Sea surface temperature fusion method and system based on ocean area

Abstract

The invention provides a sea surface temperature fusion method and system based on a South China Sea region, and the method comprises the following steps: S1, obtaining multi-source sea surface temperature data in a designated region and a time range, the multi-source sea surface temperature data comprising an infrared sea surface temperature, a microwave sea surface temperature and an actually measured sea surface temperature; s2, performing quality control on the acquired multi-source sea surface temperature data; s3, performing deviation correction between sensors on the multi-source sea surface temperature data after quality control to obtain the multi-source sea surface temperature data after deviation correction; and S4, for the multi-source sea surface temperature after quality control and deviation correction, a regularization variational method is adopted, and a wavelet coefficient is calculated by fitting a wavelet basis, so that a multi-scale variational fusion model is constructed, and a sea surface temperature fusion product at a target moment is obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of marine information processing, and particularly to a sea surface temperature fusion method and system based on a marine area. Background Art

[0002] Sea Surface Temperature (SST) is one of the extremely important physical parameters in the ocean and is also the core variable that has been the earliest concerned and widely observed in oceanographic research. Its spatio-temporal variations not only have a profound impact on the weather and climate of the local and surrounding areas through air-sea interactions, but also determine the dynamic characteristics of the local ocean environment, thus playing an important regulatory role in the marine ecosystem and biological resources.

[0003] The ocean surface is vulnerable to the influence of the monsoon system, and its circulation structure is complex and variable, and mesoscale ocean phenomena (such as vortices, fronts, upwelling, etc.) are likely to occur. The spatio-temporal scale characteristics of these dynamic processes are complex and diverse, posing a huge challenge to the management of the marine ecological environment and fishery resources. Therefore, to deeply understand the ocean processes in a marine area, a set of sea surface temperature data sets with high spatio-temporal resolution and high precision is required as the basis to support scientific research and applications.

[0004] However, the existing single remote sensing sensor data or in-situ observation data have deficiencies in terms of coverage, spatial resolution, or temporal resolution, and it is difficult to meet the needs of studying multi-scale ocean processes in a marine area. Therefore, the present invention proposes a sea surface temperature fusion method and system based on a marine area, which integrates multi-source sea surface temperature data to make up for the limitations of a single data source, provides technical support for the construction of a high-precision sea surface temperature field in a marine area, and contributes to the development of fields such as climate research, marine environmental monitoring, and resource management. Summary of the Invention

[0005] The purpose of the present invention is to provide a digital movie transmission device to solve the problems raised in the above background art.

[0006] The present invention is realized through the following technical solutions:

[0007] On the one hand, the present invention provides a sea surface temperature fusion method based on a marine area, and the method includes the following steps:

[0008] Step S1, obtain multi-source sea surface temperature data within a specified area and time range, and the multi-source sea surface temperature data includes: infrared sea surface temperature, microwave sea surface temperature, and measured sea surface temperature;

[0009] Step S2, perform quality control on the obtained multi-source sea surface temperature data;

[0010] Step S3: Perform inter-sensor bias correction on the multi-source sea surface temperature data after quality control to obtain the multi-source sea surface temperature data after bias correction;

[0011] Step S4: For the multi-source sea surface temperature after quality control and bias correction, use the regularized variational method to calculate wavelet coefficients by fitting wavelet bases, thereby constructing a multi-scale variational fusion model to obtain the sea surface temperature fusion product at the target time.

[0012] Specifically, in the step S2, the quality control of the multi-source sea surface temperature data specifically includes: pixel value inspection and spatial uniformity inspection.

[0013] Specifically, in the step S3, the calculation method of the inter-sensor bias correction includes: subtracting the bias from each SST sample value of each sensor.

[0014] Specifically, the step S4 specifically includes first determining the resolution and performing wavelet grid division, as follows:

[0015] i = (x max - x min ) / Δ n

[0016] j = (y max - y min ) / Δ n

[0017] where i and j both represent wavelet grid indices determined by the resolution, x max represents the maximum longitude of the fusion range, x min represents the minimum longitude of the fusion range, y max represents the maximum latitude of the fusion range, y min represents the minimum latitude of the fusion range, Δ n represents the reference resolution length related to the scale index, and n represents the given scale index.

[0018] Specifically, the step S4 also specifically includes using wavelet bases for data expansion and reconstruction, as follows:.

[0019] In the two-dimensional space, the sea surface temperature T N (x, y) can be decomposed by a linear combination of two-dimensional wavelet functions and scaling functions:

[0020]

[0021] where x represents longitude, y represents latitude, represents the approximate wavelet coefficient related to the wavelet, φ(x, y) represents the scaling function, The detail wavelet coefficients related to wavelets are denoted as such, and ψ(x, y) represents the wavelet function;

[0022] Since the two-dimensional basis function is extended from the one-dimensional wavelet basis, the one-dimensional wavelet basis adopts the third-order B-spline function, and the formula is as follows:

[0023]

[0024] Among them, β(t) represents the third-order B-spline function, and t represents the independent variable of the spline basis function;

[0025] In one-dimensional space, the one-dimensional wavelet function ψ(t) and the one-dimensional scaling function φ(t) are constructed through the one-dimensional wavelet basis. The specific formulas for the constructed wavelet function and scaling function are as follows:

[0026]

[0027]

[0028] Among them, f n represents the high-frequency filtering coefficients related to wavelets; h n represents the low-frequency filtering coefficients related to wavelets;

[0029] Extended from one-dimensional space to two-dimensional space through the tensor product, the constructed two-dimensional scaling function and two-dimensional wavelet function are respectively:

[0030] φ(x, y) = φ ni (x)φ nj (y)

[0031] ψ( H )(x, y) = φ ni (x)ψ nj (y)

[0032] ψ( V )(x, y) = ψ ni (x)φ nj (y)ψ( D )(x, y) = ψ ni (x)ψ nj (y)

[0033] Among them, φ ni (x) represents the low-frequency information in the x direction, φ nj (y) represents the low-frequency information in the y direction, ψ ni (x) represents the high-frequency information in the x direction, ψ nj (y) represents the high-frequency information in the y direction, φ(x, y) represents the two-dimensional scaling function, ψ( H )(x, y) represents the two-dimensional wavelet function in the horizontal direction, ψ( V)(x,y) represents the wavelet function in the two-dimensional vertical direction, ψ( D )(x,y) represents the wavelet coefficient in the two-dimensional diagonal direction;

[0034] Furthermore, according to the multi-scale wavelet decomposition, the sea surface temperature expansion component T at different scales n (x,y) is expressed as:

[0035]

[0036] wherein, represents the B-spline wavelet coefficient of a single sensor at the nth layer, Δ n represents the reference resolution length related to the scale index.

[0037] Specifically, in step S4, the regularized variational method is used to solve the optimal wavelet coefficients, as follows:

[0038]

[0039] wherein, Ω represents the spatial range of the ocean area, represents the regularization term; ε(T) represents the data fitting residual.

[0040] Specifically, in step S4, according to the wavelet expansion components obtained at different scales, the multi-scale variational fusion model of each sensor is further obtained through scale superposition:

[0041]

[0042] wherein, represents the multi-scale variational fusion model, and N represents the total number of layers of signal decomposition.

[0043] On the other hand, the present invention provides a sea surface temperature fusion system based on an ocean area, which is applied to the method described above. The system includes:

[0044] A first processing module for obtaining multi-source sea surface temperature data within a specified area and time range;

[0045] A second processing module for performing quality control on the obtained multi-source sea surface temperature data;

[0046] A third processing module for performing inter-sensor bias correction on the multi-source sea surface temperature data after quality control;

[0047] A fourth processing module for using the variational method on the multi-source sea surface temperature data after quality control and bias correction, calculating wavelet coefficients by fitting wavelet bases, and thus constructing a multi-scale variational analysis fusion model to obtain a sea surface temperature fusion product at the target time.

[0048] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0049] A sea surface temperature fusion method and system based on a marine area provided by the present invention can obtain a sea surface temperature fusion product of a marine area with high spatio-temporal resolution, high spatial coverage, and high precision by constructing a multi-scale variational fusion model. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only the preferred embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0051] Figure 1 It is a flowchart of the steps of a sea surface temperature fusion method based on a marine area provided by the present invention;

[0052] Figure 2 It is a schematic diagram of wavelet grid division provided by the present invention;

[0053] Figure 3 It is a schematic diagram of a third-order B-spline function provided by the present invention;

[0054] Figure 4 It is a schematic diagram of the structure of a sea surface temperature fusion system based on a marine area provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] In order to make the objectives, technical solutions, and advantages of the present invention more obvious, the exemplary embodiments according to the present invention will be described in detail with reference to the drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited by the exemplary embodiments described herein. Based on the embodiments of the present invention described herein, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0056] In the following description, a large number of specific details are given to provide a more thorough understanding of the present invention. However, it is obvious to those of ordinary skill in the art that the present invention can be implemented without one or more of these details. In other examples, in order to avoid confusion with the present invention, some well-known technical features are not described.

[0057] It should be understood that the present invention can be implemented in different forms and should not be construed as limited to the embodiments presented herein. On the contrary, providing these embodiments will make the disclosure thorough and complete, and will fully convey the scope of the present invention to those skilled in the art.

[0058] The purpose of the terms used herein is only to describe specific embodiments and is not a limitation of the present invention. As used herein, the singular forms "a", "an" and "the" are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the terms "comprising" and / or "including", when used in this specification, determine the presence of the stated features, integers, steps, operations, elements and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups. As used herein, the term "and / or" includes any and all combinations of the related listed items.

[0059] To thoroughly understand the present invention, detailed structures will be presented in the following description to illustrate the technical solutions proposed by the present invention. The optional embodiments of the present invention are described in detail as follows. However, in addition to these detailed descriptions, the present invention may also have other implementation manners.

[0060] See Figure 1 , on the one hand, the present invention provides a sea surface temperature fusion method based on a marine area, and the method includes the following steps:

[0061] Step S1, obtain multi-source sea surface temperature data within a specified area and time range, and the multi-source sea surface temperature data includes: infrared sea surface temperature, microwave sea surface temperature, measured sea surface temperature;

[0062] Step S2, perform quality control on the obtained multi-source sea surface temperature data;

[0063] Step S3, perform inter-sensor bias correction on the multi-source sea surface temperature data after quality control to obtain the multi-source sea surface temperature data after bias correction;

[0064] Step S4, for the multi-source sea surface temperature after quality control and bias correction, adopt the regularization variational method, calculate wavelet coefficients by fitting wavelet bases, thereby construct a multi-scale variational fusion model, and obtain a sea surface temperature fusion product at the target moment.

[0065] Exemplarily, the multi-source sea surface temperature within the specified region and time range obtained in step S1 is the sea surface temperature data of the ocean region at the target moment. The target moment is 9:00 UTC of any day in the current or historical time. The spatial range is from 110°E to 120°E and from 12°N to 20°N; the time range is the day of the target time; the infrared sea surface temperature product is the MODIS TERRA\AQUA GHRSST L2P product; the microwave sea surface temperature product is the AMSR-2 GHRSST L2P product; the measured sea surface temperature is the iQuam buoy product;

[0066] Among them, MODIS is a medium-resolution imaging spectrometer carried on the AQUA\TERRA satellite. TERRA is the morning satellite with a transit time of around 10:30 am, and AQUA is the afternoon satellite with a transit time of around 1:30 pm. It captures data in 36 spectral bands at various spatial resolutions. For the SST product, it is generated using an improved non-linear SST algorithm with the 4, 11, and 12-micron infrared bands. The MODIS GHRSST L2P retrieval data is completed by the cooperation of RSMAS, NASA, OBPG, and NASA JPL. The data includes quality flags, uncertainty estimates, metadata, and auxiliary variable information.

[0067] In addition, AMSR-2 is an advanced microwave scanning radiometer carried on the Global Change Observation Mission - Water (GCOM-W) satellite developed by JAXA. Its transit time is 1:30 pm, and the antenna rotates once every 1.5 seconds, capable of obtaining data with a swath width of 1450 km. It can obtain data covering more than 99% of the Earth's day and night every two days. The Advanced Microwave Scanning Radiometer 2 (AMSR-2) of REMSS provides two types of SST products for GHRSST, including the "near real-time" product, which is usually provided within 3 hours after measurement reception; the "final" product, which is usually released within 2 days and contains a more accurate retrieval based on atmospheric analysis data, including the wind direction from the final operational global analysis of the National Centers for Environmental Prediction (NCEP).

[0068] Therefore, the present invention selects the MODIS GHRSST L2P product with a spatial resolution of 1 km for fusion, and selects the along-track AMSR-2 L2P GHRSST "final" product with a spatial resolution of 25 km for fusion.

[0069] Moreover, the iQuam data is developed by NOAA, and this dataset includes: Argo floats, traditional drifters, high-resolution drifters, tropical moorings, coastal moorings, coral reef observation buoys, traditional ships, and ships of the International Maritime Organization. The iQuam data is stored in the NetCDF4 format, with one data file stored monthly and updated every 24 hours. The data file is divided into two parts: global attributes and data variables. In the present invention, data of observation types such as ships, drifter buoys, tropical mooring buoys, coastal mooring buoys, and high-resolution drifters are selected for accuracy verification of the fusion product.

[0070] Specifically, in step S2, the quality control of multi-source sea surface temperature data specifically includes: pixel value inspection and spatial uniformity inspection.

[0071] Exemplarily,

[0072] (1) Pixel value inspection

[0073] For satellite SST retrieval datasets using the GHRSST convention, each SST sample is associated with a quality flag whose value ranges from 0 to 5. The GHRSST quality flag generally evaluates the level of environmental conditions, such as cloud cover known to interfere with satellite SST retrieval. The quality control information provides the quality level of each pixel of the SST data, divided into 6 levels from 0 to 5, with 5 representing the highest quality level. The present invention needs to select samples marked as the highest quality.

[0074] (2) Spatial uniformity inspection

[0075] Spatially, the change in sea surface temperature is relatively slow. To eliminate abnormal sea surface temperature pixels generated by data noise in spatial changes and ensure the spatial uniformity of sea surface temperature data, the present invention eliminates abnormal sea surface temperature data based on "three times the standard deviation". During the "three times the standard deviation" abnormal data inspection process, for each SST sample, its mean value μ and standard deviation σ need to be calculated, and the current sea surface temperature within the range of (μ - 3σ, μ + 3σ) is selected. Data points outside this range are considered outliers and are eliminated.

[0076] Specifically, in step S3, the calculation method for bias correction between sensors includes: subtracting the bias from each SST sample value of each sensor.

[0077] Exemplarily, before fusion, bias correction needs to be performed on multi-source sea surface temperature data to eliminate systematic biases between different data sources. For the GHRSST dataset, this dataset contains sensor error statistics, namely bias and standard deviation. Therefore, before inputting into the multi-scale variational fusion model, each SST sample needs to subtract the bias to obtain the bias-corrected GHRSST data.

[0078] Exemplarily, for regularly sampled data, after wavelet decomposition, the signal can be divided into an average field and a series of variant fields. By solving the wavelet coefficients and then using the wavelet coefficients and two-dimensional basis functions for data reconstruction, the final fused data is obtained. However, the sea surface temperature data in ocean regions is often irregular. Therefore, the present invention introduces regularization variational method and numerical approximation for integration.

[0079] The variational method is numerically implemented based on the Gauss-Markov theorem and through the finite element implementation using wavelet bases. The numerical implementation is achieved by combining the finite element method with wavelet basis functions. The above method not only shows stronger robustness to noise data, but also can transform complex mathematical problems into discrete numerical calculations, which is convenient for implementation.

[0080] Specifically, step S4 specifically includes first determining the resolution and performing wavelet grid division, as follows:

[0081] i = (x max - x min ) / Δ n

[0082] j = (y max - y min ) / Δ n

[0083] where both i and j represent the wavelet grid indices determined by the resolution, x max represents the maximum longitude of the fusion range, x min represents the minimum longitude of the fusion range, y max represents the maximum latitude of the fusion range, y min represents the minimum latitude of the fusion range, Δ n represents the reference resolution length related to the scale index, and n represents the given scale index.

[0084] where, Δ n = 2 -n Δ0, which means that for the given scale index n and baseline scale length Δ0, in the analysis, when the baseline scale length Δ0 = 0.2°, the wavelet grid size divided is as Figure 2 shown.

[0085] Specifically, step S4 also specifically includes using wavelet bases for data expansion and reconstruction, as follows:

[0086] In two-dimensional space, the sea surface temperature T N (x, y) can be decomposed by the linear combination of two-dimensional wavelet functions and scaling functions:

[0087]

[0088] Among them, x represents longitude and y represents latitude. represents the approximate wavelet coefficient related to the wavelet, and φ(x, y) represents the scaling function. represents the detail wavelet coefficient related to the wavelet, and ψ(x, y) represents the wavelet function.

[0089] The two-dimensional basis function is extended from the one-dimensional wavelet basis. Therefore, the one-dimensional wavelet basis adopts the third-order B-spline function, and the formula is as follows:

[0090]

[0091] Among them, β(t) represents the third-order B-spline function, t represents the independent variable of the spline basis function, and the interval is [-2, 2].

[0092] In one-dimensional space, the one-dimensional wavelet function ψ(t) and the one-dimensional scaling function φ(t) are constructed through the one-dimensional wavelet basis. The specific formulas of the constructed wavelet function and scaling function are as follows:

[0093]

[0094] Among them, f n represents the high-frequency filtering coefficient related to the wavelet; h n represents the low-frequency filtering coefficient related to the wavelet.

[0095] Extended from one-dimensional space to two-dimensional space through tensor product, the constructed two-dimensional scaling function and two-dimensional wavelet function are respectively:

[0096] φ(x, y) = φ ni (x)φ nj (y)

[0097] ψ( H )(x, y) = φ ni (x)ψ nj (y)

[0098] ψ( V )(x, y) = ψ ni (x)φ nj (y)ψ( D )(x, y) = ψ ni (x)ψ nj (y)

[0099] Among them, φ ni (x) represents the low-frequency information in the x direction, φ nj (y) represents the low-frequency information in the y direction, ψ ni (x) represents the high-frequency information in the x direction, ψ nj(y) represents the high-frequency information in the y direction, φ(x, y) represents the two-dimensional scaling function, ψ( H )(x, y) represents the two-dimensional wavelet function in the horizontal direction, ψ( V )(x, y) represents the two-dimensional wavelet function in the vertical direction, ψ( D )(x, y) represents the two-dimensional wavelet coefficient in the diagonal direction;

[0100] Then, according to the multi-scale wavelet decomposition, the sea surface temperature expansion component T n (x, y) is expressed as:

[0101]

[0102] Among them, represents the B-spline wavelet coefficient of a single sensor at the nth layer, Δ n represents the reference resolution length related to the scale index.

[0103] Specifically, in step S4, the regularization variational method is used to solve the optimal wavelet coefficients, as follows:

[0104]

[0105] Among them, Ω represents the spatial range of the ocean area, represents the regularization term; ε(T) represents the data fitting residual.

[0106]

[0107] Among them, ω n is the data weight; is the reconstructed sea surface temperature data set; T n is the input sea surface temperature data, x n represents the longitude of the input sea surface temperature data, y n represents the latitude of the input sea surface temperature data;

[0108]

[0109] Among them, w1, w2, w3, w4, and w5 are all regularization weights; T is the reconstructed sea surface temperature; x is the longitude of each point; y is the latitude of each point.

[0110] Then, the minimization problem is converted into a problem of solving a sparse matrix linear equation system.

[0111] Specifically, the coefficients of each satellite data at each scale can be solved according to the sparse equation matrix. The calculation formula is as follows:

[0112]

[0113] Among them, E pq is the contribution degree of each sea surface temperature on the corresponding grid; D pqij is the coefficient matrix in the linear system. p and q represent the wavelet grid positions of each point in the input sea surface temperature data; i and j represent the control point positions of the basis functions on each wavelet grid.

[0114] Specifically, in step S4, according to the obtained wavelet expansion components at different scales, a multi-scale variational fusion model of each sensor is further obtained through scale superposition:

[0115]

[0116] Among them, represents the multi-scale variational fusion model, and N represents the total number of layers of signal decomposition.

[0117] By inputting the processed data into the multi-scale variational fusion model, a sea surface temperature fusion product at the target time is obtained.

[0118] On the other hand, the present invention provides a sea surface temperature fusion system based on an ocean area. This system applies the above method. See Figure 4 and the system includes:

[0119] A first processing module for obtaining multi-source sea surface temperature data within a specified area and time range;

[0120] A second processing module for performing quality control on the obtained multi-source sea surface temperature data;

[0121] A third processing module for performing inter-sensor bias correction on the multi-source sea surface temperature data after quality control;

[0122] A fourth processing module for using the variational method to calculate wavelet coefficients by fitting wavelet bases for the multi-source sea surface temperature data after quality control and bias correction, thereby constructing a multi-scale variational analysis fusion model to obtain a sea surface temperature fusion product at the target time.

[0123] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A sea surface temperature fusion method based on ocean areas, characterized in that: The method comprises the following steps: Step S1, obtaining multi-source sea surface temperature data within a specified area and time range, wherein the multi-source sea surface temperature data includes: infrared sea surface temperature, microwave sea surface temperature, and measured sea surface temperature; Step S2, performing quality control on the acquired multi-source sea surface temperature data; Step S3, performing inter-sensor deviation correction on the multi-source sea surface temperature data after quality control to obtain the multi-source sea surface temperature data after deviation correction; Step S4: For the multi-source sea surface temperature after quality control and bias correction, a regularized variational method is used to calculate the wavelet coefficients by fitting the wavelet basis, thereby constructing a multi-scale variational fusion model to obtain the sea surface temperature fusion product at the target time.

2. The method for fusion of sea surface temperature based on ocean area according to claim 1, characterized in that: In step S2, quality control of multi-source sea surface temperature data specifically includes: pixel value inspection and spatial uniformity inspection.

3. The method for fusion of sea surface temperature based on ocean area according to claim 2, characterized in that: In step S3, the calculation method of the deviation correction between sensors includes: subtracting the deviation from each SST sample value of each sensor.

4. The method for fusion of sea surface temperature based on ocean area according to claim 3 is characterized in that: The step S4 specifically includes first determining the resolution and performing wavelet grid division, as follows: i=(x max -x min ) / Δ n j=(y max -y min ) / Δ n Where i and j represent the wavelet grid index determined by the resolution, and x max Indicates the maximum longitude of the fusion range, x min Indicates the minimum longitude of the fusion range, y max Indicates the maximum latitude of the fusion range, y min Indicates the minimum latitude of the fusion range, Δ n represents the reference resolution length associated with the scale index, and n represents the given scale index.

5. The method for fusion of sea surface temperature based on ocean area according to claim 4 is characterized in that: The step S4 also specifically includes using a wavelet basis to perform data expansion and reconstruction, as follows: In two-dimensional space, the sea surface temperature T N (x,y) can be decomposed by a linear combination of a two-dimensional wavelet function and a scaling function: Among them, x represents longitude, y represents latitude, represents the approximate wavelet coefficients associated with the wavelet, φ(x,y) represents the scaling function, represents the detail wavelet coefficients associated with the wavelet, ψ(x,y) represents the wavelet function; The two-dimensional basis function is expanded by the one-dimensional wavelet basis. Therefore, the one-dimensional wavelet basis adopts the third-order B-spline function, and the formula is as follows: Among them, β(t) represents the third-order B-spline function, and t represents the independent variable of the spline basis function; In one-dimensional space, a one-dimensional wavelet function ψ(t) and a one-dimensional scaling function φ(t) are constructed through a one-dimensional wavelet basis. The specific formulas of the constructed wavelet function and scaling function are as follows: Among them, f n represents the high-frequency filter coefficient associated with the wavelet; h n represents the low-frequency filter coefficients associated with the wavelet; The two-dimensional scaling function and two-dimensional wavelet function constructed by extending the one-dimensional space to the two-dimensional space through the tensor product are: φ(x,y)=φ ni (x)φ nj (y) ψ (H) (x,y)=φ ni (x)ψ nj (y) ψ (V) (x,y)=ψ ni (x)φ nj (y)ψ (D) (x,y)=ψ ni (x)ψ nj (y) Among them, φ ni (x) represents the low-frequency information in the x direction, φ nj (y) represents the low-frequency information in the y direction, ψ ni (x) represents the high-frequency information in the x direction, ψ nj (y) represents the high-frequency information in the y direction, φ(x,y) represents the two-dimensional scaling function, and ψ (H) (x,y) represents the wavelet function in the two-dimensional horizontal direction, ψ (V) (x,y) represents the wavelet function in the two-dimensional vertical direction, ψ (D) (x,y) represents the wavelet coefficients in the two-dimensional diagonal direction; Then, according to the multi-scale wavelet decomposition, the sea surface temperature at different scales is expanded into components T n (x,y) is represented as: in, represents the B-spline wavelet coefficient of a single sensor on the nth layer, Δ n Indicates the base resolution length associated with the scale index.

6. The method for fusion of sea surface temperature based on ocean area according to claim 5, characterized in that: The step S4 includes using a regularized variational method to solve the optimal wavelet coefficients, as follows: Among them, Ω represents the spatial extent of the ocean area, represents the regularization term; ε(T) represents the data fitting residual.

7. The method for fusion of sea surface temperature based on ocean area according to claim 6, characterized in that: In step S4, based on the obtained wavelet expansion components at different scales, a multi-scale variational fusion model of each sensor is further obtained by scale superposition: in, represents the multi-scale variational fusion model, and N represents the total number of layers of signal decomposition.

8. A sea surface temperature fusion system based on an ocean region, the system being applied to the method according to any one of claims 1 to 7, characterized in that: The system comprises: The first processing module is used to obtain multi-source sea surface temperature data within a specified area and time range; A second processing module, used for performing quality control on the acquired multi-source sea surface temperature data; The third processing module is used to perform inter-sensor deviation correction on the multi-source sea surface temperature data after quality control; The fourth processing module is used to calculate the wavelet coefficients by fitting the wavelet basis using the variational method for the multi-source sea surface temperature after quality control and deviation correction, so as to construct a multi-scale variational analysis fusion model to obtain the sea surface temperature fusion product at the target time.

Images