A method for correcting Argo conductivity data based on the influence of internal waves in the thermohaline layer

Abstract

The present invention relates to the field of ocean conductivity data processing, and discloses an Argo conductivity data correction method based on the influence of thermohaline cline internal waves, comprising the following steps: installing a thermohaline depth sensor and an acoustic Doppler current profiler on an Argo buoy, sampling seawater in a vertical direction, and obtaining the temperature, salinity, and current profile data of the seawater; determining the upper and lower boundary depths of the thermohaline cline and the thickness of the thermohaline cline based on the acquired data, calculating the internal wave frequency and the internal wave number, calculating the temperature internal wave amplitude and the salinity internal wave amplitude, as well as the temperature internal wave phase and the salinity internal wave phase; establishing a conductivity data correction model based on the calculated parameters; and dynamically correcting the conductivity data using the model. The method disclosed by the present invention can significantly improve the accuracy and reliability of conductivity data, and can provide support for research in many fields such as the physical characteristics of the ocean thermohaline cline, the marine ecological environment, and the circulation of marine materials.

Description

Technical Field

[0001] The invention relates to the field of ocean conductivity data processing, and in particular to an Argo conductivity data correction method based on the influence of thermohaline cline internal waves. Background Art

[0002] In the past, the calibration of Argo float conductivity data was mostly based on fixed empirical models of temperature and salinity, or regular laboratory calibration data was used to compensate for the systematic deviation of the sensor. However, this traditional calibration method is difficult to cope with the complex and changeable environment of the ocean, especially in the thermohaline region. The thermohaline is a special water layer in the ocean where the temperature and salinity change dramatically with depth. There are complex physical, chemical and biological processes inside it. These processes are intertwined and jointly affect the change of seawater conductivity. In the thermohaline, due to the rapid changes in temperature and salinity, there are obvious differences in seawater density. This density stratification provides conditions for the generation of internal waves. After incorporating the relevant parameters of internal waves into the analysis, it is found that the change of conductivity is closely related to the internal wave activity. When the internal wave amplitude is large and the frequency is high, the fluctuation of conductivity in the thermohaline is also more intense. The crests and troughs of internal waves move in the vertical direction, and their movement process will cause fluctuations in parameters such as temperature, salinity and flow velocity of seawater in the thermohaline, which in turn affects physical quantities such as conductivity.

[0003] In actual ocean observations, these neglected factors often lead to large errors in the conductivity data collected by Argo floats when they cross the thermohaline layer, which cannot accurately reflect the true physical characteristics of the ocean, thus affecting the research and understanding of key ocean processes such as ocean circulation, heat transfer, and ecosystems. Therefore, a new method that can dynamically correct thermohaline conductivity data is urgently needed to improve the accuracy and reliability of ocean observation data and provide a solid data foundation for in-depth research on changes in the ocean environment. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides an Argo conductivity data correction method based on the influence of thermohaline internal waves, so as to achieve the purpose of improving the accuracy and reliability of conductivity data.

[0005] To achieve the above object, the technical solution of the present invention is as follows:

[0006] The method for correcting Argo conductivity data based on the influence of internal waves in the thermohaline layer includes the following steps:

[0007] Step 1: Install a temperature, salinity, depth sensor and an acoustic Doppler current profiler on the Argo buoy to sample and measure the seawater at fixed intervals in the vertical direction to obtain the profile data of the temperature and salinity of the seawater and the current profile data of the seawater;

[0008] Step 2: Calculate the vertical gradients of temperature and salinity based on the obtained profile data of seawater temperature and salinity; use the K-Means clustering algorithm to determine the upper and lower boundary depths of the thermohaline layer and the thickness of the thermohaline layer;

[0009] Step 3, calculating the frequency spectrum distribution of the flow velocity according to the acquired flow velocity profile data of the seawater, obtaining the internal wave frequency in the obtained frequency spectrum, and calculating the internal wave number;

[0010] Step 4: Calculate the temperature fluctuation sequence based on the acquired long-term temperature observation data sequence to obtain the temperature internal wave amplitude; calculate the salinity fluctuation sequence based on the acquired long-term salinity observation data sequence to obtain the salinity internal wave amplitude;

[0011] Step 5, defining a temperature error function and a salinity error function according to the obtained profile data of seawater temperature and salinity, and obtaining estimated values ​​of the temperature internal wave phase and the salinity internal wave phase by minimizing the temperature error function and the salinity error function;

[0012] Step 6, considering the influence of internal waves and the thermohaline cline, a conductivity data correction model is established according to the solved thermohaline cline thickness, internal wave frequency, internal wave wave number, temperature internal wave amplitude, salinity internal wave amplitude, temperature internal wave phase and salinity internal wave phase;

[0013] Step 7: Use the established conductivity data correction model to dynamically correct the Argo float conductivity data.

[0014] In the above scheme, in step 2, the central difference method is used to calculate the vertical gradients of temperature and salinity:

[0015] Temperature gradient: ;

[0016] Salinity gradient: ;

[0017] in, Indicates temperature, Indicates depth, Indicates the vertical depth The depth increment is express Temperature at depth, express Temperature at depth; Indicates salinity, express Salinity at depth, express Salinity at depth.

[0018] In the above scheme, in step 2, the K-Means clustering algorithm is used to determine the upper and lower boundary depths of the thermohaline layer and the thickness of the thermohaline layer as follows:

[0019] (1) Initialize cluster centers: Randomly select 3 data points as the initial cluster centers of different water layers , , ; For two data points and , the weighted Euclidean distance between them The calculation formula is: ,in and Respectively indicate temperature and salinity The weight coefficient of

[0020] (2) Assign data points to three clusters: For each data point , calculate its weighted Euclidean distance to the three initial cluster centers, and assign each data point to the class represented by the nearest cluster center;

[0021] (3) Update cluster centers: For each cluster , calculate the weighted mean of all data points in the cluster as the new cluster center; set cluster There are Data points , , then the new cluster center The update formula is:

[0022] ;

[0023] (4) Repeat the above assignment and update steps until the cluster center no longer changes significantly; for each cluster , find the minimum depth corresponding to the data point and maximum depth ; When the temperature gradient and salinity gradient exceed their respective thresholds at the same time, the depth interval belongs to the thermohaline cline;

[0024] (5) After determining the cluster where the thermohaline is located, the corresponding minimum depth and maximum depth As the upper and lower boundaries of the thermohaline layer, the thickness of the thermohaline layer .

[0025] In the above scheme, in step 3, the fast Fourier transform algorithm is used to perform spectrum analysis on the velocity profile data of seawater to obtain the spectrum distribution of the velocity. ,in represents the internal wave frequency, Indicates depth; find the frequency corresponding to the energy peak associated with the internal wave in the spectrum graph , which is the internal wave frequency; according to the linear internal wave theory, the internal wave number Internal wave frequency , gravitational acceleration g and the buoyancy frequency N of seawater are related as follows: , and the internal wave number is calculated .

[0026] In the above scheme, in step 4, the temperature fluctuation sequence is calculated based on the acquired long-term temperature observation data sequence, and the internal wave amplitude based on the temperature data is obtained as follows:

[0027] (1) Calculate the temperature average and temperature fluctuation based on the long-term temperature observation data series measured in the thermohaline layer:

[0028] ;

[0029] ;

[0030] Where n=1,2,…,M, M represents the total number of observed data points, Indicates the depth Department, The temperature observation value at time, Indicates the measured data at depth The average temperature at Indicates the depth Department, Temperature fluctuation value at each moment;

[0031] (2) Obtain the autocorrelation function of temperature fluctuation:

[0032] ;

[0033] in, is the time lag, represents the depth, M represents the total number of observed data points, and n represents the index variable used to traverse the entire data set M;

[0034] (3) The power spectrum density of temperature is obtained by Fourier transforming the autocorrelation function:

[0035] ;

[0036] in, represents the internal wave frequency, ;

[0037] (4) Calculate the internal wave temperature variance:

[0038] ;

[0039] in, and Respectively represent the maximum and minimum values ​​of the internal wave frequency;

[0040] (5) The internal wave amplitude solved based on the temperature data is obtained, that is, the temperature internal wave amplitude:

[0041] .

[0042] In the above scheme, in step 4, the salinity fluctuation sequence is calculated based on the acquired long-term salinity observation data sequence, and the internal wave amplitude based on the salinity data is obtained as follows:

[0043] (1) Calculate the average salinity and salinity fluctuation value based on the long-term salinity observation data series measured in the thermohaline layer:

[0044] ;

[0045] ;

[0046] Where n=1,2,…,M, M is the total number of observed data points, For the depth Department, The salinity observation value at time Indicates the measured data at depth The average salinity at For the depth Department, Salinity fluctuation value at the moment;

[0047] (2) Obtain the autocorrelation function of salinity fluctuation:

[0048] ;

[0049] in, Indicates time lag, represents the depth, M represents the total number of observed data points, and n represents the index variable used to traverse the entire data set M;

[0050] (3) The power spectrum density of salinity is obtained by Fourier transforming the autocorrelation function:

[0051] ;

[0052] in, represents the internal wave frequency, ;

[0053] (4) Calculate the internal wave salinity variance:

[0054] ;

[0055] in, and Respectively represent the maximum and minimum values ​​of the internal wave frequency;

[0056] (5) The internal wave amplitude solved based on salinity data is obtained, namely, the salinity internal wave amplitude:

[0057] .

[0058] In the above scheme, the specific method of step 5 is as follows:

[0059] (1) For temperature data, define the temperature error function:

[0060] ;

[0061] Where M represents the total number of observed data points, Indicates the depth Department, The actual observed temperature value at time, Indicates the measured data at depth The average temperature at Indicates depth The temperature internal wave amplitude at represents the internal wave frequency, represents the wave number;

[0062] For salinity data, define the salinity error function:

[0063] ;

[0064] in, Indicates the depth Department, The actual observed salinity value at time Indicates the measured data at depth The average salinity at Indicates depth The amplitude of salinity internal wave at ;

[0065] (2) Calculate the partial derivatives of the temperature error function and the salinity error function with respect to the temperature internal wave phase and the salinity internal wave phase. According to the composite function derivation method, we can obtain:

[0066] ;

[0067] (3) Use batch gradient descent optimization algorithm to minimize the error function and , through iterative updating, the estimated values ​​of the temperature internal wave phase and the salinity internal wave phase are obtained:

[0068] ;

[0069] in, represents the learning rate, represents the temperature internal wave phase before iterative update, represents the salinity internal wave phase before iterative update, represents the temperature internal wave phase after iterative update, Represents the salinity internal wave phase after iterative update.

[0070] In the above scheme, the specific method of step 6 is as follows:

[0071] Assume that the temperature fluctuation and salinity fluctuation caused by internal waves are:

[0072] ;

[0073] in, and Respectively represent the measured data at depth The average temperature and salinity at and Depth The temperature internal wave amplitude and salinity internal wave amplitude at and are the temperature internal wave phase and the salinity internal wave phase, represents depth, t represents time, represents the internal wave frequency, represents the wave number, and Respectively indicate temperature and salinity The weight coefficient of

[0074] Substituting the fluctuation values ​​of temperature and salinity into the conductivity and temperature-salinity relationship model, we can obtain the conductivity correction model considering the influence of internal waves:

[0075] ;

[0076] Among them, a, b, c, d, e, and f are coefficients, and a represents the ,salinity When all are zero and the periodic fluctuation caused by internal waves is not considered, the base value of conductivity, b and c represent the linear coefficients of temperature and salinity, respectively, reflecting their respective linear effects on conductivity; d reflects the cross-effect of temperature and salinity; e and f consider the quadratic effects of temperature and salinity, and are used to more accurately describe the complex nonlinear relationship between conductivity and temperature and salinity; Indicates the conductivity correction value considering the influence of internal waves;

[0077] (3) Consider the influence of the thickness of the thermohaline layer on the internal wave parameters:

[0078] For internal wave frequency ,set up ,in, is a constant related to the characteristics of the internal wave source, represents the thickness of the thermohaline layer;

[0079] For the temperature internal wave amplitude and salinity internal wave amplitude ,set up and ,in, and is a constant related to the intensity of the internal wave source, and Represents a specific depth position closely related to the generation or propagation of internal waves;

[0080] The calibration model based on the above updated conductivity data is:

[0081] ;

[0082] in, Updated conductivity correction value to account for the influence of thermohaline thickness.

[0083] In the above scheme, the specific method of step 7 is as follows:

[0084] The temperature internal wave amplitude, salinity internal wave amplitude, temperature internal wave phase and salinity internal wave phase are solved using the actual measured seawater temperature, salinity and depth profile data. The internal wave frequency and internal wave number are calculated based on the actual measured seawater flow velocity profile data and input into the established conductivity data correction model to calculate the corrected conductivity value.

[0085] Through the above technical solution, the Argo conductivity data correction method based on the influence of thermohaline cline internal waves provided by the present invention has the following beneficial effects:

[0086] The correction method provided by the present invention expands the functions at the algorithm level on the basis of currently mature hardware equipment such as ocean conductivity measurement sensors, temperature and salinity sensors, etc., and takes into account the influence of internal waves in the thermohaline cline on the sampling accuracy of Argo conductivity data. By adopting a combination of a clustering algorithm and a correction model, a dynamic correction of thermohaline conductivity data that does not rely on traditional empirical models and fixed parameter settings is achieved. The equipment installation and maintenance costs are low, and the influence of factors such as complex thermohaline structures and internal waves in the thermohaline cline on conductivity data can be effectively dealt with, and the accuracy and reliability of conductivity data can be significantly improved, which can provide support for related research in many fields such as the study of physical characteristics of ocean thermohaline cline, marine ecological environment, and marine material circulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] 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 are briefly introduced below.

[0088] Figure 1 The present invention discloses a flow chart of a method for correcting Argo conductivity data based on the influence of internal waves in the thermohaline cline. DETAILED DESCRIPTION

[0089] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0090] The present invention provides an Argo conductivity data correction method based on the influence of thermohaline internal waves, such as Figure 1 As shown, the following steps are included:

[0091] The method for correcting Argo conductivity data based on the influence of internal waves in the thermohaline layer includes the following steps:

[0092] Step 1: Install a temperature-salinity-depth sensor (CTD) and an acoustic Doppler current profiler (ADCP) on the Argo float, and sample the seawater at the coordinates of 19.5°N, 116.0°E at fixed intervals in the vertical direction to obtain the temperature of the seawater. ,salinity Profile data and seawater velocity profile data ,in Indicates time, Indicates depth, ranging from sea surface to the maximum depth measured.

[0093] Step 2: Calculate the vertical gradients of temperature and salinity based on the obtained profile data of seawater temperature and salinity; use the K-Means clustering algorithm to determine the upper and lower boundary depths of the thermohaline layer and the thickness of the thermohaline layer.

[0094] First, the vertical gradients of temperature and salinity are calculated using the central difference method:

[0095] Temperature gradient: ;

[0096] Salinity gradient: ;

[0097] in, Indicates temperature, Indicates depth, Indicates the vertical depth The depth increment is express Temperature at depth, express Temperature at depth; Indicates salinity, express Salinity at depth, express Salinity at depth.

[0098] Then, the K-Means clustering algorithm was used to determine the upper and lower boundary depths of the thermohaline cline and the thickness of the thermohaline cline. Since the thermohaline is expected to appear as a relatively independent category, in addition to the thermohaline cline, there are usually other relatively stable water layers in the ocean water body, such as the surface mixed layer and the deep stable layer, so the data were clustered into three categories.

[0099] (1) Initialize cluster centers: Randomly select three data points as the initial cluster centers of different water layers, denoted as , , ; For two data points and , the weighted Euclidean distance between them The calculation formula is: ,in and Respectively indicate temperature and salinity The weight coefficient of

[0100] (2) Assign data points to three clusters: For each data point , calculate its weighted Euclidean distance to the three initial cluster centers , where l=1,2,3;

[0101] Data Points and the initial cluster centers The weighted Euclidean distance between Calculated as: ; Data points and the initial cluster centers The weighted Euclidean distance between Calculated as: ; at data point and the initial cluster centers The weighted Euclidean distance between Calculated as: ;

[0102] By comparison , and , find the minimum distance value. Assume is the smallest, then the data point was assigned to In the cluster centered on ,in Represents data points Assigned to The cluster category centered on

[0103] Assign each data point to the class represented by the nearest cluster center according to the above method;

[0104] (3) Update cluster centers: For each cluster , calculate the weighted mean of all data points in the cluster as the new cluster center; set cluster There are Data points , , then the new cluster center The update formula is:

[0105] ;

[0106] (4) Repeat the above assignment and update steps until the cluster center no longer changes significantly; for each cluster , find the minimum depth corresponding to the data point and maximum depth ; When the temperature gradient and salinity gradient exceed their respective thresholds at the same time, the depth interval belongs to the thermohaline cline;

[0107] (5) After determining the cluster where the thermohaline is located, the corresponding minimum depth and maximum depth As the upper and lower boundaries of the thermohaline layer, the thickness of the thermohaline layer .

[0108] Step 3, calculate the frequency spectrum distribution of the flow velocity based on the acquired seawater flow velocity profile data, obtain the internal wave frequency in the obtained spectrum diagram, and calculate the internal wave number.

[0109] Specifically, the fast Fourier transform algorithm is used to perform spectrum analysis on the velocity profile data of seawater to obtain the spectrum distribution of the velocity. ,in represents the internal wave frequency, Indicates depth; find the frequency corresponding to the energy peak associated with the internal wave in the spectrum graph , which is the internal wave frequency; this is because the existence of internal waves will cause the seawater flow velocity to undergo periodic changes at its oscillation frequency, and this frequency component can be extracted through spectrum analysis.

[0110] Then according to the linear internal wave theory, the internal wave number Internal wave frequency , gravitational acceleration g and the buoyancy frequency N of seawater are related as follows: , and the internal wave number is calculated .

[0111] Step 4: Calculate the temperature fluctuation sequence based on the acquired long-term temperature observation data sequence to obtain the temperature internal wave amplitude; calculate the salinity fluctuation sequence based on the acquired long-term salinity observation data sequence to obtain the salinity internal wave amplitude.

[0112] Specifically, the temperature fluctuation sequence is calculated based on the acquired long-term temperature observation data sequence, and the internal wave amplitude based on the temperature data is obtained as follows:

[0113] (1) Calculate the temperature average and temperature fluctuation based on the long-term temperature observation data series measured in the thermohaline layer:

[0114] ;

[0115] ;

[0116] Where n=1,2,…,M, M represents the total number of observed data points, Indicates the depth Department, The temperature observation value at time, Indicates the measured data at depth The average temperature at Indicates the depth Department, Temperature fluctuation value at each moment;

[0117] (2) Obtain the autocorrelation function of temperature fluctuation:

[0118] ;

[0119] in, is the time lag, represents the depth, M represents the total number of observed data points, and n represents the index variable used to traverse the entire data set M;

[0120] (3) The power spectrum density of temperature is obtained by Fourier transforming the autocorrelation function:

[0121] ;

[0122] in, represents the internal wave frequency, ;

[0123] (4) Calculate the internal wave temperature variance:

[0124] ;

[0125] in, and Respectively represent the maximum and minimum values ​​of the internal wave frequency;

[0126] (5) The internal wave amplitude solved based on the temperature data is obtained, that is, the temperature internal wave amplitude:

[0127] .

[0128] Specifically, the salinity fluctuation sequence is calculated based on the acquired long-term salinity observation data sequence, and the internal wave amplitude based on the salinity data is obtained as follows:

[0129] (1) Calculate the average salinity and salinity fluctuation value based on the long-term salinity observation data series measured in the thermohaline layer:

[0130] ;

[0131] ;

[0132] Where n=1,2,…,M, M is the total number of observed data points, For the depth Department, The salinity observation value at time Indicates the measured data at depth The average salinity at For the depth Department, Salinity fluctuation value at the moment;

[0133] (2) Obtain the autocorrelation function of salinity fluctuation:

[0134] ;

[0135] in, Indicates time lag, represents the depth, M represents the total number of observed data points, and n represents the index variable used to traverse the entire data set M;

[0136] (3) The power spectrum density of salinity is obtained by Fourier transforming the autocorrelation function:

[0137] ;

[0138] in, represents the internal wave frequency, ;

[0139] (4) Calculate the internal wave salinity variance:

[0140] ;

[0141] in, and Respectively represent the maximum and minimum values ​​of the internal wave frequency;

[0142] (5) The internal wave amplitude solved based on salinity data is obtained, namely, the salinity internal wave amplitude:

[0143] .

[0144] Step 5, based on the acquired seawater temperature and salinity profile data, define the temperature error function and the salinity error function, and obtain the estimated values ​​of the temperature internal wave phase and the salinity internal wave phase by minimizing the temperature error function and the salinity error function.

[0145] The specific method is as follows:

[0146] (1) For temperature data, define the temperature error function:

[0147] ;

[0148] Where M represents the total number of observed data points, Indicates the depth Department, The actual observed temperature value at time, Indicates the measured data at depth The average temperature at Indicates depth The temperature internal wave amplitude at represents the internal wave frequency, represents the wave number;

[0149] For salinity data, define the salinity error function:

[0150] ;

[0151] in, Indicates the depth Department, The actual observed salinity value at time Indicates the measured data at depth The average salinity at Indicates depth The amplitude of salinity internal wave at ;

[0152] (2) Calculate the partial derivatives of the temperature error function and the salinity error function with respect to the temperature internal wave phase and the salinity internal wave phase. According to the composite function derivation method, we can obtain:

[0153] ;

[0154] (3) Use batch gradient descent optimization algorithm to minimize the error function and , through iterative updating, the estimated values ​​of the temperature internal wave phase and the salinity internal wave phase are obtained:

[0155] ;

[0156] in, represents the learning rate, represents the temperature internal wave phase before iterative update, represents the salinity internal wave phase before iterative update, represents the temperature internal wave phase after iterative update, Represents the salinity internal wave phase after iterative update.

[0157] Step 6, considering the influence of internal waves and the thermohaline cline, a conductivity data correction model is established based on the solved thermohaline cline thickness, internal wave frequency, internal wave wave number, temperature internal wave amplitude, salinity internal wave amplitude, temperature internal wave phase and salinity internal wave phase.

[0158] The specific method is as follows:

[0159] Under the action of internal waves, the temperature and salinity in the thermohaline layer will fluctuate, thus affecting the conductivity. Assuming that the temperature fluctuation and salinity fluctuation formula caused by internal waves are:

[0160] ;

[0161] in, and Respectively represent the measured data at depth The average temperature and salinity at and Depth The temperature internal wave amplitude and salinity internal wave amplitude at and are the temperature internal wave phase and the salinity internal wave phase, represents depth, t represents time, represents the internal wave frequency, represents the wave number, and Respectively indicate temperature and salinity The weight coefficient of

[0162] Substituting the fluctuation values ​​of temperature and salinity into the conductivity and temperature-salinity relationship model, we can obtain the conductivity correction model considering the influence of internal waves:

[0163] ;

[0164] Among them, a, b, c, d, e, and f are coefficients, and a represents the ,salinity are all zero, and the baseline value of conductivity when the periodic fluctuation caused by internal waves is not considered; b and c represent the linear coefficients of temperature and salinity, respectively, reflecting their respective linear effects on conductivity; d reflects the cross-effect of temperature and salinity; e and f consider the quadratic effects of temperature and salinity, and are used to more accurately describe the complex nonlinear relationship between conductivity and temperature and salinity; Indicates the conductivity correction value considering the influence of internal waves;

[0165] (3) Consider the influence of the thickness of the thermohaline layer on the internal wave parameters:

[0166] For internal wave frequency ,set up ,in, is a constant related to the characteristics of the internal wave source, represents the thickness of the thermohaline layer;

[0167] For the temperature internal wave amplitude and salinity internal wave amplitude ,set up and ,in, and is a constant related to the intensity of the internal wave source, and Represents a specific depth position closely related to the generation or propagation of internal waves;

[0168] The calibration model based on the above updated conductivity data is:

[0169] ;

[0170] in, Updated conductivity correction value to account for the influence of thermohaline thickness.

[0171] Then the coefficient of determination To evaluate the model. It is used to measure the degree of fit of the model to the data. It reflects the proportion of the variation of the dependent variable (corrected conductivity) that can be explained by the independent variables (input parameters in the model, such as temperature, salinity, internal wave parameters, etc.). The calculation formula is:

[0172] ;

[0173] Where n=1,2,…,M, M is the total number of observed data points, Indicates depth, Indicates time, Indicates the conductivity data correction value, Indicates the true value of conductivity data, is the average of the true conductivity values.

[0174] The value range is from 0 to 1. The closer it is to 1, the better the model fits the conductivity correction, that is, the model can make good use of the input parameters to correct the conductivity. Finally, the model is optimized and adjusted according to the evaluation results to ensure the best conductivity correction effect.

[0175] Step 7: Use the established conductivity data correction model to dynamically correct the Argo float conductivity data.

[0176] The specific method is as follows:

[0177] The temperature internal wave amplitude, salinity internal wave amplitude, temperature internal wave phase and salinity internal wave phase are solved using the actual measured seawater temperature, salinity and depth profile data. The internal wave frequency and internal wave number are calculated based on the actual measured seawater flow velocity profile data and input into the established conductivity data correction model to calculate the corrected conductivity value.

[0178] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. The Argo conductivity data correction method based on the influence of thermohaline internal waves is characterized by: The steps include: Step 1: Install a temperature, salinity, depth sensor and an acoustic Doppler current profiler on the Argo buoy to sample and measure the seawater at fixed intervals in the vertical direction to obtain the profile data of the temperature and salinity of the seawater and the current profile data of the seawater; Step 2: Calculate the vertical gradients of temperature and salinity based on the obtained profile data of seawater temperature and salinity; use the K-Means clustering algorithm to determine the upper and lower boundary depths of the thermohaline layer and the thickness of the thermohaline layer; Step 3, calculating the frequency spectrum distribution of the flow velocity according to the acquired flow velocity profile data of the seawater, obtaining the internal wave frequency in the obtained frequency spectrum, and calculating the internal wave number; Step 4: Calculate the temperature fluctuation sequence based on the acquired long-term temperature observation data sequence to obtain the temperature internal wave amplitude; calculate the salinity fluctuation sequence based on the acquired long-term salinity observation data sequence to obtain the salinity internal wave amplitude; Step 5, defining a temperature error function and a salinity error function according to the obtained profile data of seawater temperature and salinity, and obtaining estimated values ​​of the temperature internal wave phase and the salinity internal wave phase by minimizing the temperature error function and the salinity error function; Step 6, considering the influence of internal waves and the thermohaline cline, a conductivity data correction model is established according to the solved thermohaline cline thickness, internal wave frequency, internal wave wave number, temperature internal wave amplitude, salinity internal wave amplitude, temperature internal wave phase and salinity internal wave phase; Step 7: Use the established conductivity data correction model to dynamically correct the Argo float conductivity data.

2. The Argo conductivity data correction method based on the influence of thermohaline internal waves according to claim 1 is characterized in that: In step 2, the central difference method is used to calculate the vertical gradients of temperature and salinity: Temperature gradient: ; Salinity gradient: ; in, Indicates temperature, Indicates depth, Indicates the vertical depth The depth increment is express Temperature at depth, express Temperature at depth; Indicates salinity, express Salinity at depth, express Salinity at depth.

3. The Argo conductivity data correction method based on the influence of thermohaline internal waves according to claim 1 is characterized in that: In step 2, the K-Means clustering algorithm is used to determine the upper and lower boundary depths of the thermohaline layer and the thickness of the thermohaline layer as follows: (1) Initialize cluster centers: Randomly select 3 data points as the initial cluster centers of different water layers , , , for two data points and , the weighted Euclidean distance between them The calculation formula is: ,in and Respectively indicate temperature and salinity The weight coefficient of (2) Assign data points to three clusters: For each data point , calculate its weighted Euclidean distance to the three initial cluster centers, and assign each data point to the class represented by the nearest cluster center; (3) Update cluster centers: For each cluster , calculate the weighted mean of all data points in the cluster as the new cluster center; set cluster There are Data points , , then the new cluster center The update formula is: ； (4) Repeat the above assignment and update steps until the cluster center no longer changes significantly; for each cluster , find the minimum depth corresponding to the data point and maximum depth ; When the temperature gradient and salinity gradient exceed their respective thresholds at the same time, the depth interval belongs to the thermohaline cline; (5) After determining the cluster where the thermohaline is located, the corresponding minimum depth and maximum depth As the upper and lower boundaries of the thermohaline layer, the thickness of the thermohaline layer .

4. The Argo conductivity data correction method based on the influence of thermohaline internal waves according to claim 1 is characterized in that: In step 3, the fast Fourier transform algorithm is used to perform spectrum analysis on the velocity profile data of seawater to obtain the spectrum distribution of the velocity. ,in represents the internal wave frequency, Indicates depth; find the frequency corresponding to the energy peak associated with the internal wave in the spectrum graph , which is the internal wave frequency; according to the linear internal wave theory, the internal wave number Internal wave frequency , gravitational acceleration g and the buoyancy frequency N of seawater are related as follows: , and the internal wave number is calculated .

5. The Argo conductivity data correction method based on the influence of thermohaline internal waves according to claim 1 is characterized in that: In step 4, the temperature fluctuation sequence is calculated based on the acquired long-term temperature observation data sequence, and the internal wave amplitude based on the temperature data is obtained as follows: (1) Calculate the temperature average and temperature fluctuation based on the long-term temperature observation data series measured in the thermohaline layer: ； ； Where n=1,2,…,M, M represents the total number of observed data points, Indicates the depth Department, The temperature observation value at time, Indicates the measured data at depth The average temperature at Indicates the depth Department, Temperature fluctuation value at each moment; (2) Obtain the autocorrelation function of temperature fluctuation: ； in, is the time lag, represents the depth, M represents the total number of observed data points, and n represents the index variable used to traverse the entire data set M; (3) The power spectrum density of temperature is obtained by Fourier transforming the autocorrelation function: ； in, represents the internal wave frequency, ; (4) Calculate the internal wave temperature variance: ； in, and Respectively represent the maximum and minimum values ​​of the internal wave frequency; (5) The internal wave amplitude solved based on the temperature data is obtained, that is, the temperature internal wave amplitude: 。 6. The method for correcting Argo conductivity data based on the influence of thermohaline internal waves according to claim 1, characterized in that: In step 4, the salinity fluctuation sequence is calculated based on the acquired long-term salinity observation data sequence, and the internal wave amplitude based on the salinity data is obtained as follows: (1) Calculate the average salinity and salinity fluctuation value based on the long-term salinity observation data series measured in the thermohaline layer: ； ； Where n=1,2,…,M, M is the total number of observed data points, For the depth Department, The salinity observation value at time Indicates the measured data at depth The average salinity at For the depth Department, Salinity fluctuation value at the moment; (2) Obtain the autocorrelation function of salinity fluctuation: ； in, Indicates time lag, represents the depth, M represents the total number of observed data points, and n represents the index variable used to traverse the entire data set M; (3) The power spectrum density of salinity is obtained by Fourier transforming the autocorrelation function: ； in, represents the internal wave frequency, ; (4) Calculate the internal wave salinity variance: ； in, and Respectively represent the maximum and minimum values ​​of the internal wave frequency; (5) The internal wave amplitude solved based on salinity data is obtained, namely, the salinity internal wave amplitude: 。 7. The method for correcting Argo conductivity data based on the influence of thermohaline internal waves according to claim 1, characterized in that: The specific method of step 5 is as follows: (1) For temperature data, define the temperature error function: ； Where M represents the total number of observed data points, Indicates the depth Department, The actual observed temperature value at time, Indicates the measured data at depth The average temperature at Indicates depth The temperature internal wave amplitude at represents the internal wave frequency, represents the wave number; For salinity data, define the salinity error function: ； in, Indicates the depth Department, The actual observed salinity value at time Indicates the measured data at depth The average salinity at Indicates depth The amplitude of salinity internal wave at ; (2) Calculate the partial derivatives of the temperature error function and the salinity error function with respect to the temperature internal wave phase and the salinity internal wave phase. According to the composite function derivation method, we can obtain: ； (3) Use batch gradient descent optimization algorithm to minimize the error function and , through iterative updating, the estimated values ​​of the temperature internal wave phase and the salinity internal wave phase are obtained: ； in, represents the learning rate, represents the temperature internal wave phase before iterative update, represents the salinity internal wave phase before iterative update, represents the temperature internal wave phase after iterative update, Represents the salinity internal wave phase after iterative update.

8. The method for correcting Argo conductivity data based on the influence of thermohaline internal waves according to claim 1, characterized in that: The specific method of step 6 is as follows: Assume that the temperature fluctuation and salinity fluctuation caused by internal waves are: ； in, and Respectively represent the measured data at depth The average temperature and salinity at and Depth The temperature internal wave amplitude and salinity internal wave amplitude at and are the temperature internal wave phase and the salinity internal wave phase, represents depth, t represents time, represents the internal wave frequency, represents the wave number, and Respectively indicate temperature and salinity The weight coefficient of Substituting the fluctuation values ​​of temperature and salinity into the conductivity and temperature-salinity relationship model, we can obtain the conductivity correction model considering the influence of internal waves: ； Among them, a, b, c, d, e, and f are coefficients, and a represents the ,salinity are all zero, and the baseline value of conductivity when the periodic fluctuation caused by internal waves is not considered; b and c represent the linear coefficients of temperature and salinity, respectively, reflecting their respective linear effects on conductivity; d reflects the cross-effect of temperature and salinity; e and f consider the quadratic effects of temperature and salinity, and are used to more accurately describe the complex nonlinear relationship between conductivity and temperature and salinity; Indicates the conductivity correction value considering the influence of internal waves; (3) Consider the influence of the thickness of the thermohaline layer on the internal wave parameters: For internal wave frequency ,set up ,in, is a constant related to the characteristics of the internal wave source, represents the thickness of the thermohaline layer; For the temperature internal wave amplitude and salinity internal wave amplitude ,set up and ,in, and is a constant related to the intensity of the internal wave source, and Represents a specific depth position closely related to the generation or propagation of internal waves; According to the above, the updated conductivity data correction model is: ； in, Updated conductivity correction value to account for the influence of thermohaline thickness.

9. The method for correcting Argo conductivity data based on the influence of thermohaline internal waves according to claim 1, characterized in that: The specific method of step 7 is as follows: The temperature internal wave amplitude, salinity internal wave amplitude, temperature internal wave phase and salinity internal wave phase are solved using the actual measured seawater temperature, salinity and depth profile data. The internal wave frequency and internal wave number are calculated based on the actual measured seawater flow velocity profile data and input into the established conductivity data correction model to calculate the corrected conductivity value.

Images