A sea surface brightness temperature differential reconstruction method and system based on double-region fitting of a spaceborne synthetic aperture microwave radiometer

Abstract

The present application relates to the field of marine science and microwave remote sensing technology, and particularly relates to a kind of satellite-borne synthetic aperture microwave radiometer sea surface brightness temperature difference reconstruction method and system based on double region fitting.The method first selects two fixed reference regions with different spatial distribution patterns of brightness temperature, calculates the simulated brightness temperature and visibility vector;Then calculate the target simulated brightness temperature of the observed region, establish a weighted least squares fitting model to solve the fitting coefficient, construct the synthetic reference brightness temperature and visibility that match the physical characteristics and observation geometry of the target region;Finally, calculate the difference visibility and retrieve the difference brightness temperature, and obtain the reconstructed brightness temperature by compensating the synthetic reference brightness temperature.The present application retains the basis of traditional G matrix physical reconstruction, reduces the brightness temperature to be reconstructed through fitting algorithm, suppresses the noise amplification effect of underdetermined problem, and supports multi-polarization weighting design, enhances the applicability.

Description

Technical Field

[0001] This invention relates to the fields of marine science and microwave remote sensing technology, and in particular to a method and system for differential reconstruction of sea surface brightness temperature based on a spaceborne integrated aperture microwave radiometer using dual-region fitting. Background Technology

[0002] An interferometric synthetic aperture microwave radiometer is a passive microwave remote sensing system that obtains high-resolution brightness temperature distribution in space through interferometry. Its basic principle is to receive radiation signals from the target area using a spatial baseline array, obtain a visibility function (VF) through cross-correlation, and then recover the brightness temperature distribution of the target scene from the visibility function using a brightness temperature reconstruction method. Assuming the visibility function has M sample values ​​and the spatial domain discretization grid has N grid points, the relationship between visibility and brightness temperature can be discretized and represented as a matrix equation: in and These are column vectors representing the visibility function sample values ​​and the discretized scene brightness temperature, respectively. In reality, the system impulse response matrix, known as the G matrix, is formed by integrating the system parameters and array configuration of the aperture radiometer. The G matrix is ​​typically constructed directly from the antenna patterns of each element and the measured values ​​of the normalized spatial decorrelation effect function. The brightness temperature distribution of the observed scene is reconstructed by solving a system of linear equations. in To reconstruct brightness temperature, The pseudo-inverse of the system's impulse response matrix. Using the visibility function, this brightness temperature reconstruction method is also known as the G-matrix method. The main sources of error in the G-matrix method are as follows: 1. Underdetermined solution error: When the number of samples M in the observation equation is less than the number of grid points N in the spatial domain discrete reconstruction, the system is an underdetermined system of equations. In this case, there is no unique solution, and it needs to be solved through pseudo-inverse form. Therefore, the reconstruction result is sensitive to noise and will produce the Gibbs effect.

[0003] 2. Antenna pattern error: Due to measurement errors and antenna deformation, the antenna pattern on the track is inconsistent with the ground measurement pattern.

[0004] To address the aforementioned issues, a differential reconstruction method is introduced. This method utilizes prior information about the observation scene and system parameters to model instrument observation results, and performs brightness temperature reconstruction on the differential visibility between actual and simulated observation results, reducing system errors introduced by the ill-posedness of the system response matrix. Differential reconstruction further utilizes prior scene information, such as geographical location information like Earth regions, oceans, and landmasses, to further mitigate the Gibbs effect by reducing the brightness temperature contrast of the original scene's brightness temperature distribution, thereby reducing the inversion method's sensitivity to system errors and improving inversion performance.

[0005] Several external calibration and reconstruction methods based on the differential principle have been proposed internationally. A typical example is the Flat Target Transformation (FTT) algorithm. The FTT algorithm originated from the SMOS mission (see reference: M. Martín-Neira, M. Suess, J. Kainulainen, and F. Martín-Porqueras, “The flattarget transformation,” IEEE Trans. Geosci. Remote Sens., vol. 46, no. 3, pp.613–620, Mar. 2008.). According to the Corbella equation, the amplification of system error is proportional to the temperature difference between the scene brightness temperature and the receiver temperature. FTT introduces a reference “flat target temperature” to reduce the brightness temperature to be reconstructed, thereby significantly reducing the amplification factor. Its essence is to perform differential transformation on visibility, construct the “flat target” visibility (Flat Target Response, FTR), and subtract its proportional component from the observed visibility. The flat target commonly used is the cold sky. However, due to the significant differences between the brightness temperature distribution in cold-space scenes and ocean scenes, and the influence of Earth's brightness temperature radiation from the back lobe on cold-space observation conditions, it is difficult to meet the requirements for high-precision brightness temperature reconstruction. Therefore, a "spaceborne synthetic aperture microwave radiometer imaging correction method based on differential measurement" (reference patent number: CN202511376618.2) is proposed. This differential algorithm proposes to use a modelable "reference area brightness temperature" "Using cold air as the differential reference, its reconstruction equation is:" in The pseudo-inverse of the system's impulse response matrix. Let be the visibility function of the observed area. For the visibility function of the reference area, The brightness temperature of the reference region is used. Because this method uses a single fixed reference scene, it will be referred to as the "single reference region differential reconstruction method" in the following text.

[0006] A fixed, single reference source cannot adapt to the dynamic changes in brightness temperature across global observation scenarios. Whether using cold air or a single calm sea area as a reference, its physical properties (brightness temperature distribution, polarization characteristics, etc.) are relatively fixed in time and space. However, the global ocean brightness temperature distribution is influenced by various factors such as sea surface physical parameters (e.g., sea surface temperature, sea surface wind speed, sea surface salinity), observation geometry, and atmospheric environment, exhibiting a large dynamic range and spatial variability.

[0007] In differential reconstruction theory, reconstruction accuracy is highly dependent on the degree of matching between the reference scene and the target scene. When the brightness temperature characteristics of the target area to be observed (such as a highly dynamic sea area affected by the environment) differ significantly from the selected single reference scene, the residual visibility after differential (i.e., the difference between the target visibility and the reference visibility) still contains a large signal component. This will lead to the inability to effectively suppress the noise amplification effect caused by the underdetermined problem in the subsequent solution of the ill-conditioned inverse (inversion of the G matrix), and the residual systematic error cannot be completely canceled out.

[0008] Therefore, a single reference standard cannot meet the high-precision reconstruction requirements of large brightness temperature dynamic change areas globally, limiting the inversion performance of interferometric synthetic aperture radiometers on a global scale. Summary of the Invention

[0009] The purpose of this application is to overcome the above-mentioned defects of the prior art and thus provide a method and system for differential reconstruction of sea surface brightness temperature based on dual-region fitting of a spaceborne integrated aperture microwave radiometer.

[0010] To address the aforementioned technical problems, the technical solution provided in this application offers a method for differential reconstruction of sea surface brightness temperature based on dual-region fitting of a spaceborne synthetic aperture microwave radiometer, comprising: Step 1: Select two fixed regions with different brightness temperature spatial distribution patterns as the first reference region and the second reference region, and calculate the simulated brightness temperature of the first reference region respectively. Simulated brightness temperature of the second reference region and the corresponding first reference region visibility vector Second reference region visibility vector ; Step 2: Calculate the simulated brightness temperature of the target area to be observed. Establish a simulated brightness temperature based on the first reference region. Simulated brightness temperature of the second reference region Using a weighted least squares fitting model with a basis, a dynamic weighted fitting method is used to solve for the first fitting coefficient. Second fitting coefficient In order to construct a synthetic reference benchmark that matches the physical characteristics and observation geometry of the area to be observed; Step 3: Based on the first fitting coefficients Second fitting coefficient Simulated brightness temperature for the first reference region Simulated brightness temperature of the second reference region Linear combination construction of synthetic reference brightness temperature and the visibility vector of the first reference region Second reference region visibility vector Constructing a synthetic reference visibility through linear combination ; Step 4: Calculate the actual observed visibility of the area to be observed. With synthetic reference visibility Differential visibility is obtained by inverting and reconstructing the differential brightness temperature, and then a synthetic reference brightness temperature is used. Compensation is performed on the differential brightness temperature to obtain the reconstructed brightness temperature of the target area. .

[0011] As an improvement to the above technical solution, step 1 specifically includes: determining a first candidate sea area and a second candidate sea area in geographic space as a first reference area and a second reference area, respectively; collecting observation snapshots of the satellite passing through the first reference area and the second reference area within a preset update period of M days; performing quality control on the observation snapshots using environmental background field data to remove transient snapshots with high wind speeds, precipitation, or sea surface temperature anomalies; calculating the full polarization simulated brightness temperature of the observation snapshots after quality control based on the radiative transfer model, and taking the average value of the snapshots as the simulated brightness temperature of the first reference area. Simulated brightness temperature of the second reference region ; Calculate the visibility vector of the corresponding first reference region Second reference region visibility vector .

[0012] As an improvement to the above technical solution, the environmental background field data includes parameters such as sea temperature, sea salinity, wind speed, total water vapor, total liquid water, and offshore distance. The observation snapshots are filtered by setting absolute value threshold ranges and standard deviation threshold ranges.

[0013] As an improvement to the above technical solution, when no observation snapshot meeting the screening criteria is obtained within a certain update cycle, the simulated brightness temperature of the first reference region from the previous update cycle is inherited. Simulated brightness temperature of the second reference region Visibility vector of the first reference region Second reference region visibility vector .

[0014] As an improvement to the above technical solution, in step 2, the first fitting coefficient Second fitting coefficient Satisfy constraints The weighted least squares fitting model satisfies: ; in, This indicates the value of the variable that makes the function below reach its minimum. The Euclidean norm of a vector. It is a positive definite weighted matrix.

[0015] As an improvement to the above technical solution, the positive definite weighting matrix The weighting matrix is ​​used to adjust the weights of different polarization brightness temperature components during the fitting process, wherein the positive definite weighting matrix... Represented as: ; in, For XX polarization weights, For YY polarization weights, For XY polarization weights, for 3D identity matrix The number of brightness temperature pixels used in calculating the fitting coefficients.

[0016] As an improvement to the above technical solution, in step 3, the reference brightness temperature for:

[0017] The reference visibility for: .

[0018] As an improvement to the above technical solution, step 4 specifically includes: Calculate the actual observed visibility of the area to be observed. With synthetic reference visibility Differential visibility : ; Using the pseudo-inverse algorithm of the system response matrix to determine differential visibility Inversion reconstruction is performed to obtain differential brightness temperature. : ; in, This is the pseudo-inverse matrix of the system response matrix calculated from the antenna pattern; Based on the synthetic reference brightness temperature Compensation is performed on the differential brightness temperature to obtain the reconstructed brightness temperature of the target area. : .

[0019] To achieve another objective of the present invention, the present invention also provides a spaceborne integrated aperture microwave radiometer sea surface brightness temperature differential reconstruction system based on dual-region fitting, comprising: The reference region calculation module is used to select two fixed regions with different brightness temperature spatial distribution patterns as the first reference region and the second reference region, and to calculate the simulated brightness temperature of the first reference region respectively. Simulated brightness temperature of the second reference region and the corresponding first reference region visibility vector Second reference region visibility vector ; The fitting coefficient solution module is used to calculate the simulated brightness temperature of the target in the area to be observed. Establish a simulated brightness temperature based on the first reference region. Simulated brightness temperature of the second reference region Using a weighted least squares fitting model with a basis, a dynamic weighted fitting method is used to solve for the first fitting coefficient. Second fitting coefficient In order to construct a synthetic reference benchmark that matches the physical characteristics and observation geometry of the area to be observed; The synthetic reference quantity construction module is used to construct reference quantities based on the first fitting coefficients. Second fitting coefficient Simulated brightness temperature for the first reference region Simulated brightness temperature of the second reference region Linear combination construction of synthetic reference brightness temperature and the visibility vector of the first reference region Second reference region visibility vector Constructing a synthetic reference visibility through linear combination ;and The brightness temperature difference reconstruction module is used to calculate the actual observed visibility of the area to be observed. With synthetic reference visibility Differential visibility is obtained by inverting and reconstructing the differential brightness temperature, and then a synthetic reference brightness temperature is used. Compensation is performed on the differential brightness temperature to obtain the reconstructed brightness temperature of the target area. .

[0020] Compared with the prior art, the advantages of this invention are that while retaining the foundation of traditional G matrix physical reconstruction, it reduces the brightness temperature to be reconstructed through fitting algorithm, suppresses the noise amplification effect of underdetermined problem, and supports multi-polar weighted design, thus enhancing applicability. Attached Figure Description

[0021] Figure 1 A flowchart of the sea surface brightness temperature differential reconstruction method based on dual-region fitting for a spaceborne integrated aperture microwave radiometer provided in Embodiment 1 of the invention; Figure 2 This is a map showing the distribution of reference regions in the Northern and Southern Hemispheres. Figure 3(a) shows the reference brightness temperature of XX polarization in the Northern and Southern Hemispheres; Figure 3(b) shows the YY polarization reference brightness temperature in the Northern and Southern Hemispheres. Figure 4(a) shows a comparison of the fitting effect between the XX polarization synthetic reference brightness temperature and the simulated brightness temperature of the target in the observation area; Figure 4(b) shows a comparison of the fitting effect between the YY polarization synthesized reference brightness temperature and the simulated brightness temperature of the target in the observation area; Figure 5(a) shows the error between the reconstructed brightness temperature and the true brightness temperature using the XX polarization dual-scene dynamic fitting difference reconstruction method. Figure 5(b) shows the error between the reconstructed brightness temperature and the true brightness temperature of the YY polarization dual-scene dynamic fitting difference reconstruction method; Figure 5(c) shows the error between the reconstructed brightness temperature and the true brightness temperature of the XX polarization single-reference region differential reconstruction method; Figure 5(d) shows the error between the reconstructed brightness temperature and the true brightness temperature using the YY polarization single-reference region differential reconstruction method. Detailed Implementation

[0022] The technical solutions provided in this application are further illustrated below with reference to the embodiments.

[0023] Example 1 This embodiment proposes a differential reconstruction method for sea surface brightness temperature of a spaceborne integrated aperture microwave radiometer based on dual-region fitting. The method further reduces the brightness temperature to be reconstructed by dynamic fitting, thereby reducing the reconstruction algorithm error.

[0024] The present invention solves the above-mentioned technical problem through the following technical solution: like Figure 1 As shown in this embodiment, the sea surface brightness temperature differential reconstruction method based on dual-region fitting for spaceborne integrated aperture microwave radiometers includes: Step 1: Select a reference region and calculate the simulated brightness temperature of the reference region. and the corresponding reference visibility vector obtained from equipment observation. ; Step 2: Calculate the simulated brightness temperature of the target area to be observed. Establish a simulated brightness temperature based on the reference region. For a weighted least squares fitting model with a basis of 1, calculate the fitting coefficients. , ; Step 3: Use the obtained fitting coefficients and Simulated brightness temperature of the reference region and reference visibility vector Linear weighted combinations are performed separately to construct a synthetic reference brightness temperature adapted to the target region. and synthetic reference visibility ; Step 4: Calculate the actual visibility of the target area to be observed. With synthetic reference visibility The differential visibility between them is then used for differential reconstruction, and the result is based on the synthetic reference brightness temperature. Compensation is performed to obtain the reconstructed brightness temperature of the target area to be observed. .

[0025] Step 1: Select two fixed regions with significantly different spatial distribution patterns of brightness temperature as reference regions. The selection and screening process specifically includes: determining two candidate regions in geographic space that avoid fixed interference sources, and collecting satellite observation snapshots of the candidate regions within a preset update cycle; using environmental background field data to perform quality control on the observation snapshots, eliminating transient snapshots with high wind speeds, precipitation, or sea surface temperature anomalies; calculating the full polarization simulated brightness temperature (X polarization / Y polarization / XY polarization) of the remaining snapshots after quality control based on the radiative transfer model, and determining the average value of the snapshots as the simulated brightness temperature vector of the reference region for the current update cycle. and the corresponding reference visibility vector The update cycle for the reference area is M days, where M is determined by the stability of the brightness temperature of the reference area over time. The snapshots selected during the update process must simultaneously meet conditions such as low wind speed, no precipitation, and small sea surface temperature standard deviation to ensure the stability of the reference area.

[0026] Step 2: Calculate the simulated brightness temperature of the target in the area to be observed. Establish a simulated brightness temperature based on the reference region. The weighted least squares fitting model with basis , the fitting coefficients a and b are given by satisfying the constraints. Under the following conditions, find the optimal solution that minimizes the following weighted residual norm: in The L2 norm of a vector is used to represent the Euclidean norm. This is a positive definite weighting matrix used to adjust the weights of different polarization brightness temperature components during the fitting process, specifically:

[0027] in, For XX polarization weights, For YY polarization weights, For XY polarization weights, It is an N-dimensional identity matrix, where N is the number of brightness temperature pixels involved in the calculation of fitting coefficients; Step 3: Use the obtained fitting coefficients and Simulated brightness temperature of the reference region and the corresponding reference visibility vector obtained from equipment observation. A weighted combination is performed to construct a synthetic reference brightness temperature adapted to the target region. and synthetic reference visibility The construction process is represented as follows:

[0028] Step 4: Calculate the actual observed visibility of the area to be observed. With synthetic reference visibility Differential visibility between:

[0029] The differential brightness temperature is obtained by inverting and reconstructing the differential visibility using the pseudo-inverse algorithm of the system response matrix.

[0030] in This is the pseudo-inverse matrix of the system response matrix calculated from the antenna pattern; Then the differential brightness temperature and the synthesis reference brightness temperature were compared. Superimposed to obtain the final reconstructed brightness temperature of the observed region:

[0031] When the reference region fails to obtain a snapshot that meets the conditions within a certain update cycle, the simulated brightness temperature of the reference region from the previous cycle is inherited. and the corresponding reference visibility vector obtained from equipment observation. .

[0032] This method overcomes the limitations of traditional fixed single reference sources. Based on the dynamic distribution characteristics of brightness temperature of the radiometer-observed target globally, it selects two fixed regions with significantly different spatial distribution patterns of brightness temperature as reference benchmarks to construct a differential reference system covering the dynamic range of observations. On this basis, using the brightness temperature information from these two reference regions, the simulated brightness temperature of any target region is subjected to least-squares linear fitting to dynamically construct a 'synthetic reference brightness temperature' and 'synthetic reference visibility' that match the observed geometric and physical characteristics of the current target scene. Subsequently, the actual observed visibility of the target region is differentiated from this synthetic reference visibility, and the differential brightness temperature is reconstructed using the pseudo-inverse of the G matrix. Finally, the reference brightness temperature is superimposed to obtain the final brightness temperature result.

[0033] The specific implementation process of this invention will be described below using an example of an L-band spaceborne interferometric radiometer. L-band observations are significantly affected by Faraday rotation, exhibiting unique differences in brightness temperature distribution between the Northern and Southern Hemispheres. This is one of the typical manifestations of the "large dynamic range difference" of this invention. Specifically, the spaceborne L-band microwave radiation signal undergoes significant Faraday rotation when passing through the ionosphere. The magnitude and direction of this rotation angle are strictly dependent on the geomagnetic field distribution and the vertical total electron content (VTEC). This geophysical effect leads to a significant asymmetry in observed brightness temperature between the Northern and Southern Hemispheres and introduces a large dynamic range of brightness temperature globally.

[0034] Single-reference-area differential reconstruction methods typically select a reference scene within a fixed, calm sea area (single latitudinal zone). The polarization state and brightness temperature characteristics of this single reference scene cannot encompass the large dynamic range of global variations caused by Faraday rotation. When the target area is located in a different hemisphere or in a region with significant geomagnetic differences, the spatial difference in brightness temperature distribution between the reference scene and the target scene is substantial. This results in the inability to effectively cancel systematic errors during the differential process, ultimately severely impacting the accuracy of global sea surface salinity inversion.

[0035] In this embodiment, the dual reference areas will be specifically selected as specific sea areas in the Northern and Southern Hemispheres. The specific steps are as follows: Reference area and snapshot filtering To ensure the brightness temperature of the reference region obtained by fitting the reference region Visibility with corresponding reference area To ensure physical reliability and effective constraint on the reconstruction of the target region, the reference region must meet two requirements: first, good repeatability of radiation physics, allowing the radiative transfer model to stably simulate under the same conditions with minimal model error; second, spatially / temporally, it should be considered "background" rather than a source of transient events compared to the target region (avoiding extreme brightness temperatures caused by waves, foam, sea ice, river inflows, sea surface oil pollution, or radio frequency interference). Calm sea surfaces and areas with minimal intra-month variations can minimize the impact of sea surface roughness, foam, and nonlinear scattering on brightness temperature, minimizing the simulation error of the radiative transfer model at these locations, thereby improving the physical reliability and long-term stability of the fitting coefficients. Region selection involves two steps: primary screening and fine screening.

[0036] Geographical and spatial criteria for candidate regions (preliminary screening): Hemispherical distribution: A fixed sea area was selected in both the Southern and Northern Hemispheres: 34°~38°S, 112°~122°W in the Southern Hemisphere and 25°~29°N, 149°~157°W in the Northern Hemisphere to ensure that the fitted brightness temperature can cover the brightness temperature distribution of different sea areas around the world.

[0037] Keep away from the coastline: The center of the candidate area is more than 1000km away from the nearest coastline to avoid the land edge entering the antenna's field of view.

[0038] Keep away from fixed sources of interference (radio frequency interference, oil field platforms, densely populated fishing areas, and large shipping lanes, etc.): reduce transient radiation caused by human activities.

[0039] Spatial uniformity: The region should have uniform radiation properties at the selected scale (no strong gradient in the background field), and ensure that there are more than 5 consecutive snapshot samples in a region to average and reduce system noise.

[0040] Background field control (fine filtering): To quantify the properties of calm sea surfaces and couple them with radiative transfer models, relevant sea surface parameters need to be extracted from the fifth-generation global atmospheric reanalysis dataset ERA5 (ECMWFReanalysisv5) released by the European Centre for Medium-Range Weather Forecasts (ECMWF) to formulate threshold quantification screening criteria. Calm sea areas are selected by establishing absolute value thresholds and standard deviation thresholds for parameters such as wind speed, sea surface temperature, water vapor content, and liquid water content. The specific criteria are as follows:

[0041] Next, snapshot screening is performed. The fitting coefficient update period is M days. Therefore, all snapshots of the satellite passing through the candidate sea area are collected within the M-day period. ERA5 background field data interpolation is performed on each snapshot, and the snapshots are screened according to the ERA5 threshold. The selected snapshots are used to generate simulated brightness temperatures through radiative transfer model simulation. If there are no qualified snapshots in a certain period, the simulated brightness temperature of the reference area in the previous period is inherited. and the corresponding reference visibility vector obtained from equipment observation. .

[0042] Figure 2 The final selected reference regions for the Northern and Southern Hemispheres are given, and Figures 3(a) and 3(b) show the reference brightness temperature tangents for the Northern and Southern Hemispheres.

[0043] Solving for fitting coefficients After obtaining a snapshot of the reference region, its brightness temperature information is used to perform a linear fit on the brightness temperature of the target region. Multi-polarization weights are introduced during the fitting process to ensure that the contributions of different polarization channels are reasonably reflected. The objective function for optimization is to minimize the weighted sum of squared errors.

[0044] in, Simulate brightness temperature for the target in the area to be observed. Simulated brightness temperatures for reference regions in the Northern and Southern Hemispheres, respectively. For polarization weights, specifically in, For XX polarization weights, For YY polarization weights, For XY polarization weights, It is an N-dimensional identity matrix, where N is the number of brightness temperature pixels used to calculate the fitting coefficients.

[0045] The constraints are This optimization problem is solved using the constrained least squares method.

[0046] Reference brightness temperature and visibility calculation After determining the fitting coefficients a and b, the synthetic reference brightness temperature of the region to be observed can be calculated: Simultaneously, the synthetic reference visibility of the observed area is calculated using the same linear combination relationship:

[0047] in, These are the visibility functions for the two reference regions, respectively.

[0048] Figures 4(a) and 4(b) show a comparison of the fitting effect between the synthesized reference brightness temperature and the theoretical simulated brightness temperature of the target region. It can be seen that after dual-region weighted fitting, the synthesized reference brightness temperature and the simulated brightness temperature of the target region to be reconstructed highly overlap across the entire field of view, with minimal residuals. This means that by using the 'synthetic visibility' corresponding to the synthesized reference brightness temperature for differential calculation, the background brightness temperature can be offset to the maximum extent, retaining only the high-frequency disturbance information of the target scene and the system residual error, thus meeting the input requirements for high-precision reconstruction.

[0049] Brightness temperature differential reconstruction Measured visibility of the observation area With reference visibility Perform the difference: The difference visibility is then reconstructed using the pseudo-inverse of the G matrix: Finally, the differential reconstruction result is added to the reference brightness temperature to obtain the final reconstructed brightness temperature of the observed area: To verify the effectiveness of the dual-region fitting differential reconstruction method proposed in this invention, this embodiment selected a set of typical L-band simulation observation data from around the world for testing and compared it with the single-reference-region differential reconstruction method. Figures 5(a)-5(d) show the comparison of the reconstruction brightness temperature error between the two reconstruction methods and the true brightness temperature error.

[0050] Through the above steps, the present invention achieves brightness temperature differential reconstruction using a fixed reference area, effectively suppressing the influence of noise and improving the stability and accuracy of reconstruction.

[0051] Example 2 This embodiment presents a spaceborne integrated aperture microwave radiometer sea surface brightness temperature differential reconstruction system based on dual-region fitting, comprising: The reference region calculation module is used to select two fixed regions with different brightness temperature spatial distribution patterns as the first reference region and the second reference region, and to calculate the simulated brightness temperature of the first reference region respectively. Simulated brightness temperature of the second reference region and the corresponding first reference region visibility vector Second reference region visibility vector ; The fitting coefficient solution module is used to calculate the simulated brightness temperature of the target in the area to be observed. Establish a simulated brightness temperature based on the first reference region. Simulated brightness temperature of the second reference region Using a weighted least squares fitting model with a basis, a dynamic weighted fitting method is used to solve for the first fitting coefficient. Second fitting coefficient In order to construct a synthetic reference benchmark that matches the physical characteristics and observation geometry of the area to be observed; The synthetic reference quantity construction module is used to construct reference quantities based on the first fitting coefficients. Second fitting coefficient Simulated brightness temperature for the first reference region Simulated brightness temperature of the second reference region Linear combination construction of synthetic reference brightness temperature and the visibility vector of the first reference region Second reference region visibility vector Constructing a synthetic reference visibility through linear combination ;and The brightness temperature difference reconstruction module is used to calculate the actual observed visibility of the area to be observed. With synthetic reference visibility Differential visibility is obtained by inverting and reconstructing the differential brightness temperature, and then a synthetic reference brightness temperature is used. Compensation is performed on the differential brightness temperature to obtain the reconstructed brightness temperature of the target area. .

[0052] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications or equivalent substitutions to the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for differential reconstruction of sea surface brightness temperature based on dual-region fitting of a spaceborne synthetic aperture microwave radiometer, comprising: Step 1: Select two fixed regions with different brightness temperature spatial distribution patterns as the first reference region and the second reference region, and calculate the simulated brightness temperature of the first reference region respectively. Simulated brightness temperature of the second reference region and the corresponding first reference region visibility vector Second reference region visibility vector ; Step 2: Calculate the simulated brightness temperature of the target area to be observed. Establish a simulated brightness temperature based on the first reference region. Simulated brightness temperature of the second reference region Using a weighted least squares fitting model with a basis, a dynamic weighted fitting method is used to solve for the first fitting coefficient. Second fitting coefficient In order to construct a synthetic reference benchmark that matches the physical characteristics and observation geometry of the area to be observed; Step 3: Based on the first fitting coefficient Second fitting coefficient Simulated brightness temperature for the first reference region Simulated brightness temperature of the second reference region Linear combination construction of synthetic reference brightness temperature and the visibility vector of the first reference region Second reference region visibility vector Constructing a synthetic reference visibility through linear combination ; Step 4: Calculate the actual observed visibility of the area to be observed. With synthetic reference visibility Differential visibility is obtained by inverting and reconstructing the differential brightness temperature, and then a synthetic reference brightness temperature is used. Compensation is performed on the differential brightness temperature to obtain the reconstructed brightness temperature of the target area. .

2. The method for differential reconstruction of sea surface brightness temperature based on dual-region fitting of a spaceborne synthetic aperture microwave radiometer according to claim 1, characterized in that, Step 1 specifically includes: Geographically, the first candidate sea area and the second candidate sea area are respectively designated as the first reference area and the second reference area; Collect observation snapshots of the satellite passing through the first and second reference areas within a preset update cycle of M days; The quality control of the observation snapshots was performed using environmental background field data to remove transient snapshots with high wind speeds, precipitation, or sea surface temperature anomalies. The simulated brightness temperature of the fully polarized observation snapshots after quality control was calculated based on the radiative transfer model, and the average value of the snapshots was taken as the simulated brightness temperature of the first reference region. Simulated brightness temperature of the second reference region ; Calculate the visibility vector of the corresponding first reference region Second reference region visibility vector .

3. The method for differential reconstruction of sea surface brightness temperature based on dual-region fitting of a spaceborne synthetic aperture microwave radiometer according to claim 2, characterized in that, The environmental background field data includes parameters such as sea temperature, sea salinity, wind speed, total water vapor, total liquid water, and offshore distance. The observation snapshots are filtered by setting absolute value threshold ranges and standard deviation threshold ranges.

4. The method for differential reconstruction of sea surface brightness temperature based on dual-region fitting of a spaceborne synthetic aperture microwave radiometer according to claim 2, characterized in that, If no observation snapshots meeting the screening criteria are obtained within a certain update cycle, the simulated brightness temperature of the first reference region from the previous update cycle is inherited. Simulated brightness temperature of the second reference region Visibility vector of the first reference region Second reference region visibility vector .

5. The method for differential reconstruction of sea surface brightness temperature based on dual-region fitting of a spaceborne synthetic aperture microwave radiometer according to claim 1, characterized in that, In step 2, the first fitting coefficient Second fitting coefficient Satisfy constraints The weighted least squares fitting model satisfies: ； in, This indicates the value of the variable that makes the function below reach its minimum. The Euclidean norm of a vector. It is a positive definite weighted matrix.

6. The method for differential reconstruction of sea surface brightness temperature based on dual-region fitting of a spaceborne synthetic aperture microwave radiometer according to claim 5, characterized in that, The positive definite weighting matrix The weighting matrix is ​​used to adjust the weights of different polarization brightness temperature components during the fitting process, wherein the positive definite weighting matrix... Represented as: ； in, For XX polarization weights, For YY polarization weights, For XY polarization weights, for 3D identity matrix The number of brightness temperature pixels used in calculating the fitting coefficients.

7. The method for differential reconstruction of sea surface brightness temperature based on dual-region fitting of a spaceborne synthetic aperture microwave radiometer according to claim 1, characterized in that, In step 3, the reference brightness temperature for: The reference visibility for: 。 8. The method for differential reconstruction of sea surface brightness temperature based on dual-region fitting of a spaceborne synthetic aperture microwave radiometer according to claim 1, characterized in that, Step 4 specifically includes: Calculate the actual observed visibility of the area to be observed. With synthetic reference visibility Differential visibility : ； Using the pseudo-inverse algorithm of the system response matrix to determine differential visibility Inversion reconstruction is performed to obtain differential brightness temperature. : ； in, This is the pseudo-inverse matrix of the system response matrix calculated from the antenna pattern; Based on synthetic reference brightness temperature Compensation is performed on the differential brightness temperature to obtain the reconstructed brightness temperature of the target area. : 。 9. A spaceborne synthetic aperture microwave radiometer sea surface brightness temperature differential reconstruction system based on dual-region fitting, characterized in that, include: The reference region calculation module is used to select two fixed regions with different brightness temperature spatial distribution patterns as the first reference region and the second reference region, and to calculate the simulated brightness temperature of the first reference region respectively. Simulated brightness temperature of the second reference region and the corresponding first reference region visibility vector Second reference region visibility vector ; The fitting coefficient solution module is used to calculate the simulated brightness temperature of the target in the area to be observed. Establish a simulated brightness temperature based on the first reference region. Simulated brightness temperature of the second reference region Using a weighted least squares fitting model with a basis, a dynamic weighted fitting method is used to solve for the first fitting coefficient. Second fitting coefficient In order to construct a synthetic reference benchmark that matches the physical characteristics and observation geometry of the area to be observed; The synthetic reference quantity construction module is used to construct reference quantities based on the first fitting coefficients. Second fitting coefficient Simulated brightness temperature for the first reference region Simulated brightness temperature of the second reference region Linear combination construction of synthetic reference brightness temperature and the visibility vector of the first reference region Second reference region visibility vector Constructing a synthetic reference visibility through linear combination ; and The brightness temperature difference reconstruction module is used to calculate the actual observed visibility of the area to be observed. With synthetic reference visibility Differential visibility is obtained by inverting and reconstructing the differential brightness temperature, and then a synthetic reference brightness temperature is used. Compensation is performed on the differential brightness temperature to obtain the reconstructed brightness temperature of the target area. .

Images