Regional three-dimensional sound velocity field inversion and grid encryption processing method based on multi-source ocean data

Abstract

The invention discloses a regional three-dimensional sound velocity field inversion and grid encryption processing method based on multi-source ocean data, belongs to the technical field of ocean detection and data processing, is used for three-dimensional sound velocity field inversion, and comprises regional three-dimensional sound velocity field inversion based on multi-source ocean data, temperature and salt field calculation through a regression model, and grid encryption in an Argo buoy data coverage area. Correcting the calculated temperature and salt field by using Argo buoy data, and correcting the temperature and salt field corrected by the Argo buoy data by using high-precision thermosalinity-depth instrument data; performing multi-source data fusion based on hierarchical processing; performing adaptive grid optimization based on gradient analysis; performing regional adjustment on a sound velocity field inversion result by using sea level height abnormal data; and updating the regional three-dimensional sound velocity field in real time. Compared with the prior art, the method has the advantages that a new high-resolution data source is provided for construction of the sound velocity field of the deep sea area, the spatial resolution and the calculation efficiency of the sound velocity field are improved, and the application prospect is wide.

Description

Technical Field

[0001] The invention discloses a regional three-dimensional sound velocity field inversion and grid encryption processing method based on multi-source ocean data, belonging to the technical field of ocean detection and data processing. Background Art

[0002] The sound velocity field is the basic data in the fields of ocean acoustic detection, underwater communication, sonar positioning, etc. At present, the inversion of the sound velocity field mainly relies on a single or a small number of data sources (such as remote sensing, Argo buoys, and shipborne observations), which have problems of insufficient coverage and limited resolution. Although manned submersibles can provide high-precision profile data, their coverage is limited. Autonomous underwater vehicles (AUVs) have the ability to be flexible and maneuverable and operate over a large range. Their coordinated observations with manned submersibles and other data sources can significantly improve the spatial distribution uniformity and coverage of data.

[0003] The existing technology has deficiencies in multi-source data fusion and grid optimization processing, and cannot give full play to the advantages of multi-submersible collaborative observation. The present invention proposes a method for fusion inversion of temperature, salinity and depth data of collaborative multi-submersibles, and improves the resolution of the sound velocity field through dynamic grid optimization, providing an efficient solution for acoustic applications in complex marine environments. Summary of the invention

[0004] The purpose of the present invention is to provide a regional three-dimensional sound velocity field inversion and grid encryption processing method based on multi-source ocean data, so as to solve the problems of insufficient coverage and limited resolution of sound velocity field inversion in the prior art.

[0005] The regional three-dimensional sound velocity field inversion and grid encryption processing method based on multi-source ocean data includes:

[0006] S1. Perform regional three-dimensional sound velocity field inversion based on multi-source ocean data, including data acquisition, data preprocessing, multi-source data fusion and sound velocity field inversion;

[0007] S2. Calculate the temperature-salinity field through the regression model. In the area covered by Argo float data, use Argo float data to correct the calculated temperature-salinity field. Use high-precision temperature-salinity-depth instrument data to correct the temperature-salinity field corrected by Argo float data.

[0008] S3. Multi-source data fusion based on hierarchical processing: remote sensing data and Argo float data are interpolated and fused in shallow waters, and Argo float profile data and submersible temperature-salinity-depth instrument CTD measurement data are fused through weighted averaging and spatial interpolation in medium and deep waters;

[0009] S4. Adaptive mesh optimization based on gradient analysis, with high gradient areas being refined first, followed by medium gradient areas, and low gradient areas maintaining the initial resolution;

[0010] S5. Use sea level anomaly data to make regional adjustments to the sound velocity field inversion results;

[0011] S6. The regional three-dimensional sound velocity field is updated in real time.

[0012] The data acquisition of S1 includes remote sensing data, Argo buoy data and submersible data. The remote sensing data includes sea level height anomaly SSH, sea surface temperature SST and sea surface salinity SSS. The Argo buoy data includes the temperature and salinity profile data provided by the Argo buoy. The submersible data includes the temperature, salinity and depth instrument measurement data of manned submersibles and autonomous underwater vehicles AUV.

[0013] Data preprocessing includes denoising, temporal and spatial alignment, outlier removal, and missing data completion for multi-source data.

[0014] S1's multi-source data fusion includes combining remote sensing data, Argo buoy data and submersible data, and using a dynamic weight algorithm to calculate the observation accuracy factor, spatial coverage factor and timeliness factor based on data accuracy, spatial coverage and time consistency, and perform layered weighted fusion to generate a three-dimensional temperature and salinity field.

[0015] The observation precision factor representing the accuracy level of each type of data source is:

[0016]

[0017] Where P i is the observation precision factor of data source i, σ i is the observation error of data source i, σ j is the observation error of the jth data source, and N is the total number of data sources;

[0018] The spatial coverage factor that represents the spatial distribution characteristics of each type of data source is:

[0019]

[0020] In the formula, A i is the coverage factor of data source i, C i is the coverage area of ​​data source i, C j is the coverage area of ​​the jth data source;

[0021] The timeliness factor indicating the time validity of data is:

[0022]

[0023] Where, T i is the timeliness factor of data source i, Δt iis the time difference between data source i and the current time, λ is the time-effectiveness attenuation coefficient, which can be adjusted according to actual needs, Δt j is the time difference between the jth data source and the current time, and e is a natural constant;

[0024] Calculate the final dynamic weight W of each data source i :

[0025] W i =αA i +βP i +γT i ;

[0026] In the formula, α, β, and γ are weight adjustment coefficients, which are set according to the importance of each indicator in the application, α+β+γ=1;

[0027] Perform a weighted fusion of the temperature and salinity values ​​for each grid cell:

[0028]

[0029]

[0030] In the formula, t f is the weighted fusion value of the grid cell temperature, t i is the temperature value of data source i in the grid cell, s f is the weighted fusion value of grid cell salinity, s i is the salinity value of data source i in the grid cell;

[0031] For data-sparse areas at different depths, interpolation methods are used to complete the temperature and salinity values ​​to ensure the integrity of the three-dimensional field. The interpolation methods include inverse distance weighted method and Kriging interpolation method.

[0032] The inversion of the sound velocity field of S1 includes using the Chen-Millero sound velocity formula to calculate the sound velocity field from the temperature-salinity-depth instrument data:

[0033] c(T,S,P)=c0+α1T+β1S+γ1P+δT 2 +∈TS;

[0034] Where c is the speed of sound, which is a function of seawater temperature T, salinity S and pressure P, c0 is the speed of sound constant under the reference state, α1, β1, γ1 are the first-order linear coefficients of temperature, salinity and pressure, δ is the second-order coefficient of temperature, and ∈ is the temperature-salinity cross-term coefficient.

[0035] S2 includes the calculation of temperature and salt field through regression model:

[0036] T c (x,y,z)=g(SST,z);

[0037] S c (x,y,z)=f(SSH,SST);

[0038] Where, T c (x,y,z) and S c (x, y, z) are the temperature and salinity calculated by the regression model, f and g are empirical functions obtained by fitting historical observation data, and (x, y, z) are the longitude, latitude and depth coordinates;

[0039] In the area covered by Argo float data, the estimated temperature and salinity field is corrected using Argo float data:

[0040] T'(x,y,z)=T c (x,y,z)+ΔT Argo ;

[0041] S'(x,y,z)=S c (x,y,z)+ΔS Argo ;

[0042] Where T'(x,y,z) and S'(x,y,z) are the temperature and salinity corrected by Argo float data, ΔT Argo and ΔS Argo are the temperature and salinity corrected by Argo floats, respectively;

[0043] The area with a radius of 10 km between the manned submersible and the AUV path is selected as the local optimization area, and the temperature and salinity field corrected by the Argo buoy data is corrected using the high-precision temperature and salinity depth instrument data:

[0044] T final (x,y,z)=T'(x,y,z)+ΔT sub ;

[0045] S final (x,y,z)=S'(x,y,z)+ΔS sub ;

[0046] Where, T final (x,y,z) and S final (x, y, z) are the final corrected temperature and salinity, ΔT sub and ΔS sub Temperature and salinity corrected using manned submersibles and AUVs, respectively.

[0047] S3 includes dividing the ocean area into shallow, middle and deep layers according to depth, with the shallow layer being between 0 and 200 meters, the middle layer being between 200 and 1000 meters, and the deep layer being above 1000 meters;

[0048] Shallow data include remote sensing-based SST, SSS, and SSH data. The shallow observation data of Argo floats are interpolated and fused. The shallow observation points of Argo floats are aligned to the remote sensing grid. The temperature and salinity values ​​of each grid point are combined with the remote sensing estimated values ​​and Argo measured values ​​and fused through interpolation methods:

[0049] T fusion (x,y,z)=W RS T RS (x,y,z)+W Argo T Argo (x,y,z);

[0050] S fusion (x,y,z)=W RS S RS (x,y,z)+W Argo S Argo (x,y,z);

[0051] Where, T fusion (x,y,z) and S fusion (x, y, z) are the temperature and salinity of the shallow observation point after interpolation and fusion of remote sensing data and Argo float data, T RS (x,y,z) and S RS (x, y, z) are the temperature and salinity calculated by remote sensing at the shallow observation point, T Argo (x,y,z) and S Argo (x, y, z) are the temperature and salinity measured by the Argo float at the shallow observation point, respectively. RS and W Argo are the weights of remote sensing data and Argo float data, respectively, RS +W Argo =1;

[0052] The middle and deep layers use Argo float profile data and submersible temperature-salinity-depth instrument CTD measurement data to generate the middle and deep temperature-salinity fields through weighted averaging and spatial interpolation:

[0053] T f ' usion (x,y,z)=W A ' rgo T A ' rgo (x,y,z)+W sub T sub (x,y,z);

[0054] S' fusion (x,y,z)=W A ' rgo S' Argo (x,y,z)+Wsub S sub (x,y,z);

[0055] Where, T f ' usion (x,y,z) and S' fusion (x, y, z) are the temperature and salinity of the mid-deep observation point after weighted average and spatial interpolation of the Argo float profile data and the submersible temperature-salinity-depth instrument CTD measurement data, respectively. A ' rgo and W sub are the weights of Argo float profile data and submersible temperature-salinity-depth instrument CTD measurement data, T A ' rgo (x,y,z) and S' Argo (x, y, z) are the temperature and salinity measured by the Argo float at the mid-deep observation point, respectively. sub (x,y,z) and S sub (x, y, z) are the temperature and salinity measured by the submersible temperature-salinity-depth instrument CTD at the mid-deep observation point.

[0056] S3 includes a bidirectional interpolation algorithm in the ocean stratification boundary area to make the transition between shallow, middle and deep layers smooth.

[0057] S4 includes analyzing the initial three-dimensional sound velocity field gradient change rate G:

[0058]

[0059] High gradient areas are refined first, increasing the mesh resolution to 1 / 4 of the original size. Refine the mesh to its original size in the medium gradient region The low gradient regions maintain the initial resolution to reduce computational redundancy;

[0060] Analyze data sparsity, dynamically add interpolation nodes in data sparse areas to improve coverage, construct spatial grid division units for temperature and salt fields, and calculate the number of valid observation points in each unit:

[0061]

[0062] Where k is the data point density, N p is the number of valid observation points in the unit, V cell is the volume of the grid cell;

[0063] Nodes are added preferentially to the areas with large gradient changes in the sparse grid, and the temperature and salinity values ​​are calculated using the interpolation method of the surrounding observed data points:

[0064]

[0065] Where T(x,y,z) and S(x,y,z) are the temperature and salinity calculated by interpolation of surrounding observed data points, respectively. ii and S ii are the temperature and salinity of the observed data point ii, d ii is the distance between the observed data point ii and the interpolation node, p is the weight index, p = 2, and N' is the total number of observed data points.

[0066] S5 includes the use of SSH to perform regional adjustments to the inversion results of the sound velocity field:

[0067] c'=c1+k'SSH;

[0068] In the formula, c' is the corrected sound velocity field value, c1 is the initial sound velocity field value, and k' is the empirical coefficient obtained by fitting historical observation data;

[0069] In the current frontal zone of the high SSH variation area, the local refinement correction Δc is calculated g :

[0070]

[0071] Where kk is the enhancement coefficient, which is adjusted by historical data. is an intermediate parameter;

[0072] In high dynamic areas, the sound velocity field can be locally enhanced and corrected. enhanced (x,y,z):

[0073] C enhanced (x,y,z)=C corrected (x,y,z)+Δc g ;

[0074] In the formula, C corrected (x, y, z) is the corrected sound velocity field calculated from the corrected temperature-salinity field.

[0075] S6 includes the use of remote sensing satellites to update SST, SSS, and SSH in real time, and the superposition of data sources with dynamic observations from Argo floats and CTD data from submersibles;

[0076] Use real-time data to dynamically reconstruct the sound velocity field to ensure the accuracy of the sound velocity field changing over time;

[0077] Data is updated first in high-change areas, and historical data is used for interpolation and completion in low-change areas.

[0078] Compared with the prior art, the present invention has the following beneficial effects: the high-precision profile data of the submersible is incorporated into the inversion system, providing a new high-resolution data source for the construction of the sound velocity field in the deep sea area; the adaptive grid encryption and optimization technology based on gradient analysis and data sparsity is used to dynamically encrypt and optimize the high-gradient area, sparse area and local area around the submersible path, thereby improving the spatial resolution and computational efficiency of the sound velocity field, and is suitable for deep-sea acoustic detection, underwater communication and navigation, marine resource exploration and other fields, and has broad application prospects and significant technical advantages. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 It is a technical flow chart of the present invention. DETAILED DESCRIPTION

[0080] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention is described clearly and completely below. Obviously, the described embodiments are part of the embodiments of the present invention, but 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 are within the scope of protection of the present invention.

[0081] The regional three-dimensional sound velocity field inversion and grid encryption processing method based on multi-source ocean data includes:

[0082] S1. Perform regional three-dimensional sound velocity field inversion based on multi-source ocean data, including data acquisition, data preprocessing, multi-source data fusion and sound velocity field inversion;

[0083] S2. Calculate the temperature-salinity field through the regression model. In the area covered by Argo float data, use Argo float data to correct the calculated temperature-salinity field. Use high-precision temperature-salinity-depth instrument data to correct the temperature-salinity field corrected by Argo float data.

[0084] S3. Multi-source data fusion based on hierarchical processing: remote sensing data and Argo float data are interpolated and fused in shallow waters, and Argo float profile data and submersible temperature-salinity-depth instrument CTD measurement data are fused through weighted averaging and spatial interpolation in medium and deep waters;

[0085] S4. Adaptive mesh optimization based on gradient analysis, with high gradient areas being refined first, followed by medium gradient areas, and low gradient areas maintaining the initial resolution;

[0086] S5. Use sea level anomaly data to make regional adjustments to the sound velocity field inversion results;

[0087] S6. The regional three-dimensional sound velocity field is updated in real time.

[0088] The data acquisition of S1 includes remote sensing data, Argo buoy data and submersible data. The remote sensing data includes sea level height anomaly SSH, sea surface temperature SST and sea surface salinity SSS. The Argo buoy data includes the temperature and salinity profile data provided by the Argo buoy. The submersible data includes the temperature, salinity and depth instrument measurement data of manned submersibles and autonomous underwater vehicles AUV.

[0089] Data preprocessing includes denoising, temporal and spatial alignment, outlier removal, and missing data completion for multi-source data.

[0090] S1's multi-source data fusion includes combining remote sensing data, Argo buoy data and submersible data, and using a dynamic weight algorithm to calculate the observation accuracy factor, spatial coverage factor and timeliness factor based on data accuracy, spatial coverage and time consistency, and perform layered weighted fusion to generate a three-dimensional temperature and salinity field.

[0091] The observation precision factor representing the accuracy level of each type of data source is:

[0092]

[0093] Where P i is the observation precision factor of data source i, σ i is the observation error of data source i, σ j is the observation error of the jth data source, and N is the total number of data sources;

[0094] The spatial coverage factor that represents the spatial distribution characteristics of each type of data source is:

[0095]

[0096] In the formula, A i is the coverage factor of data source i, C i is the coverage area of ​​data source i, C j is the coverage area of ​​the jth data source;

[0097] The timeliness factor indicating the time validity of data is:

[0098]

[0099] Where, T i is the timeliness factor of data source i, Δt i is the time difference between data source i and the current time, λ is the time-effectiveness attenuation coefficient, which can be adjusted according to actual needs, Δt j is the time difference between the jth data source and the current time, and e is a natural constant;

[0100] Calculate the final dynamic weight W of each data sourcei :

[0101] W i =αA i +βP i +γT i ;

[0102] In the formula, α, β, and γ are weight adjustment coefficients, which are set according to the importance of each indicator in the application, α+β+γ=1;

[0103] Perform a weighted fusion of the temperature and salinity values ​​for each grid cell:

[0104]

[0105] In the formula, t f is the weighted fusion value of the grid cell temperature, t i is the temperature value of data source i in the grid cell, s f is the weighted fusion value of grid cell salinity, s i is the salinity value of data source i in the grid cell;

[0106] For data-sparse areas at different depths, interpolation methods are used to complete the temperature and salinity values ​​to ensure the integrity of the three-dimensional field. The interpolation methods include inverse distance weighted method and Kriging interpolation method.

[0107] The inversion of the sound velocity field of S1 includes using the Chen-Millero sound velocity formula to calculate the sound velocity field from the temperature-salinity-depth instrument data:

[0108] c(T,S,P)=c0+α1T+β1S+γ1P+δT 2 +∈TS;

[0109] Where c is the speed of sound, which is a function of seawater temperature T, salinity S and pressure P, c0 is the speed of sound constant under the reference state, α1, β1, γ1 are the first-order linear coefficients of temperature, salinity and pressure, δ is the second-order coefficient of temperature, and ∈ is the temperature-salinity cross-term coefficient.

[0110] S2 includes the calculation of temperature and salt field through regression model:

[0111] T c (x,y,z)=g(SST,z);

[0112] S c (x,y,z)=f(SSH,SST);

[0113] Where, T c (x,y,z) and S c(x, y, z) are the temperature and salinity calculated by the regression model, f and g are empirical functions obtained by fitting historical observation data, and (x, y, z) are the longitude, latitude and depth coordinates;

[0114] In the area covered by Argo float data, the estimated temperature and salinity field is corrected using Argo float data:

[0115] T'(x,y,z)=T c (x,y,z)+ΔT Argo ;

[0116] S'(x,y,z)=S c (x,y,z)+ΔS Argo ;

[0117] Where T'(x,y,z) and S'(x,y,z) are the temperature and salinity corrected by Argo float data, ΔT Argo and ΔS Argo are the temperature and salinity corrected by Argo floats, respectively;

[0118] The area with a radius of 10 km between the manned submersible and the AUV path is selected as the local optimization area, and the temperature and salinity field corrected by the Argo buoy data is corrected using the high-precision temperature and salinity depth instrument data:

[0119] T final (x,y,z)=T'(x,y,z)+ΔT sub ;

[0120] S final (x,y,z)=S'(x,y,z)+ΔS sub ;

[0121] Where, T final (x,y,z) and S final (x, y, z) are the final corrected temperature and salinity, ΔT sub and ΔS sub Temperature and salinity corrected using manned submersibles and AUVs, respectively.

[0122] S3 includes dividing the ocean area into shallow, middle and deep layers according to depth, with the shallow layer being between 0 and 200 meters, the middle layer being between 200 and 1000 meters, and the deep layer being above 1000 meters;

[0123] Shallow data include remote sensing-based SST, SSS, and SSH data. The shallow observation data of Argo floats are interpolated and fused. The shallow observation points of Argo floats are aligned to the remote sensing grid. The temperature and salinity values ​​of each grid point are combined with the remote sensing estimated values ​​and Argo measured values ​​and fused through interpolation methods:

[0124] T fusion (x,y,z)=W RS T RS (x,y,z)+W Argo T Argo (x,y,z);

[0125] S fusion (x,y,z)=W RS S RS (x,y,z)+W Argo S Argo (x,y,z);

[0126] Where, T fusion (x,y,z) and S fusion (x, y, z) are the temperature and salinity of the shallow observation point after interpolation and fusion of remote sensing data and Argo float data, T RS (x,y,z) and S RS (x, y, z) are the temperature and salinity calculated by remote sensing at the shallow observation point, T Argo (x,y,z) and S Argo (x, y, z) are the temperature and salinity measured by the Argo float at the shallow observation point, respectively. RS and W Argo are the weights of remote sensing data and Argo float data, respectively, RS +W Argo =1;

[0127] The middle and deep layers use Argo float profile data and submersible temperature-salinity-depth instrument CTD measurement data to generate the middle and deep temperature-salinity fields through weighted averaging and spatial interpolation:

[0128] T f ' usion (x,y,z)=W A ' rgo T A ' rgo (x,y,z)+W sub T sub (x,y,z);

[0129] S' fusion (x,y,z)=W A ' rgo S' Argo (x,y,z)+W sub S sub (x,y,z);

[0130] Where, T f ' usion (x,y,z) and S'fusion (x, y, z) are the temperature and salinity of the mid-deep observation point after weighted average and spatial interpolation of the Argo float profile data and the submersible temperature-salinity-depth instrument CTD measurement data, respectively. A ' rgo and W sub are the weights of Argo float profile data and submersible temperature-salinity-depth instrument CTD measurement data, T A ' rgo (x,y,z) and S' Argo (x, y, z) are the temperature and salinity measured by the Argo float at the mid-deep observation point, respectively. sub (x,y,z) and S sub (x, y, z) are the temperature and salinity measured by the submersible temperature-salinity-depth instrument CTD at the mid-deep observation point.

[0131] S3 includes a bidirectional interpolation algorithm in the ocean stratification boundary area to make the transition between shallow, middle and deep layers smooth.

[0132] S4 includes analyzing the initial three-dimensional sound velocity field gradient change rate G:

[0133]

[0134] High gradient areas are refined first, increasing the mesh resolution to 1 / 4 of the original size. Refine the mesh to its original size in the medium gradient region The low gradient regions maintain the initial resolution to reduce computational redundancy;

[0135] Analyze data sparsity, dynamically add interpolation nodes in data sparse areas to improve coverage, construct spatial grid division units for temperature and salt fields, and calculate the number of valid observation points in each unit:

[0136]

[0137] Where k is the data point density, N p is the number of valid observation points in the unit, V cell is the volume of the grid cell;

[0138] Nodes are added preferentially to the areas with large gradient changes in the sparse grid, and the temperature and salinity values ​​are calculated using the interpolation method of the surrounding observed data points:

[0139]

[0140] Where T(x,y,z) and S(x,y,z) are the temperature and salinity calculated by interpolation of surrounding observed data points, respectively. ii and S iiare the temperature and salinity of the observed data point ii, d ii is the distance between the observed data point ii and the interpolation node, p is the weight index, p = 2, and N' is the total number of observed data points.

[0141] S5 includes the use of SSH to perform regional adjustments to the inversion results of the sound velocity field:

[0142] c'=c1+k'SSH;

[0143] In the formula, c' is the corrected sound velocity field value, c1 is the initial sound velocity field value, and k' is the empirical coefficient obtained by fitting historical observation data;

[0144] In the current frontal zone of the high SSH variation area, the local refinement correction Δc is calculated g :

[0145]

[0146]

[0147] Where kk is the enhancement coefficient, which is adjusted by historical data. is an intermediate parameter;

[0148] In high dynamic areas, the sound velocity field can be locally enhanced and corrected. enhanced (x,y,z):

[0149] C enhanced (x,y,z)=C corrected (x,y,z)+Δc g ;

[0150] In the formula, C corrected (x, y, z) is the corrected sound velocity field calculated from the corrected temperature-salinity field.

[0151] S6 includes the use of remote sensing satellites to update SST, SSS, and SSH in real time, and the superposition of data sources with dynamic observations from Argo floats and CTD data from submersibles;

[0152] Use real-time data to dynamically reconstruct the sound velocity field to ensure the accuracy of the sound velocity field changing over time;

[0153] Data is updated first in high-change areas, and historical data is used for interpolation and completion in low-change areas.

[0154] The technical flow chart of the present invention is as follows: Figure 1As shown in the figure, remote sensing data features wide spatial coverage and high timeliness, which are suitable for shallow large-scale fusion; Argo float data covers a depth range from the sea surface to about 2000m, with dense horizontal distribution and stable and reliable data, which are suitable for the construction of temperature and salinity fields in the middle to deep layers; submersible data has the highest accuracy and is suitable for local high-resolution areas; it supplements local high-precision temperature and salinity profile data and provides high-precision temperature, salinity and depth data in local areas, which are mainly used for correction and verification.

[0155] The larger the spatial distribution coverage of each type of data source, the higher the weight of the spatial coverage factor; the higher the accuracy of each type of data source, the higher the weight of the observation accuracy factor; the closer the data time is to the current time, the higher the weight of the timeliness factor.

[0156] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some or all of the technical features may be replaced by equivalents, and 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 method for regional three-dimensional sound velocity field inversion and grid encryption based on multi-source ocean data, characterized in that: include: S1. Perform regional three-dimensional sound velocity field inversion based on multi-source ocean data, including data acquisition, data preprocessing, multi-source data fusion and sound velocity field inversion; S2. Calculate the temperature-salinity field through the regression model. In the area covered by Argo float data, use Argo float data to correct the calculated temperature-salinity field. Use high-precision temperature-salinity-depth instrument data to correct the temperature-salinity field corrected by Argo float data. S3. Multi-source data fusion based on hierarchical processing: remote sensing data and Argo float data are interpolated and fused in shallow waters, and Argo float profile data and submersible temperature-salinity-depth instrument CTD measurement data are fused through weighted averaging and spatial interpolation in medium and deep waters; S4. Adaptive mesh optimization based on gradient analysis, with high gradient areas being refined first, followed by medium gradient areas, and low gradient areas maintaining the initial resolution; S5. Use sea level anomaly data to make regional adjustments to the sound velocity field inversion results; S6. The regional three-dimensional sound velocity field is updated in real time.

2. The method for regional three-dimensional sound velocity field inversion and grid encryption processing based on multi-source ocean data according to claim 1 is characterized in that: The data acquisition of S1 includes remote sensing data, Argo buoy data and submersible data. The remote sensing data includes sea level height anomaly SSH, sea surface temperature SST and sea surface salinity SSS. The Argo buoy data includes the temperature and salinity profile data provided by the Argo buoy. The submersible data includes the temperature, salinity and depth instrument measurement data of manned submersibles and autonomous underwater vehicles AUV. Data preprocessing includes denoising, temporal and spatial alignment, outlier removal, and missing data completion for multi-source data.

3. The method for regional three-dimensional sound velocity field inversion and grid encryption processing based on multi-source ocean data according to claim 2 is characterized in that: S1's multi-source data fusion includes combining remote sensing data, Argo buoy data and submersible data, and using a dynamic weight algorithm to calculate the observation accuracy factor, spatial coverage factor and timeliness factor based on data accuracy, spatial coverage and time consistency, and perform layered weighted fusion to generate a three-dimensional temperature and salinity field. The observation precision factor representing the accuracy level of each type of data source is: Where P i is the observation precision factor of data source i, σ i is the observation error of data source i, σ j is the observation error of the jth data source, and N is the total number of data sources; The spatial coverage factor that represents the spatial distribution characteristics of each type of data source is: In the formula, A i is the coverage factor of data source i, C i is the coverage area of ​​data source i, C j is the coverage area of ​​the jth data source; The timeliness factor indicating the time validity of data is: Where, T i is the timeliness factor of data source i, Δt i is the time difference between data source i and the current time, λ is the time-effectiveness attenuation coefficient, which can be adjusted according to actual needs, Δt j is the time difference between the jth data source and the current time, and e is a natural constant; Calculate the final dynamic weight W of each data source i : W i =αA i +βP i +γT i ; In the formula, α, β, and γ are weight adjustment coefficients, which are set according to the importance of each indicator in the application, α+β+γ=1; Perform a weighted fusion of the temperature and salinity values ​​for each grid cell: Where, t f is the weighted fusion value of the grid cell temperature, t i is the temperature value of data source i in the grid cell, s f is the weighted fusion value of grid cell salinity, s i is the salinity value of data source i in the grid cell; For data-sparse areas at different depths, interpolation methods are used to complete the temperature and salinity values ​​to ensure the integrity of the three-dimensional field. The interpolation methods include inverse distance weighted method and Kriging interpolation method.

4. The method for regional three-dimensional sound velocity field inversion and grid encryption processing based on multi-source ocean data according to claim 3 is characterized in that: The inversion of the sound velocity field of S1 includes using the Chen-Millero sound velocity formula to calculate the sound velocity field from the temperature-salinity-depth instrument data: c(T,S,P)=c0+α1T+β1S+γ1P+δT 2 +∈TS; Where c is the speed of sound, which is a function of seawater temperature T, salinity S and pressure P, c0 is the speed of sound constant under the reference state, α1, β1, γ1 are the first-order linear coefficients of temperature, salinity and pressure, δ is the second-order coefficient of temperature, and ∈ is the temperature-salinity cross-term coefficient.

5. The method for regional three-dimensional sound velocity field inversion and grid encryption processing based on multi-source ocean data according to claim 4 is characterized in that: S2 includes the calculation of temperature and salt field through regression model: T c (x,y,z)=g(SST,z); S c (x,y,z)=f(SSH,SST); Where, T c (x,y,z) and S c (x, y, z) are the temperature and salinity calculated by the regression model, f and g are empirical functions obtained by fitting historical observation data, and (x, y, z) are the longitude, latitude and depth coordinates; In the area covered by Argo float data, the estimated temperature and salinity field is corrected using Argo float data: T'(x,y,z)=T c (x,y,z)+ΔT Argo ; S'(x,y,z)=S c (x,y,z)+ΔS Argo ; Where T'(x,y,z) and S'(x,y,z) are the temperature and salinity corrected by Argo float data, ΔT Argo and ΔS Argo are the temperature and salinity corrected by Argo floats, respectively; The area with a radius of 10 km between the manned submersible and the AUV path is selected as the local optimization area, and the temperature and salinity field corrected by the Argo buoy data is corrected using the high-precision temperature and salinity depth instrument data: T final (x,y,z)=T'(x,y,z)+ΔT sub ; S final (x,y,z)=S'(x,y,z)+ΔS sub ; Where, T final (x,y,z) and S final (x, y, z) are the final corrected temperature and salinity, ΔT sub and ΔS sub Temperature and salinity corrected using manned submersibles and AUVs, respectively.

6. The method for regional three-dimensional sound velocity field inversion and grid encryption processing based on multi-source ocean data according to claim 5 is characterized in that: S3 includes dividing the ocean area into shallow, middle and deep layers according to depth, with the shallow layer being between 0 and 200 meters, the middle layer being between 200 and 1000 meters, and the deep layer being above 1000 meters; Shallow data include remote sensing-based SST, SSS, and SSH data. The shallow observation data of Argo floats are interpolated and fused. The shallow observation points of Argo floats are aligned to the remote sensing grid. The temperature and salinity values ​​of each grid point are combined with the remote sensing estimated values ​​and Argo measured values ​​and fused through interpolation methods: T fusion (x,y,z)=W RS T RS (x,y,z)+W Argo T Argo (x,y,z); S fusion (x,y,z)=W RS S RS (x,y,z)+W Argo S Argo (x,y,z); Where, T fusion (x,y,z) and S fusion (x, y, z) are the temperature and salinity of the shallow observation point after interpolation and fusion of remote sensing data and Argo float data, T RS (x,y,z) and S RS (x, y, z) are the temperature and salinity calculated by remote sensing at the shallow observation point, T Argo (x,y,z) and S Argo (x, y, z) are the temperature and salinity measured by the Argo float at the shallow observation point, respectively. RS and W Argo are the weights of remote sensing data and Argo float data, respectively, RS +W Argo =1; The middle and deep layers use Argo float profile data and submersible temperature-salinity-depth instrument CTD measurement data to generate the middle and deep temperature-salinity fields through weighted averaging and spatial interpolation: T f ' usion (x,y,z)=W A ' rgo T A ' rgo (x,y,z)+W sub T sub (x,y,z); S' fusion (x,y,z)=W A ' rgo S' Argo (x,y,z)+W sub S sub (x,y,z); Where, T f ' usion (x,y,z) and S' fusion (x, y, z) are the temperature and salinity of the mid-deep observation point after weighted average and spatial interpolation of the Argo float profile data and the submersible temperature-salinity-depth instrument CTD measurement data, respectively. A ' rgo and W sub are the weights of Argo float profile data and submersible temperature-salinity-depth instrument CTD measurement data, T A ' rgo (x,y,z) and S' Argo (x, y, z) are the temperature and salinity measured by the Argo float at the mid-deep observation point, respectively. sub (x,y,z) and S sub (x, y, z) are the temperature and salinity measured by the submersible temperature-salinity-depth instrument CTD at the mid-deep observation point.

7. The method for regional three-dimensional sound velocity field inversion and grid encryption processing based on multi-source ocean data according to claim 6 is characterized in that: S3 includes a bidirectional interpolation algorithm in the ocean stratification boundary area to make the transition between shallow, middle and deep layers smooth.

8. The method for regional three-dimensional sound velocity field inversion and grid encryption processing based on multi-source ocean data according to claim 7 is characterized in that: S4 includes analyzing the initial three-dimensional sound velocity field gradient change rate G: High gradient areas are refined first, increasing the mesh resolution to 1 / 4 of the original size. Refine the mesh to its original size in the medium gradient region The low gradient regions maintain the initial resolution to reduce computational redundancy; Analyze data sparsity, dynamically add interpolation nodes in data sparse areas to improve coverage, construct spatial grid division units for temperature and salt fields, and calculate the number of valid observation points in each unit: Where k is the data point density, N p is the number of valid observation points in the unit, V cell is the volume of the grid cell; Nodes are added preferentially to the areas with large gradient changes in the sparse grid, and the temperature and salinity values ​​are calculated using the interpolation method of the surrounding observed data points: Where T(x,y,z) and S(x,y,z) are the temperature and salinity calculated by interpolation of surrounding observed data points, respectively. ii and S ii are the temperature and salinity of the observed data point ii, d ii is the distance between the observed data point ii and the interpolation node, p is the weight index, p = 2, and N' is the total number of observed data points.

9. The method for regional three-dimensional sound velocity field inversion and grid encryption processing based on multi-source ocean data according to claim 8 is characterized in that: S5 includes the use of SSH to perform regional adjustments to the inversion results of the sound velocity field: c'=c1+k'SSH; In the formula, c' is the corrected sound velocity field value, c1 is the initial sound velocity field value, and k' is the empirical coefficient obtained by fitting historical observation data; In the current frontal zone of the high SSH variation area, the local refinement correction Δc is calculated g : Δc g =kk▽SSH; Where kk is the enhancement coefficient, which is adjusted by historical data, and ▽SSH is the intermediate parameter; In high dynamic areas, the sound velocity field can be locally enhanced and corrected. enhanced (x,y,z): C enhanced (x,y,z)=C corrected (x,y,z)+Δc g ; In the formula, C corrected (x, y, z) is the corrected sound velocity field calculated from the corrected temperature-salinity field.

10. The method for regional three-dimensional sound velocity field inversion and grid encryption processing based on multi-source ocean data according to claim 9, characterized in that: S6 includes the use of remote sensing satellites to update SST, SSS, and SSH in real time, and the superposition of data sources with dynamic observations from Argo floats and CTD data from submersibles; Use real-time data to dynamically reconstruct the sound velocity field to ensure the accuracy of the sound velocity field changing over time; Data is updated first in high-change areas, and historical data is used for interpolation and completion in low-change areas.

Images