Brightness temperature correction method based on improved ocean target transformation

Abstract

The invention discloses a brightness temperature correction method based on improved ocean target transformation. A sea area reference area is selected; extracting data snapshots in the defined longitude and latitude range of the reference area; for pixel points in each extracted data snapshot, screening effective pixel points by taking a wind speed, a Faraday rotation angle and a polarization rotation angle as constraint conditions, and determining an effective data snapshot according to a screening result; for each effective data snapshot, acquiring the brightness temperature of an observation point and corresponding sea-land surface physical parameters, calculating a simulated brightness temperature by using a forward model, and performing pixel-by-pixel subtraction on the simulated brightness temperature and the brightness temperature of the observation point to generate a brightness temperature deviation field; dividing an antenna space cosine coordinate system into grids, and interpolating the brightness temperature deviation into the grids; the brightness temperature deviation values in the effective grids are accumulated, the median in each grid is taken, the grid data quality is evaluated, a deviation field OTT is generated, the OTT correction value is mapped to a global measurement point, and the corrected brightness temperature is generated.

Description

Technical Field

[0001] This invention relates to a method for salinity inversion correction of a spaceborne integrated aperture microwave radiometer, which is applicable to dynamically eliminating brightness temperature system errors during salinity inversion and belongs to the field of satellite microwave ocean remote sensing technology. Background Technology

[0002] The integrated aperture radiometer (API) is an important payload of salinity sounding satellites, acquiring ocean radiation brightness temperature data and retrieving sea surface salinity. During satellite data inversion, a forward model predicts brightness temperature values ​​under specific parameter conditions, establishing a theoretical relationship between surface physical parameters (such as ocean salinity, temperature, and wind speed) and measured brightness temperature, providing a benchmark reference for the inversion algorithm. Analysis during the SMOS satellite's commissioning phase revealed significant discrepancies between predicted and observed brightness temperatures in certain scenarios. Specifically, in sea areas far from land, comparisons based on the antenna coordinate system show a systematic error characteristic in the difference between the two. Such errors directly affect the reliability of the salinity inversion algorithm, resulting in data products whose accuracy fails to meet practical application requirements.

[0003] To effectively eliminate the impact of inconsistencies between the forward model and observed data on subsequent salinity inversion, the SMOS satellite employs the Ocean Target Transform (OTT) algorithm to correct systematic biases. Its core components are: selecting data from stable sea states, calculating the deviation between the model-predicted brightness temperature and the observed brightness temperature, and establishing a global systematic bias model. Considering the high sensitivity of the forward model to parameters such as sea surface roughness and ionospheric interference, traditional OTT algorithms do not directly consider these factors during data selection. This may lead to non-systematic environmental noise being mistakenly included in the systematic bias calculation, resulting in uncontrollable spatial heterogeneity in the final OTT correction field, thus affecting the accuracy of salinity inversion. Summary of the Invention

[0004] The technical problem solved by this invention is: addressing the problems of traditional OTT algorithms, this invention uses wind speed, Faraday rotation angle and polarization rotation angle as additional constraints to reduce the impact of non-systematic brightness temperature fluctuations and obtain a more accurate systematic deviation field.

[0005] The solution of this invention is: a brightness temperature correction method based on improved marine target transformation, comprising:

[0006] Select a reference area of ​​the sea; read satellite observation brightness temperature data and observation indicators, and at the same time obtain the physical parameters of the sea and land surfaces, interpolate the physical parameters to the observation point location and store them to ensure spatial alignment with the measured data;

[0007] Extract data snapshots within the latitude and longitude range defined in the reference area;

[0008] For each extracted data snapshot, valid pixels are filtered using wind speed, Faraday rotation angle, and polarization rotation angle as constraints, and valid data snapshots are determined based on the filtering results.

[0009] For each valid data snapshot, the brightness temperature of the observation point and the corresponding land and sea surface physical parameters are obtained. The simulated brightness temperature is calculated using a forward model and then subtracted from the brightness temperature of the observation point pixel by pixel to generate a brightness temperature deviation field.

[0010] The antenna space cosine coordinate system is divided into grids, and the brightness temperature deviation is interpolated into the grids. The brightness temperature deviation values ​​in the effective grids are accumulated and the median of each grid is taken. The grid data quality is evaluated, and the deviation field OTT is generated. The OTT correction is mapped to global measurement points to generate the corrected brightness temperature. The effective grids refer to the grids that contain deviation data after interpolation.

[0011] Preferably, effective pixels are selected by the following methods: removing pixels with wind speeds greater than 10 m / s; removing pixels with Faraday rotation angles greater than 10°; and removing pixels with polarization rotation angles between 42° and 48°.

[0012] Preferably, the effective data snapshot is a data snapshot in which the proportion of effective pixels exceeds 50%.

[0013] Preferably, the deviation field OTT is generated in the following manner:

[0014] The median of the deviation values ​​within each grid is taken; the root mean square error of the deviation values ​​within each grid is calculated to evaluate the quality of the grid data; grids with root mean square errors exceeding a preset threshold are identified as noisy grids, and a zero correction operation is performed, i.e., the deviation field value of the grid is set to 0; the median of the remaining grids excluding the noisy grids is taken as the deviation field OTT.

[0015] Preferably, the grid containing data with a root mean square error greater than or equal to 10K is identified as a noisy grid.

[0016] Preferably, the reference area includes an ascending rail area and a descending rail area.

[0017] Preferably, the reference area is a calm sea area with significant salinity variation, low radio frequency interference, and far from the coast; the significant salinity variation is defined as an annual average variation of 0.1-0.5 psu (psu is a unit of salinity measurement, representing the number of grams of dissolved salt per kilogram of seawater), the low radio frequency interference is defined as an RFI pollution rate of less than 10%, and far from the coast is defined as a distance from the shore greater than 500 km.

[0018] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method.

[0019] A computer software product includes: a processor and a storage device;

[0020] Storage device for storing one or more programs.

[0021] When the one or more programs are executed by one or more processors, the one or more processors implement the brightness temperature correction method based on the improved marine target transformation.

[0022] The advantages of this invention compared to the prior art are:

[0023] This invention uses wind speed, Faraday rotation angle, and polarization rotation angle as additional constraints to suppress sea surface roughness interference, eliminate brightness temperature anomalies caused by extreme sea states, and improve the reliability of bias statistics. It also reduces the significant distortion of polarization signals caused by ionospheric irregularities or strong disturbances, avoids brightness temperature inversion biases due to abnormal changes in polarization state, and improves the stability of the correlation between brightness temperature and geophysical parameters. This strategy is particularly helpful in improving data quality and the reliability of inversion results, especially in low latitudes or during periods of peak solar activity. Furthermore, it avoids potential nonlinear effects or interference signals within this angular range, preventing unexpected polarization state transitions from affecting the brightness temperature calculation model. Excluding pixels with a polarization rotation angle near 45° also avoids potential nonlinear effects or interference signals within this angular range, preventing unexpected polarization state transitions from affecting the brightness temperature calculation model. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating an improved OTT algorithm according to an exemplary embodiment of the present disclosure;

[0025] Figure 2 This is a schematic diagram illustrating the location of a selected reference sea area according to an exemplary embodiment of the present disclosure;

[0026] Figure 3 The observed brightness temperature of the filtered Snapshot ID 608333430 as illustrated in this disclosure according to an exemplary embodiment;

[0027] Figure 4 This disclosure provides a simulated brightness temperature for the filtered Snapshot ID 608333430, as illustrated in an exemplary embodiment.

[0028] Figure 5 The brightness temperature deviation of the filtered Snapshot ID 608333430 is shown in this disclosure according to an exemplary implementation. Detailed Implementation

[0029] The present invention will be further described below with reference to the embodiments.

[0030] A brightness temperature correction method based on improved marine target transformation, such as Figure 1 As shown, the steps are as follows:

[0031] (1.1) Step S1. Select a calm sea area in the Pacific Ocean that meets the requirements of significant salinity variation, low radio frequency interference (RFI) and far from the coast. The selected reference area must meet the orbital coverage requirements and be divided into ascending orbit and descending orbit areas to ensure data spatiotemporal consistency.

[0032] like Figure 2 As shown, in this example, the ascending orbit region 1 and the descending orbit region 2 are defined as reference areas: Region 1 (ascending orbit) covers latitudes from 5°S to 45°S, with the top centered at 116°W and the bottom centered at 126°W, and longitudes distributed in a rectangular pattern with a width of 42°; Region 2 (descending orbit) has the top centered at 128°W and the bottom at 117°W, and longitudes distributed in a rectangular pattern with a width of 42°, forming an orbital coverage area symmetrical to the ascending orbit region.

[0033] Step S2. For the ten half-orbit data within two weeks, first read the brightness temperature data T observed by the onboard integrated aperture microwave radiometer. L1C It also acquires observation indicators such as orbital parameters, satellite position, orientation, and scene coordinates, as well as physical parameters of the sea and land surfaces such as sea surface temperature (SST), wind speed (SSW), sea surface salinity (SSS), wind direction, and land-sea mask. The physical parameters are then interpolated to the observation pixel positions using a spatial interpolation algorithm to ensure spatial alignment with the measured data.

[0034] Step S3. The read and integrated data is categorized and stored according to the ascending and descending orbits. Then, data snapshots are extracted within the latitude and longitude range defined in Step S1. For each pixel in the data snapshot, a strict filtering operation is performed. Pixels with wind speeds greater than 10 m / s are removed because high wind speeds increase sea surface roughness, causing sea surface conditions to deviate from the model assumptions. Setting a wind speed threshold can suppress sea surface roughness interference, eliminate abnormal brightness temperature fluctuations caused by extreme sea conditions, and improve the reliability of deviation statistics.

[0035] Removing pixels with a Faraday rotation angle greater than 10° can distort polarization signals due to strong ionospheric disturbances. Removing pixels with a Faraday rotation angle greater than 10° can reduce the significant distortion of polarization signals caused by ionospheric irregularities or strong disturbances, avoid brightness temperature inversion bias caused by abnormal changes in polarization state, and improve the stability of the correlation between brightness temperature and geophysical parameters. Especially in low latitudes or during peak solar activity, this strategy helps to improve data quality and the reliability of inversion results.

[0036] In L-band (1.4 GHz) satellite remote sensing, when the polarization rotation angle is 45°, the combined effect of Faraday rotation and geometric rotation may cause linearly polarized light to transform into a circular or elliptical polarization state. When the system demodulates only linearly polarized signals, it cannot accurately resolve such signals, resulting in energy distribution deviations or signal loss during brightness temperature inversion. Simultaneously, simplifying the model by ignoring higher-order terms and medium dispersion effects exacerbates the deviation between theoretical and measured rotation angles, leading to polarization correction errors, which are more pronounced at low latitudes or during periods of ionospheric disturbance. These uncorrected polarization-related errors affect the quality of brightness temperature data and reduce the reliability of geophysical products. This invention removes pixels with polarization rotation angles between 42° and 48°, as this angle range exhibits polarization symmetry and easily leads to inversion blind zones. Only snapshots with more than 50% of valid pixels are retained, thus filtering out noise-dominated or severely missing data scenarios and ensuring high-quality data for subsequent analysis.

[0037] The Fast Forward Atmospheric and Surface Emission Model (FASTEM-5) is employed, a high-precision model for simulating the microwave radiation characteristics of the ocean surface. Based on radiative transfer theory, FASTEM-5 fully considers the physical processes of the atmosphere, sea surface, and ocean interior. Inputting the physical parameters of the land and sea surfaces and the antenna observation geometry (incident angle θ, polarization p, etc.), the simulated brightness temperature Tb is calculated. model .

[0038] First, FASTEM-5 calculates the complex permittivity of seawater based on its physical properties. The commonly used calculation formula can be based on a modified form of the Debye model:

[0039]

[0040] Where ω is the angular frequency, ε ∞ It is the high-frequency relative permittivity, τ is the relaxation time, and ε is the high-frequency relative permittivity. s ε is the static relative permittivity, ε0 ​​is the vacuum permittivity, a is an empirical coefficient related to the concentration of the sample, and σ is the conductivity.

[0041] For the emissivity ε of a calm sea surface p (θ,ω,SST,SSS), the emissivity formulas for horizontal polarization (p=H) and vertical polarization (p=V) based on Fresnel formulas are as follows:

[0042] Where θ is the incident angle of the satellite observation.

[0043] Taking further consideration of sky radiation reflected from the surface, based on the above emissivity and combined with the Rayleigh-Jeans approximation, the model brightness temperature formula for a calm sea surface is:

[0044] T model =ε V ·SST+R V ·T sky T BH =ε H ·SST+R H ·T sky

[0045] Among them, T sky The sky was bright and warm.

[0046] Based on the observation parameters of the spaceborne integrated aperture microwave radiometer, the radiation brightness temperature coordinates in latitude and longitude are converted into antenna coordinate grid points in the spatial cosine plane, and a deviation expression between the observed brightness temperature and the simulated brightness temperature is constructed pixel by pixel:

[0047] ΔT(ξ,η)=T model (ξ,η)-T L1C (ξ,η)

[0048] Step S4. Divide the antenna spatial cosine coordinate system (-0.7<ξ<0.7, -0.7<η<0.4) into a uniform grid of 129×129 (the grid size is obtained according to the sampling rules of the spatial cosine coordinate system of the antenna array, which is related to the satellite's antenna array and sampling rules). Accumulate the brightness temperature deviation value of the effective pixel in each grid according to the coordinate position (ξ,η). Use the nearest neighbor interpolation method to interpolate the original deviation data of multiple snapshots to the same spatial cosine coordinate system.

[0049] To reduce the interference of isolated outliers on the statistical results, the median of the deviation values ​​within each grid is taken to generate an outlier-resistant bias field. Next, the standard deviation of the deviation values ​​within each grid is calculated to assess the data quality. For noisy grids with a standard deviation exceeding a preset threshold, a zero-correction operation is performed, setting the bias field value of that grid to 0.

[0050] Figure 3 To implement the observed brightness temperature of the filtered Snapshot ID 608333430 shown in the example, Figure 4 To simulate the brightness temperature of the filtered SnapshotID 608333430 shown in the example, Figure 5 To demonstrate the brightness temperature deviation of the filtered Snapshot ID 608333430 shown in the example.

[0051] Step S5. Using a spherical interpolation algorithm, the pre-constructed bias field in the antenna coordinate system is mapped to each measurement point globally. Based on the coordinates (ξ, η) of each measurement point, the corresponding correction amount T is obtained through interpolation. correct (ξ,η)=TL1C (ξ,η)-OTT(ξ,η).

[0052] This invention improves the accuracy of ocean salinity detection by using an improved brightness temperature correction for ocean target transformation, which can help my country achieve a global ocean salinity detection accuracy of 0.1 psu as soon as possible.

[0053] The present invention further provides a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the steps of the method.

[0054] The present invention also provides a computer software product, comprising: a processor and a storage device;

[0055] Storage device for storing one or more programs.

[0056] When the one or more programs are executed by one or more processors, the one or more processors implement the brightness temperature correction method based on the improved marine target transformation.

[0057] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.

Claims

1. A brightness temperature correction method based on improved marine target transformation, characterized in that... include: Select a reference area for the sea area; Read satellite observation brightness temperature data and observation indicators, and simultaneously obtain physical parameters of the land and sea surfaces. Interpolate the physical parameters to the observation point location and store them to ensure spatial alignment with the measured data. Extract data snapshots within the latitude and longitude range defined in the reference area; For each extracted data snapshot, valid pixels are filtered using wind speed, Faraday rotation angle, and polarization rotation angle as constraints, and valid data snapshots are determined based on the filtering results. For each valid data snapshot, the brightness temperature of the observation point and the corresponding land and sea surface physical parameters are obtained. The simulated brightness temperature is calculated using a forward model and then subtracted from the brightness temperature of the observation point pixel by pixel to generate a brightness temperature deviation field. The antenna space cosine coordinate system is divided into grids, and the brightness temperature deviation is interpolated into the grids. The brightness temperature deviation values ​​in the effective grids are accumulated and the median of each grid is taken. The grid data quality is evaluated, and the deviation field OTT is generated. The OTT correction is mapped to global measurement points to generate the corrected brightness temperature. The effective grids refer to the grids that contain deviation data after interpolation.

2. The method according to claim 1, characterized in that: Valid pixels are filtered as follows: pixels with wind speeds greater than 10 m / s are removed; pixels with Faraday rotation angles greater than 10° are removed; pixels with polarization rotation angles between 42° and 48° are removed.

3. The method according to claim 1, characterized in that: The effective data snapshot is a data snapshot in which the percentage of effective pixels exceeds 50%.

4. The method according to claim 1, characterized in that: The deviation field OTT is generated in the following manner: The median of the deviation values ​​within each grid is taken; the root mean square error of the deviation values ​​within each grid is calculated to evaluate the quality of the grid data; grids with root mean square errors exceeding a preset threshold are identified as noisy grids, and a zero correction operation is performed, i.e., the deviation field value of the grid is set to 0; the median of the remaining grids excluding the noisy grids is taken as the deviation field OTT.

5. The method according to claim 4, characterized in that: Data with a root mean square error greater than or equal to 10K is classified as a noisy grid.

6. The method according to claim 1, characterized in that: The reference area includes the ascending rail area and the descending rail area.

7. The method according to claim 1, characterized in that: The reference area is a calm sea area with significant salinity variation, low radio frequency interference, and far from the coast; the significant salinity variation is defined as an annual average variation of 0.1-0.5 psu, the low radio frequency interference is defined as an RFI pollution rate of less than 10%, and far from the coast is defined as a distance from the shore greater than 500 km.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7.

9. A computer software product, characterized in that... include: Processors and storage devices; Storage device for storing one or more programs. When the one or more programs are executed by one or more processors, the one or more processors implement the brightness temperature correction method based on improved marine target transformation as described in any one of claims 1 to 7.

Images