A method and system for calculating the profile structure of particulate organic carbon in upper ocean waters

Abstract

The present invention provides a method and system for calculating the profile structure of particulate organic carbon in upper seawater. By using high-frequency observational data of particulate attenuation coefficient, rich on-site data is provided for the construction of a neural network model. On this basis, considering the model accuracy and the availability of water color satellite remote sensing and numerical models of input parameters, a prediction model of water body particulate attenuation coefficient is constructed by using sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and particulate organic carbon profile type as input parameters; and finally, according to the conversion relationship between water body particulate organic carbon concentration and particulate attenuation coefficient, the three-dimensional distribution of water body particulate organic carbon concentration in the sea basin area is obtained. Thus, by virtue of the large-scale quasi-real-time long-time-series observation advantages of remote sensing, the dynamic monitoring and historical change assessment of the three-dimensional field of water body particulate organic carbon concentration in the sea basin area are realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of remote sensing, and particularly relates to a method and system for calculating the profile structure of particulate organic carbon in upper seawater. Background Art

[0002] The ocean carbon pump refers to the process by which the ocean regulates the absorption of carbon dioxide in the atmosphere and stores it inside the ocean or buries it in seafloor sediments. Among them, the biological pump process driven by the photosynthesis of phytoplankton in the euphotic zone converts inorganic carbon in seawater into organic carbon and exports it to the deep sea in the form of particulate organic carbon (POC), thereby achieving carbon sequestration in the ocean. Although phytoplankton only accounts for 0.2%-0.3% of the ocean biomass, due to its high turnover rate, this process can explain approximately 70% of the vertical gradient change of dissolved inorganic carbon. The effective operation of the biological pump plays an important regulatory role in the concentration of atmospheric carbon dioxide ( ), and if this process stops, the atmospheric concentration will increase by nearly 200 ppm, which is equivalent to an increase of about 40% compared to the current level. Therefore, improving the understanding of the distribution, composition of POC in the euphotic zone and its changes under different climate conditions is one of the key areas in climate change research.

[0003] Currently, ocean color satellite remote sensing is the only feasible means to meet the continuous observation requirements of POC in the upper water layer at the global or regional scale. However, the observation ability of traditional ocean color satellites is theoretically limited to the first optical depth, that is, the upper 20% area of the euphotic zone, which poses limitations to the observation of the POC profile throughout the euphotic zone. Existing studies mainly use two types of methods to evaluate the POC distribution in the upper water layer: one is to establish a statistical relationship between the sea surface POC concentration and the integrated or average concentration of the water layer (the first optical depth, the euphotic zone depth, or the mixed layer depth); the other is to construct a typical profile model by combining satellite observations at the sea surface. Early studies found that for seawater dominated by phytoplankton, the integrated storage of chlorophyll Chla in the euphotic zone can be estimated through the power function relationship of sea surface Chla, and further constructed 7 Gaussian profile distribution models of Chla in the euphotic zone, thereby realizing the three-dimensional remote sensing observation of Chla in the upper water layer at the global scale. On this basis, the structural characteristics of the POC profile in the euphotic zone were studied, and empirical power function relationships between the integrated storage of POC in the euphotic zone and the sea surface POC concentration were constructed under stratified and well-mixed conditions respectively to achieve the remote sensing estimation of POC at the global scale.

[0004] However, when applying empirical methods at the global scale in marginal sea areas, it is necessary to re-tune the model constants, and even the model structure (linear or power function) itself may not be applicable to regional variation characteristics. In addition, Patent (CN202110272580.X) proposes a method for classifying the POC profile distribution that depends on the sea surface POC concentration, the mixed layer depth, and the water depth. However, this method is based on fixed seasonal inputs and is applicable to monthly average scale observations. There may be jump phenomena at the intra-month scale, and using fixed constants within a season makes it difficult to meet the dynamic observation requirements of remote sensing images pixel by pixel.

[0005] In recent years, research on POC has gradually focused on the inherent optical properties of water with high vertical resolution to better characterize the profile characteristics of traditional discrete sampling POC. However, there is currently a lack of a dynamic observation method for the profile structure of the inherent optical properties of water based on satellite remote sensing, and related research still mainly relies on in-situ observation data or directly uses optical profile data measured by in-situ instruments as model inputs. Therefore, there is an urgent need to develop remote sensing dynamic observation technology based on the inherent optical properties of water to improve the three-dimensional remote sensing observation ability of POC in the upper water layer. Summary of the Invention

[0006] The present invention provides a method and system for calculating the profile structure of particulate organic carbon in the upper seawater to solve the problem that the existing technology lacks a dynamic observation method for the profile structure of the inherent optical properties of water based on remote sensing and mainly relies on in-situ observation data or directly uses optical profile data measured by in-situ instruments.

[0007] To solve the above technical problems, the embodiments of the present invention disclose the following technical solutions:

[0008] One aspect of the present invention provides a method for calculating the profile structure of particulate organic carbon in the upper seawater, which is applied to the ocean basin area. The method includes:

[0009] Obtain seawater characteristic data of multiple observation points in the ocean basin area. The seawater characteristic data of each observation point includes at least sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, particulate organic carbon profile type, and particulate attenuation coefficient profile data;

[0010] Train a neural network based on the seawater characteristic data of all observation points to obtain a particulate attenuation coefficient profile prediction model;

[0011] Identify the ocean basin area pixels in the target remote sensing image, and determine the particulate organic carbon profile type of each ocean basin area pixel in a preset manner;

[0012] For each pixel in the ocean basin area, the obtained sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and the corresponding particulate organic carbon profile type are input into the prediction model to obtain particulate attenuation coefficient profile data;

[0013] Pre-obtain the conversion relationship between particulate organic carbon concentration and particulate attenuation coefficient in the ocean basin area;

[0014] Based on the conversion relationship, obtain the particulate organic carbon profile data of each pixel in the ocean basin area according to the corresponding particulate attenuation coefficient profile data, and construct the three-dimensional field of particulate organic carbon in the upper layer of seawater in the target ocean basin area.

[0015] Optionally, the training of the neural network based on the seawater characteristic data of all observation points to obtain the particulate attenuation coefficient profile prediction model includes:

[0016] Construct a training data set based on the seawater characteristic data of all observation points. Each sample includes the sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and particulate attenuation coefficient profile data observed at any observation point during the same period, as well as the particulate organic carbon profile type at the observation point;

[0017] Use the sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and particulate organic carbon profile type in the sample as input data, and the particulate attenuation coefficient profile data as output data to train the neural network model and use the 5-fold cross-validation method for verification, and finally obtain the particulate attenuation coefficient profile prediction model.

[0018] Optionally, before performing the step of training the neural network based on the seawater characteristic data of all observation points to obtain the particulate attenuation coefficient profile prediction model, the method further includes:

[0019] Preprocess the seawater characteristic data of all observation points, and the preprocessing at least includes outlier removal and normalization processing.

[0020] Optionally, identifying the pixels in the ocean basin area in the target remote sensing image and determining the particulate organic carbon profile type of each pixel in the ocean basin area in a preset manner includes:

[0021] Obtain the water depth at each pixel in the target remote sensing image;

[0022] Determine the pixels with water depth exceeding the preset depth threshold as the pixels in the ocean basin area;

[0023] Calculate the euphotic layer depth corresponding to each pixel in the ocean basin area, and judge whether the euphotic layer depth is greater than the preset classification threshold;

[0024] If so, determine that the particulate organic carbon profile type corresponding to the pixel is Gaussian-like;

[0025] If not, determine that the particulate organic carbon profile type corresponding to the pixel is an exponential decay type.

[0026] Optionally, the euphotic layer depth corresponding to each pixel in the ocean basin area is calculated using the following formula:

[0027] kvis = k1 + k2 / (1 + Z)^0.5

[0028] TE = kvis × Z

[0029] Zeu = min(abs(TE - 4.605))

[0030] Where Z is the water depth at the pixel; kvis is the diffuse attenuation coefficient of the water body; TE is the optical depth; k1 and k2 are obtained by calculating the following formula from the inherent optical quantities of the water body:

[0031] k1=(x0 + x1×at490^0.5 + x2×bbt490)×(1 + a0×sin( ));

[0032] k2=(c0 + c1×at490 + c2×bbt490)×(a1 + a2×cos( ));

[0033] Where x0 = -0.057; x1 = 0.482; x2 = 4.221; c0 = 0.183; c1 = 0.702; c2 = -2.567; a0 = 0.090; a1 = 1.465; a2 = -0.667; at490 and bbt490 are the total absorption coefficient and the backscattering coefficient at a wavelength of 490 nm below the sea surface, respectively, obtained from the remote sensing image metadata; is the solar zenith angle.

[0034] Optionally, the conversion relationship between the particulate organic carbon concentration and the particulate attenuation coefficient in the ocean basin area is determined according to the following formula:

[0035] POC = 10 ^ ((log(cp660) + 1.71) / 1.27)

[0036] Where cp660 is the particulate attenuation coefficient; POC is the particulate organic carbon concentration in the upper layer of seawater.

[0037] Optionally, the method further includes:

[0038] Preprocess the target remote sensing image;

[0039] Obtain the sea surface chlorophyll concentration, the seawater temperature profile data, and the seawater salinity profile data corresponding to each pixel, respectively.

[0040] Optionally, the separately obtaining the sea surface chlorophyll concentration, seawater temperature profile data, and seawater salinity profile data corresponding to each pixel includes:

[0041] Obtaining the sea surface chlorophyll concentration using publicly available remote sensing products;

[0042] Obtaining the seawater temperature profile data and seawater salinity profile data based on the HYCOM global ocean model.

[0043] Optionally, construct the three-dimensional field of particulate organic carbon in the upper layer of seawater in the target ocean basin area in the following manner:

[0044] For the particulate organic carbon profile data of each pixel in the ocean basin area, perform a moving average process to obtain the vertical distribution data of the particulate organic carbon concentration of the pixel in the ocean basin area;

[0045] Smooth the planar data composed of the vertical distribution data of the particulate organic carbon concentration of all pixels in the ocean basin area using median filtering;

[0046] Construct a three-dimensional field based on the processed particulate organic carbon concentration data.

[0047] Another aspect of the present invention discloses a system for calculating the particulate organic carbon profile structure in the upper layer of seawater, and the system executes the method for calculating the particulate organic carbon profile structure in the upper layer of seawater described in any one of the foregoing aspects.

[0048] A method and system for calculating the particulate organic carbon profile structure in the upper layer of seawater disclosed by the present invention provide rich in-situ data for constructing a neural network model through high-frequency particulate attenuation coefficient observation data. On this basis, considering the model accuracy and the availability of water color satellite remote sensing and numerical models of input parameters, using the sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and particulate organic carbon profile type as input parameters, construct a prediction model for the particulate attenuation coefficient of water bodies; finally, based on the conversion relationship between the particulate organic carbon concentration and the particulate attenuation coefficient of water bodies, obtain the three-dimensional distribution of the particulate organic carbon concentration in the water bodies in the ocean basin area. Thus, by leveraging the advantages of large-scale, quasi-real-time, and long-time series observations of remote sensing, realize the dynamic monitoring and historical change assessment of the three-dimensional field of the particulate organic carbon concentration in the water bodies in the ocean basin area.

[0049] The invention content part is provided to introduce the selection of concepts in a simplified form, which will be further described in the specific implementation manners below. The invention content part is not intended to identify the important features or essential features of the present disclosure, nor is it intended to limit the scope of the present disclosure. Brief Description of the Drawings

[0050] The above and other objects, features, and advantages of the present disclosure will become more apparent by describing the exemplary embodiments of the present disclosure in more detail with reference to the accompanying drawings, in which, in the exemplary embodiments of the present disclosure, the same reference numerals generally represent the same components.

[0051] Figure 1 It is a schematic flowchart of a method for calculating the profile structure of particulate organic carbon in upper seawater provided by an embodiment of the present invention;

[0052] Figure 2 It is for realizing Figure 1 the flowchart of step S200 in

[0053] Figure 3 It is for realizing Figure 1 the flowchart of step S300 in

[0054] Figure 4 It is a schematic flowchart of another method for calculating the profile structure of particulate organic carbon in upper seawater provided by an embodiment of the present invention.

[0055] Figure 5 It is for realizing Figure 1 the flowchart of step S600 in Detailed implementation manners

[0056] Hereinafter, embodiments of the present disclosure will be described in more detail with reference to the accompanying drawings. Although the embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure will be more thorough and complete, and can fully convey the scope of the present disclosure to those skilled in the art.

[0057] [[ID=^{34]]The term "including" and its variations used herein mean open inclusion, that is, "including but not limited to". Unless otherwise stated, the term "or" means "and / or". The term "based on" means "at least partially based on". The terms "an exemplary embodiment" and "an embodiment" mean "at least one exemplary embodiment". The term "another embodiment" means "at least one additional embodiment". The terms "first", "second", etc. may refer to different or the same objects. There may also be other explicit and implicit definitions hereinafter.

[0058] Figure 1 It is a schematic flowchart of a method for calculating the profile structure of particulate organic carbon in upper seawater provided by an embodiment of the present invention. This method is applied to the ocean basin area, and the ocean basin is a low-lying area in the ocean topography. As Figure 1 shown, this method includes the following steps:

[0059] Step S100: Obtain the seawater characteristic data of multiple observation points in the sea basin area.

[0060] In the disclosed embodiments of the present invention, the seawater characteristic data of each observation point at least includes sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, particulate organic carbon profile type, and particulate attenuation coefficient profile data.

[0061] The sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and particulate attenuation coefficient profile data of the observation points can be obtained through remote sensing data or on-site measurement.

[0062] For example, the sea surface chlorophyll concentration can be obtained by collecting samples with a water sampler and then measuring them by a laboratory using fluorescence method, or can be obtained from publicly available remote sensing products;

[0063] The seawater temperature profile data and seawater salinity profile data can be obtained by measuring with an on-site CTD probe, or can be obtained through the HYCOM model. HYCOM is a global ocean numerical model that provides temperature profile data with high spatio-temporal resolution.

[0064] The particulate attenuation coefficient profile data can be obtained by using an AC-S water body measurement device through profile detection. AC-S can measure the light absorption and scattering of the water body.

[0065] Step S200: Train a neural network based on the seawater characteristic data of all observation points to obtain a particulate attenuation coefficient profile prediction model.

[0066] In an embodiment disclosed in the present invention, as Figure 2 shown, step S200 can be implemented in the following manner:

[0067] Step S201: Construct a training data set based on the seawater characteristic data of all observation points. Each sample includes the sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and particulate attenuation coefficient profile data observed at any observation point during the same time period, as well as the particulate organic carbon profile type at the observation point.

[0068] In an embodiment disclosed in the present invention, the seawater characteristic data of each observation point is composed of matching samples according to the observation time period. Taking one sample as an example, this sample includes the sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and particulate attenuation coefficient profile data observed at a certain observation point during a certain time period, as well as the particulate organic carbon profile type at this observation point.

[0069] All samples are divided into a training data set and a test data set. For example, 70% of the samples are used to construct the training data set, and the remaining 30% of the samples are used to construct the test data set.

[0070] Step S202: Use the sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and particulate organic carbon profile type in the sample as input data, and the particulate attenuation coefficient profile data as output data to train the neural network model and verify it using the 5-fold cross-validation method. Finally, obtain the particulate attenuation coefficient profile prediction model.

[0071] In an embodiment disclosed by the present invention, before training the prediction model, first preprocess the seawater characteristic data of all observation points. For example:

[0072] Detect outliers and remove abnormal samples; use Min-Max normalization or Z-score normalization to normalize the samples to ensure that data with different dimensions can be input into the neural network.

[0073] The network architecture of the neural network model includes an input layer, a hidden layer, an activation function, and an output layer. Among them, the input layer includes 4 input variables (sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and particulate organic carbon profile type); the hidden layer can have 2 to 3 layers, with 16 to 64 neurons in each layer (adjusted according to actual needs); the activation function is ReLU, which is used to enhance the non-linear modeling ability; the output layer includes 1 output variable (particulate attenuation coefficient profile data).

[0074] The neural network model can use the mean squared error (MSE) as the loss function and the Adam optimizer as the optimization algorithm.

[0075] In addition, in the embodiment of the present invention, 5-fold cross-validation is used for the neural network model to improve the model stability and prevent overfitting. The verification process is as follows: divide the training data into 5 parts, select 4 parts as the training set and 1 part as the verification set each time, train 5 times, and finally take the average of the performance.

[0076] After completing the training of the model, use the test data set to test the model performance. For example, calculate the following performance metrics: mean squared error (MSE), root mean squared error (RMSE), coefficient of determination, etc.

[0077] Step S300: Identify the ocean basin pixels in the target remote sensing image and determine the particulate organic carbon profile type of each ocean basin pixel in a preset manner.

[0078] In an embodiment disclosed by the present invention, as Figure 3 shown, the following sub-steps can be used to complete Step S300:

[0079] Step S301: Obtain the water depth corresponding to each pixel in the target remote sensing image.

[0080] The water depth at each pixel can be obtained from the ETOPO1 global relief dataset, which records the global terrain and ocean depth dataset, covering all land and ocean areas of the world.

[0081] According to the geographic coordinate information of the remote sensing image and the row and column numbers of the pixel in the image, the corresponding longitude and latitude of the pixel are determined, and then the water depth at the pixel is obtained through the ETOPO1 global relief dataset.

[0082] Step S302: Determine the pixels with water depth exceeding the preset depth threshold as the pixels in the ocean basin area.

[0083] In the embodiment disclosed in the present invention, the preset depth threshold is 200 meters, and the pixels with water depth exceeding 200 meters are determined as the pixels representing the ocean basin area.

[0084] Step S303: Calculate the euphotic layer depth corresponding to each pixel in the ocean basin area, and judge whether the euphotic layer depth is greater than the preset classification threshold.

[0085] In an embodiment disclosed in the present invention, the following method is used to determine the euphotic layer depth corresponding to each pixel in the ocean basin area.

[0086] kvis = k1 + k2 / (1 + Z)^0.5

[0087] TE = kvis × Z

[0088] Zeu = min(abs(TE - 4.605))

[0089] Wherein, Z is the water depth at the pixel; kvis is the diffuse attenuation coefficient of the water body; TE is the optical depth; Zeu is obtained from the depth when TE is closest to 4.605; k1 and k2 are calculated from the inherent optical quantities of the water body through the following formulas:

[0090] k1 = (x0 + x1 × at490^0.5 + x2 × bbt490) × (1 + a0 × sin( ));

[0091] k2 = (c0 + c1 × at490 + c2 × bbt490) × (a1 + a2 × cos( ));

[0092] Wherein, x0 = -0.057; x1 = 0.482; x2 = 4.221; c0 = 0.183; c1 = 0.702; c2 = -2.567; a0 = 0.090; a1 = 1.465; a2 = -0.667; at490 and bbt490 are the total absorption coefficient and the backscattering coefficient at a wavelength of 490 nm below the sea surface, respectively, and are obtained from the remote sensing image metadata; is the solar zenith angle.

[0093] For each pixel in the ocean basin area, it is determined whether the euphotic layer depth is greater than a preset classification threshold.

[0094] In a specific embodiment disclosed by the present invention, through the research and analysis of the euphotic layer depth of the water body and the type of particulate organic carbon profile over the years, the preset classification threshold is set to 80 meters, that is, it is determined whether the euphotic layer depth at the pixel is greater than 80 meters.

[0095] If so, it is determined that the type of particulate organic carbon profile corresponding to the pixel is the Gaussian-like type;

[0096] If not, it is determined that the type of particulate organic carbon profile corresponding to the pixel is the exponential decay type.

[0097] The present invention is applied to the ocean basin area, and the types of particulate organic carbon profiles in the ocean basin area are mainly the Gaussian-like type and the exponential decay type. Therefore, other types are not distinguished in the present invention.

[0098] Step S400: For each pixel in the ocean basin area, the obtained sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and the corresponding particulate organic carbon profile type are input into the prediction model to obtain particulate attenuation coefficient profile data.

[0099] In an embodiment disclosed by the present invention, as Figure 4 shown, the method disclosed by the present invention further includes the following steps:

[0100] Step S010: Preprocess the target remote sensing image.

[0101] For example, the following methods can be used to preprocess the remote sensing image.

[0102] (1) Radiometric correction: Convert the digital number (DN) of the image to a physical unit (such as radiance, reflectance, etc.).

[0103] (2) Atmospheric correction: For example, Rayleigh scattering correction to remove the influence of atmospheric molecular scattering; or aerosol correction to reduce the interference of suspended particles on the water body reflectance.

[0104] (3) Geometric correction: Project the image onto a unified coordinate system (WGS84, UTM) to correct the geometric distortion caused by terrain and sensor tilt.

[0105] (4) Noise removal: For example, cloud removal processing, using a cloud mask to remove the cloud-covered area in the image; or water body masking processing to exclude areas with serious influence of water body edges and sediment.

[0106] Step S020: Obtain the sea surface chlorophyll concentration, seawater temperature profile data, and seawater salinity profile data corresponding to each pixel respectively.

[0107] In an embodiment disclosed in the present invention, the sea surface chlorophyll concentration at each pixel can be obtained using publicly available remote sensing products. Alternatively, a remote sensing inversion algorithm can be used to calculate the sea surface chlorophyll concentration at the pixel.

[0108] In addition, the seawater temperature profile data and seawater salinity profile data are obtained based on the HYCOM global ocean model. HYCOM is a global three-dimensional ocean numerical model. After obtaining the longitude and latitude corresponding to the pixel, the seawater temperature profile data and seawater salinity profile data at the pixel can be determined through this model.

[0109] Step S500: Obtain in advance the conversion relationship between the particulate organic carbon concentration and the particulate attenuation coefficient in the sea basin area.

[0110] Obtain in advance multiple sets of historical data of the particulate organic carbon concentration and the particulate attenuation coefficient matching at multiple observation points in the sea basin area, and perform regression analysis on their historical data to obtain the following conversion relationship:

[0111] POC = 10 ^ ((log(cp660) + 1.71) / 1.27)

[0112] where cp660 is the particulate attenuation coefficient; POC is the particulate organic carbon concentration in the upper layer of seawater.

[0113] Step S600: Based on the conversion relationship, obtain the particulate organic carbon profile data of each pixel in the sea basin area according to the corresponding particulate attenuation coefficient profile data, and construct a three-dimensional field of the particulate organic carbon in the upper layer of seawater in the target sea basin area.

[0114] In an embodiment disclosed in the present invention, as Figure 5 shown, construct the three-dimensional field of the particulate organic carbon in the upper layer of seawater in the target sea basin area in the following manner:

[0115] Step S601: For the particulate organic carbon profile data of each pixel in the sea basin area, perform a moving average process to obtain the vertical distribution data of the particulate organic carbon concentration of the pixels in the sea basin area.

[0116] MATLAB can be used to implement the moving average process on the particulate organic carbon profile data to obtain the vertical distribution data. For example, the vertical distribution data can be obtained at intervals of 1 meter to reduce fluctuations, eliminate extreme outliers, make the profile smoother, and at the same time, make the hierarchical structure of the profile data clear, which is helpful for analyzing the distribution law of the particulate organic carbon concentration at different depths.

[0117] Step S602: Perform median filtering for smoothing on the planar data composed of the vertical distribution data of the particulate organic carbon concentration of all pixel cells in the ocean basin area.

[0118] To improve data quality, make the vertical distribution of the particulate organic carbon concentration smoother, and avoid the influence of extreme values, the present invention performs median filtering and smoothing on the planar data composed of the vertical distribution data of the particulate organic carbon concentration of all pixel cells in the ocean basin area using MATLAB to reduce noise and maintain the distribution trend of the particulate organic carbon concentration.

[0119] Step S603: Construct a three-dimensional field based on the processed particulate organic carbon concentration data.

[0120] Based on the data of all depth layers, construct a complete three-dimensional field of the particulate organic carbon concentration, and intuitively output the distribution of the particulate organic carbon concentration data at different latitudes, longitudes, and depths.

[0121] Another embodiment of the present invention provides an upper-layer seawater particulate organic carbon profile structure calculation system, which can implement the upper-layer seawater particulate organic carbon profile structure calculation method disclosed in the foregoing embodiments.

[0122] The various embodiments of the present disclosure have been described above. The above description is exemplary and not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations are obvious to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The choice of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements to the technologies in the market, or to enable other ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for calculating the profile structure of particulate organic carbon in upper seawater, characterized in that, Applied to the ocean basin area, the method includes: Obtaining seawater characteristic data of multiple observation points in the ocean basin area, and the seawater characteristic data of each observation point at least includes sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, particulate organic carbon profile type, and particulate attenuation coefficient profile data; Training a neural network based on the seawater characteristic data of all observation points to obtain a particulate attenuation coefficient profile prediction model; Identifying ocean basin area pixels in the target remote sensing image, and determining the particulate organic carbon profile type of each ocean basin area pixel by a preset method; For each ocean basin area pixel, inputting the obtained sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and the corresponding particulate organic carbon profile type into the prediction model to obtain particulate attenuation coefficient profile data; Pre-obtaining the conversion relationship between particulate organic carbon concentration and particulate attenuation coefficient in the ocean basin area, including: Determining the conversion relationship between particulate organic carbon concentration and particulate attenuation coefficient in the ocean basin area according to the following formula: POC = 10^((log(cp660)+1.71) / 1.27) where cp660 is the particulate attenuation coefficient; POC is the particulate organic carbon concentration in the upper layer of seawater; Based on the conversion relationship, obtaining the particulate organic carbon profile data of each ocean basin area pixel according to the corresponding particulate attenuation coefficient profile data, and constructing a three-dimensional field of particulate organic carbon in the upper layer of seawater in the target ocean basin area.

2. The method according to claim 1, wherein The training the neural network based on the seawater characteristic data of all observation points to obtain a particulate attenuation coefficient profile prediction model includes: Constructing a training data set based on the seawater characteristic data of all observation points, and each sample includes the sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and particulate attenuation coefficient profile data observed at any observation point in the same time period, as well as the particulate organic carbon profile type at the observation point; Using the sea surface chlorophyll concentration, seawater temperature profile data, seawater salinity profile data, and particulate organic carbon profile type in the sample as input data, and the particulate attenuation coefficient profile data as output data, training the neural network model and validating it by the 5-fold cross-validation method, and finally obtaining a particulate attenuation coefficient profile prediction model.

3. The method according to claim 2, wherein Before performing the step of training the neural network based on the seawater characteristic data of all observation points to obtain a particulate attenuation coefficient profile prediction model, the method further includes: Preprocessing the seawater characteristic data of all observation points, and the preprocessing at least includes outlier removal and normalization processing.

4. The method according to claim 1, wherein The identifying ocean basin area pixels in the target remote sensing image and determining the particulate organic carbon profile type of each ocean basin area pixel by a preset method includes: Obtaining the water depth at each pixel in the target remote sensing image; Determining the pixels with water depth exceeding the preset depth threshold as ocean basin area pixels; Calculating the euphotic layer depth corresponding to each ocean basin area pixel, and determining whether the euphotic layer depth is greater than the preset classification threshold; If so, determining that the particulate organic carbon profile type corresponding to the pixel is Gaussian-like; If not, determining that the particulate organic carbon profile type corresponding to the pixel is exponential decay type.

5. The method according to claim 4, wherein Calculating the euphotic layer depth corresponding to each ocean basin area pixel by the following formula: kvis = k1 + k2 / (1 + Z)^0.5 TE = kvis × Z Zeu = min(abs(TE - 4.605)) Wherein, Z is the water depth at the pixel; kvis is the diffuse attenuation coefficient of the water body; TE is the optical depth; k1 and k2 are obtained by calculating the inherent optical quantities of the water body through the following formula: k1 = (x0 + x1 × at490^0.5 + x2 × bbt490) × (1 + a0 × sin(θ)); k2 = (c0 + c1 × at490 + c2 × bbt490) × (a1 + a2 × cos(θ)); Wherein, x0 = -0.057; x1 = 0.482; x2 = 4.221; c0 = 0.183; c1 = 0.702; c2 = -2.567; a0 = 0.090; a1 = 1.465; a2 = -0.667; at490 and bbt490 are the total absorption coefficient and the backscattering coefficient at a wavelength of 490 nm below the sea surface, respectively, and are obtained from the remote sensing image metadata; θ is the solar zenith angle.

6. The method according to claim 1, characterized in that, The method further includes: Preprocessing the target remote sensing image; Respectively obtaining the sea surface chlorophyll concentration, the seawater temperature profile data, and the seawater salinity profile data corresponding to each pixel.

7. The method according to claim 6, wherein The respectively obtaining the sea surface chlorophyll concentration, the seawater temperature profile data, and the seawater salinity profile data corresponding to each pixel includes: Obtaining the sea surface chlorophyll concentration by using publicly available remote sensing products; Obtaining the seawater temperature profile data and the seawater salinity profile data based on the HYCOM global ocean model.

8. The method according to claim 1, wherein Construct the three-dimensional field of the upper-layer seawater particulate organic carbon in the target sea basin area in the following manner: For the particulate organic carbon profile data of each pixel in the sea basin area, perform a moving average process to obtain the vertical distribution data of the particulate organic carbon concentration of the pixel in the sea basin area; Perform smoothing processing on the planar data composed of the vertical distribution data of the particulate organic carbon concentration of all pixels in the sea basin area by using median filtering; Construct a three-dimensional field based on the processed particulate organic carbon concentration data.

9. A system for calculating the profile structure of particulate organic carbon in upper seawater, characterized in that, The system executes the method for calculating the upper-layer seawater particulate organic carbon profile structure according to any one of claims 1-8.

Images