Systematic bias correction method for improving the accuracy of salinity retrieval from interferometric microwave radiometer

Abstract

The present application relates to the system deviation correction method for improving the precision of interferometric microwave radiometer salinity inversion, and relates to the satellite ocean remote sensing data processing technical field, a specific calm sea area is demarcated and selected as a dynamic calibration reference area in global sea area, and the observation brightness temperature data of a preset time window in the area is strictly purified and selected to eliminate the influence of radio frequency interference, sun and moon pollution, etc., to collect high-purity original data, calculate simulated brightness temperature by using a radiation transfer model, and obtain brightness temperature deviation samples by subtracting the purified observation brightness temperature from the simulated brightness temperature, then a robust statistical method based on a sliding window and a median is used to statistically calculate the deviation samples to obtain stable and reliable deviation correction values, and finally a two-dimensional lookup table is generated and applied to sea surface salinity inversion. The present application can efficiently and accurately correct the observation brightness temperature used for salinity inversion from the source, and solves the contradiction between stability and adaptability in the prior art.

Description

Technical Field

[0001] This invention belongs to the field of satellite ocean remote sensing data processing technology, specifically, it relates to a method for correcting system biases to improve the accuracy of salinity inversion using interferometric microwave radiometers. Background Technology

[0002] Sea surface salinity (SSS) is a key physical parameter influencing ocean dynamics and global thermohaline circulation. Many phenomena and processes occurring in the ocean are closely related to its distribution and changes. Therefore, studying the distribution and variation patterns of sea surface salinity has irreplaceable application value for improving the predictive capabilities of ocean and climate models, monitoring extreme weather events, and conducting marine scientific research.

[0003] Satellite remote sensing has become a core pillar of sea surface salinity monitoring. Compared with traditional in-situ measurements relying on ships and buoys, satellite microwave remote sensing technology, especially L-band (approximately 1.4 GHz) microwave radiometers, can penetrate clouds and achieve all-weather, large-scale, and high-frequency global sea surface salinity observations.

[0004] However, retrieving sea surface salinity using microwave radiometers faces a fundamental technical challenge: the underlying physical quantity measured—brightness temperature (TB)—is extremely insensitive to salinity changes. For every 1 practical salinity unit (PSU) change in salinity, the brightness temperature changes by only about 0.5 K. This weak signal response makes observed brightness temperature highly susceptible to noise interference from various sources. Any tiny deviation can be amplified into a huge salinity error, thus obscuring the true ocean salinity signal. These deviations mainly originate from uncertainties in geophysical models and contamination from external environmental signals. Currently, existing bias correction schemes either perform large-scale statistical correction using sparse in-situ data after salinity products are generated, failing to address the quality issue of the input brightness temperature at its source; or they rely on complex numerical weather prediction (NWP) data assimilation systems, which are computationally expensive and difficult to deploy independently. These methods struggle to balance stability and adaptability to time-varying biases, and the reference benchmarks used for statistical bias are susceptible to contamination from complex global sea conditions and noise, resulting in insufficient correction accuracy and robustness. Therefore, employing a high-precision, highly stable, and computationally efficient method to correct the bias in the original observed brightness temperature before salinity inversion is of significant technical importance and application value for improving the overall accuracy and reliability of sea surface salinity products. Summary of the Invention

[0005] The purpose of this invention is to address the aforementioned problems in the prior art by proposing a brightness temperature deviation correction method based on a stable reference region and robust statistics. This aims to provide a high-precision, high-stability, and computationally efficient solution. The core of this solution lies in establishing a stable and pure deviation correction benchmark and generating a quasi-dynamic correction lookup table through robust statistical methods.

[0006] This invention proposes a systematic bias correction method to improve the accuracy of salinity inversion in interferometric microwave radiometers, comprising the following steps:

[0007] S1. Select a stable calibration reference area for the interferometric microwave radiometer;

[0008] S2. Collect the observed brightness temperature data within the preset time window of the reference area;

[0009] S3. The observed brightness temperature data is purified and filtered to obtain a valid dataset;

[0010] S4. Calculate the simulated brightness temperature corresponding to each observation point in the effective dataset using the microwave radiative transfer model and subtract it from the observed brightness temperature to obtain a brightness temperature deviation sample.

[0011] S5. Perform statistical calculations on the deviation samples to obtain the final deviation correction value;

[0012] S6. Store the final deviation correction value described in S5 into a two-dimensional lookup table;

[0013] S7. When performing sea surface salinity inversion, the final deviation correction value is obtained from the two-dimensional lookup table, the observed brightness temperature is corrected, and then the sea surface salinity inversion is performed.

[0014] Preferably, in step S1, the selection of the stable calibration reference area is based on the following four conditions:

[0015] The marine dynamic environment is stable, and the temporal and spatial variations of sea surface temperature and salinity are gradual.

[0016] Located far from land and areas of human activity, with low risk of radio frequency interference and land pollution;

[0017] Weather conditions are uniform;

[0018] The shape of the region is similar to the orbital path of satellite observation.

[0019] Preferably, in step S2, the preset time window is a sliding time window centered on the target correction date, and the sliding time window covers a total data period of 10 days.

[0020] Preferably, in step S5, the geographic coordinates are converted into cosine coordinates of the center observation direction using the observation point information, and then mapped to a two-dimensional coordinate system. Index; within the sliding time window, for each The index is used to calculate the median of all corresponding brightness temperature deviation samples, and the median is used as the final deviation correction value.

[0021] Preferably, the geographic coordinates of the instrument's field of view center point are discretized into... The indexing process includes the following steps: First, based on the satellite's real-time position and velocity vectors and the instrument's pointing angle, a satellite platform coordinate system and an instrument coordinate system are established as references. Next, the geographic coordinates of the field of view's center point are converted to geocentric coordinates and then transformed to the instrument coordinate system to obtain the field of view direction vector. Subsequently, the direction cosine of this direction vector in the instrument coordinate system is calculated. Finally, through normalization, the direction cosine is mapped to a preset grid. index.

[0022] Preferably, the steps for establishing the satellite platform coordinate system and the instrument coordinate system are as follows: using the satellite position vector and velocity vector Define platform coordinate basis vector , , :

[0023]

[0024]

[0025]

[0026] Using the instrument's pitch angle Rotation angle Generate instrument coordinate system rotation matrix :

[0027] .

[0028] Preferably, the step of transforming the geographic coordinates of the field of view center point to the instrument coordinate system specifically involves: first, converting the latitude of the geographic coordinates to the instrument coordinate system. Longitude is Converting ground points to ECEF coordinate vectors :

[0029]

[0030] Where R is the Earth's radius;

[0031] The field of view direction vector in the instrument coordinate system is then obtained using the following formula. :

[0032]

[0033] , and These are the field of view direction vectors. Projected components on the x-axis, y-axis, and z-axis of the instrument coordinate system.

[0034] Preferably, the steps for calculating the direction cosine are as follows:

[0035]

[0036]

[0037] in, , , express The cosine of the angle between the instrument coordinate system and the x-axis. express The cosine of the angle between the instrument coordinate system and the y-axis. Zenith angle, representing The angle between the instrument coordinate system and the positive z-axis. Azimuth, representing The angle between the projection onto the xy plane and the positive x-axis.

[0038] Preferably, the direction cosine is mapped to an index. The specific steps are as follows: First, determine the maximum value of the direction cosine within the calibration reference area ( , ) and minimum value ( , Within this range, a 129×129 standard grid is generated to store the final deviation correction value; The index is calculated using the following formula:

[0039] ;

[0040] .

[0041] Preferably, in step S7, during the application stage of sea surface salinity inversion, the direction cosine coordinates of each observation point need to be calculated first; then, the coordinates are used to perform linear interpolation on the two-dimensional lookup table to accurately obtain the final deviation correction value of the coordinate position, and this interpolation result is subtracted from the original observed brightness temperature to obtain the brightness temperature data after systematic deviation correction.

[0042] The beneficial effects of this invention are as follows:

[0043] The purpose of this invention is to solve the problems existing in the prior art and to propose a high-precision and high-stability brightness temperature deviation correction method for L-band microwave radiometers. It specifically utilizes the characteristics of specific calm sea areas as natural calibration fields and generates a quasi-dynamic deviation correction lookup table through an innovative calibration scheme based on a stable reference area and robust statistics. This achieves efficient, accurate and stable correction of observed brightness temperature, thereby significantly improving the accuracy and reliability of sea surface salinity inversion products. Attached Figure Description

[0044] Figure 1 This is a schematic flowchart of the brightness temperature deviation correction method for improving longitude inversion based on sea surface salinity according to the present invention.

[0045] Figure 2 This is a schematic diagram of the deviation correction values ​​of the present invention;

[0046] Figure 3 This is a schematic diagram illustrating the correction effect of the present invention. Detailed Implementation

[0047] To further understand the present invention, it will be further described below with reference to the accompanying drawings and embodiments, and the technical solutions of the present invention will be clearly and completely described. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments. This embodiment proposes a brightness temperature deviation correction method based on a stable reference zone and robust statistics, and applies it to sea surface salinity inversion to verify its effectiveness.

[0049] like Figure 1 The diagram shown is a schematic flowchart of the brightness temperature deviation correction method for improving longitude inversion based on sea surface salinity according to the present invention, which includes the following steps:

[0050] S1. Select a stable calibration reference area for the interferometric microwave radiometer.

[0051] The brightness temperature data used in this embodiment is the SMOSL1C grade product, and the auxiliary geophysical parameters (such as sea surface temperature, wind field, etc.) are from the ECMWF auxiliary dataset released in conjunction with SMOS data.

[0052] Before the experiment began, a large and relatively calm quadrilateral area was selected near the center of the South Pacific subtropical circulation as the reference area for dynamic calibration, based on the four conditions for the stable calibration reference area proposed in this invention.

[0053] The selection of a stable calibration reference area is based on the following four conditions:

[0054] The marine dynamic environment is stable, and the temporal and spatial variations of sea surface temperature and salinity are gradual.

[0055] Located far from land and areas of human activity, with low risk of radio frequency interference and land pollution;

[0056] Weather conditions are uniform;

[0057] The shape of the region is similar to the orbital path of satellite observation.

[0058] S2. Collect the observed brightness temperature data within the preset time window in the reference area and perform data preprocessing.

[0059] This step aims to collect raw data within the calibration reference area, with the preset time window being a sliding time window centered on the target calibration date. Specifically, a 10-day sliding time window (D-5 to D+5) is constructed centered on any target calibration date (D). All SMOSL1C brightness temperature observation data within this window that are located within the calibration reference area are collected.

[0060] S3. The observed brightness temperature data is purified and filtered to obtain a valid dataset.

[0061] Any geographical sampling points that do not meet the quality requirements will be removed. Removal criteria include: 1) External signal interference: radio frequency interference, solar or lunar contamination; 2) Adverse geophysical conditions: rainfall, sea ice coverage, extreme wind speeds (too high or too low); 3) Data defects: instrument errors, missing key auxiliary data, or insufficient amount of effective brightness temperature data.

[0062] After the above filtering, a high-purity original dataset is obtained, denoted as dataset A.

[0063] S4. Calculate the simulated brightness temperature corresponding to each observation point in the effective dataset using the microwave radiative transfer model and subtract it from the observed brightness temperature to obtain the brightness temperature deviation ΔTB sample.

[0064] Using the radiative transfer model, the simulated brightness temperature is calculated for each valid observation point in dataset A. Then, the difference between the observed brightness temperature and the simulated brightness temperature is used to obtain the brightness temperature deviation ΔTB sample for that point.

[0065] S5. Perform robust statistics on the biased samples to obtain the final bias correction value.

[0066] To ensure the stability and reliability of the correction values, this step employs a robust statistical method.

[0067] Using observation point information, geographic coordinates are converted into cosine coordinates of the central observation direction, and then mapped to two dimensions. Index; within the sliding time window, for each The index is used to calculate the median of all corresponding brightness temperature deviation samples, and the median is used as the final deviation correction value.

[0068] Discretize the geographic coordinates of the instrument's field of view center point as follows: The indexing process includes the following steps: First, based on the satellite's real-time position and velocity vectors and the instrument's pointing angle, a satellite platform coordinate system and an instrument coordinate system are established as references. Next, the geographic coordinates of the field of view's center point are converted to geocentric coordinates and then transformed to the instrument coordinate system to obtain the field of view direction vector. Subsequently, the direction cosine of this direction vector in the instrument coordinate system is calculated. Finally, through normalization, the direction cosine is mapped to a preset grid. index.

[0069] The specific steps for establishing the satellite platform coordinate system and the instrument coordinate system are as follows: using the satellite position vector and velocity vector Define platform coordinate basis vector , , :

[0070]

[0071]

[0072]

[0073] Using the instrument's pitch angle Rotation angle Generate instrument coordinate system rotation matrix :

[0074] .

[0075] The specific steps for converting the geographic coordinates of the field of view center point to the instrument coordinate system are as follows: First, convert the latitude... Longitude is Converting ground points to ECEF coordinate vectors :

[0076]

[0077] Where R is the Earth's radius;

[0078] The field of view direction vector in the instrument coordinate system is then obtained using the following formula. That is, the direction vector from the satellite to the center point of the observation field of view:

[0079]

[0080] , and These are the field of view direction vectors. Projected components on the x-axis, y-axis, and z-axis of the instrument coordinate system.

[0081] Calculate direction cosine The specific steps are as follows:

[0082]

[0083]

[0084] in, express The cosine of the angle between the instrument coordinate system and the x-axis. express The cosine of the angle between the instrument coordinate system and the y-axis; Its physical meaning is the straight-line distance between the center point of the satellite and the ground observation field of view; The zenith angle represents... The angle between the instrument coordinate system and the positive z-axis. The azimuth angle represents the... The angle between the projection onto the xy-plane and the positive x-axis can be expressed by the formula... , Calculated.

[0085] Map direction cosine to index The specific steps are as follows: First, determine the maximum value of the direction cosine within the calibration reference area ( , ) and minimum value ( , Within this range, a 129×129 standard grid is generated to store the final deviation correction value; The index is calculated using the following formula:

[0086] ;

[0087] .

[0088] Specifically, all ΔTB samples obtained within a 10-day sliding window are mapped onto a regular two-dimensional (x,y) grid using a direction cosine transform. Then, each grid cell is calculated... The median of all samples within the range is used to determine the final deviation correction value for the target day (D). A diagram illustrating the final deviation correction value is shown below. Figure 2 As shown.

[0089] S6. Store the final deviation correction value in a two-dimensional lookup table.

[0090] The calculated deviation correction value is used in the grid cells of step 5. Use it as an index and store it in a two-dimensional lookup table.

[0091] S7. When performing sea surface salinity inversion, the final deviation correction value is obtained from the two-dimensional lookup table, the observed brightness temperature is corrected, and then the sea surface salinity inversion is performed.

[0092] The lookup table generated in step S6 is applied to the global sea surface salinity inversion process to verify its correction effect. Before performing salinity inversion on any observed brightness temperature, the corresponding bias correction value is obtained by interpolation from the lookup table, and this value is subtracted from the observed brightness temperature. The corrected brightness temperature is then used as input to perform sea surface salinity inversion, thereby obtaining a more accurate sea surface salinity value.

[0093] like Figure 3 As shown, (a) is the sea surface salinity map retrieved using the original brightness temperature data, (b) is the sea surface salinity map retrieved after bias correction using the method of this invention, and (c) is a schematic diagram of the sea surface salinity difference before and after correction. The comparison shows that the corrected sea surface salinity field has a clearer structure in the main ocean current areas, eliminates the previously existing large-scale artifacts and biases, and better matches known oceanographic characteristics, proving the effectiveness of the method of this invention.

[0094] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, alterations, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for correcting systematic biases to improve the accuracy of salinity inversion in interferometric microwave radiometers, characterized in that, Includes the following steps: S1. Select a stable calibration reference area for the interferometric microwave radiometer; S2. Collect the observed brightness temperature data within the preset time window of the reference area; S3. Clean and filter the observed brightness temperature data to obtain a valid dataset; S4. Calculate the simulated brightness temperature corresponding to each observation point in the effective data using the microwave radiative transfer model and subtract it from the observed brightness temperature to obtain the brightness temperature deviation sample. S5. Perform statistical calculations on the deviation samples to obtain the final deviation correction value; Using observation point information, geographic coordinates are converted into cosine coordinates of the central observation direction, and then mapped to two dimensions. Introducing, within the sliding time window, for each Index, calculate the median of all corresponding brightness temperature deviation samples, and use the median as the final deviation correction value; Discretize the geographic coordinates of the instrument's field of view center point as follows: The indexing process includes the following steps: First, based on the satellite's real-time position and velocity vectors and the instrument's pointing angle, a satellite platform coordinate system and an instrument coordinate system are established as references. Next, the geographic coordinates of the field of view's center point are converted to geocentric coordinates and then transformed to the instrument coordinate system to obtain the field of view direction vector. Subsequently, the direction cosine of this direction vector in the instrument coordinate system is calculated. Finally, through normalization, the direction cosine is mapped to a preset grid. Index; Map direction cosines to index The specific steps are as follows: First, determine the maximum value of the direction cosine within the calibration reference area. and minimum value Within this range, a standard grid of 129×129 is generated to store the final deviation correction value; The index is calculated using the following formula: The specific steps for calculating the direction cosine are as follows: in, Represents the field of view direction vector The cosine of the angle between the instrument coordinate system and the x-axis. express The cosine of the angle between the y-axis and the instrument coordinate system, the zenith angle. represent The angle between the azimuth and the positive z-axis of the instrument coordinate system. represent The angle between the projection onto the xy plane and the positive x-axis; These are the field of view direction vectors. Projected components on the x-axis, y-axis and z-axis of the instrument coordinate system; S6. Store the final deviation correction value of S5 into a two-dimensional lookup table; S7. When performing sea surface salinity inversion, obtain the final deviation correction value from the two-dimensional lookup table, correct the observed brightness temperature, and then perform sea surface salinity inversion.

2. The method for correcting system bias to improve the accuracy of salinity inversion in an interferometric microwave radiometer according to claim 1, characterized in that, In step S1, the selection of the stable calibration reference area is based on the following four conditions: stable marine dynamic environment, and gentle spatiotemporal changes in sea surface temperature and salinity. Located far from land and areas of human activity, with low risk of radio frequency interference and land pollution; Weather conditions are uniform; The shape of the region is similar to the orbital path of satellite observation.

3. The method for correcting systematic biases to improve the accuracy of salinity inversion in an interferometric microwave radiometer according to claim 1, characterized in that, In step S2, the preset time window is a sliding time window centered on the target correction date, and the sliding time window covers a total data period of 10 days.

4. The method for correcting system bias to improve the accuracy of salinity inversion in an interferometric microwave radiometer according to claim 1, characterized in that, The specific steps for establishing the satellite platform coordinate system and the instrument coordinate system are as follows: using the satellite position vector and velocity vector Define platform coordinate basis vector Using the instrument's pitch angle Rotation angle Generate instrument coordinate system rotation matrix 5. The method for correcting system bias to improve the accuracy of salinity inversion in an interferometric microwave radiometer according to claim 4, characterized in that, The specific steps for converting the geographic coordinates of the field of view center point to the instrument coordinate system are as follows: First, convert the latitude... Longitude is Converting ground points to ECEF coordinate vectors Where R is the Earth's radius; The field of view direction vector in the instrument coordinate system is then obtained using the following formula.

6. The method for correcting system biases to improve the accuracy of salinity inversion in an interferometric microwave radiometer according to claim 1, characterized in that, In step S7, during the application stage of sea surface salinity inversion, the direction cosine coordinates of each observation point need to be calculated first. Then, the two-dimensional lookup table is linearly interpolated using the direction cosine coordinates to accurately obtain the final deviation correction value of the coordinate position. This interpolation result is subtracted from the original observed brightness temperature to obtain the brightness temperature data after systematic deviation correction.

Images