coefficient; taking the band corresponding to the absorption coefficient with the highest determination coefficient as the model band; Taking the absorption coefficient at the model band as the independent variable and the measured particulate organic carbon concentration as the dependent variable, a logarithmic model is established, and the logarithmic model is a remote sensing inversion model of particulate organic carbon concentration based on the absorption coefficient; The concentration of particulate organic carbon in the ocean surface at different geographical locations is obtained by using the absorption coefficient-based particulate organic carbon concentration remote sensing inversion model and the satellite remote sensing image to be measured.

2. A method for estimating the concentration of particulate organic carbon in the ocean surface layer according to claim 1, characterized in that: The absorption coefficient a(λ) with the highest determination coefficient corresponds to a band λ of 490 nm, and the absorption coefficient of the 490 nm band is expressed as a(490).

3. A method for estimating the concentration of particulate organic carbon in the surface layer of the ocean according to claim 2, characterized in that: The remote sensing inversion model of particulate organic carbon concentration based on the absorption coefficient is: Wherein, X=log10a(490).

4. A method for estimating the concentration of particulate organic carbon in the ocean surface layer according to claim 1, characterized in that: The data of the measured data set of particulate organic carbon concentration are matched with the optical information image in time and space, including: the sampling time of the measured data of particulate organic carbon concentration and the satellite transit time are on the same day, and the pixel corresponding to the measured data sampling point is taken as the center pixel to extract the 3×3 pixels around it and take the average value.

5. The method for estimating the concentration of particulate organic carbon in the ocean surface layer according to claim 1, characterized in that: The optical information images representing the global ocean surface water body include: OC-CCI remote sensing images.

6. The method for estimating the concentration of particulate organic carbon in the ocean surface layer according to claim 1, characterized in that: The satellite remote sensing image to be measured includes: OC-CCI remote sensing image or OLCI L1B remote sensing image.

7. A method for estimating the concentration of particulate organic carbon in the ocean surface layer according to claim 6, characterized in that: When the satellite remote sensing image to be measured is an OLCIL1B remote sensing image, the particulate organic carbon concentration remote sensing inversion model based on the absorption coefficient and the satellite remote sensing image to be measured are used to obtain the concentration of particulate organic carbon in the ocean surface at different geographical locations, including: Acquire OLCI L1B remote sensing images; Performing geographic correction and atmospheric correction on the OLCIL1B remote sensing image; The QAA_V5 semi-analytical algorithm is used to extract the absorption coefficient of all matching points in the OLCI L1B remote sensing image under the model band; Based on the absorption coefficient and the remote sensing inversion model of particulate organic carbon concentration based on the absorption coefficient, the concentration of particulate organic carbon in the ocean surface at different geographical locations in the OLCIL1B remote sensing image is obtained.

8. A method for estimating the concentration of particulate organic carbon in the ocean surface layer according to claim 6 or 7, characterized in that: When the satellite remote sensing image to be measured is an OC-CCI remote sensing image, the particulate organic carbon concentration remote sensing inversion model based on the absorption coefficient and the satellite remote sensing image to be measured are used to obtain the concentration of particulate organic carbon in the ocean surface at different geographical locations, including: Obtain OC-CCI remote sensing images; Extracting the absorption coefficient of all matching points in the OC-CCI remote sensing image under the model band; Based on the absorption coefficient and the remote sensing inversion model of particulate organic carbon concentration based on the absorption coefficient, the concentration of particulate organic carbon in the ocean surface at different geographical locations in the OC-CCI remote sensing image is obtained.

9. A water color remote sensing device for estimating the concentration of particulate organic carbon in the ocean surface, characterized in that: include: The first acquisition module: obtain the measured data set of global ocean surface organic carbon concentration; The second acquisition module: acquires optical information images representing the global ocean surface water body; Matching module: performing temporal and spatial matching of the data of the measured organic carbon concentration data set with the optical information image to obtain a matching point data set, and dividing the matching point data set into a matching point training set and a matching point test set; the matching point training set is used to construct a remote sensing inversion model of particulate organic carbon concentration based on absorption coefficient; the matching point test set is used to verify the accuracy of the remote sensing inversion model of particulate organic carbon concentration based on absorption coefficient; Extraction module: extracting the absorption coefficients at different wavelengths of all matching points in the optical information image; Construction module: construct the relationship between the measured organic carbon concentration in the matching point data set and the absorption coefficient at each wavelength, and perform multivariate nonlinear regression fitting to obtain the band corresponding to the absorption coefficient with the highest determination coefficient; use the band corresponding to the absorption coefficient with the highest determination coefficient as the model band; use the measured organic carbon concentration as the independent variable and the absorption coefficient at the model band as the dependent variable to establish a logarithmic model, which is a remote sensing inversion model of particulate organic carbon concentration based on the absorption coefficient; Estimation module: using the remote sensing inversion model of particulate organic carbon concentration based on absorption coefficient and the satellite remote sensing image to be measured, the concentration of particulate organic carbon in the surface layer of the ocean at different geographical locations is obtained.

Images