Method for identifying and eliminating internal interference of satellite-borne microwave radiometer
By using an interference identification method based on two-dimensional features between scan lines and observation points, internal interference of the spaceborne microwave radiometer is identified and replaced, solving the problem that external interference detection is not applicable in existing technologies, and improving data quality and inversion accuracy.
Patent Information
- Application Number
- CN202511695835.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-19
- Publication Date
- 2026-02-13
AI Technical Summary
Existing interference detection methods for spaceborne microwave radiometers mainly target external electromagnetic radiation, making it difficult to effectively identify and eliminate internal interference. This leads to a decline in data quality and affects the accurate inversion of physical parameters of the Earth's surface and atmosphere.
An interference identification method based on two-dimensional features between scan lines and observation points is adopted. By calculating the data change rate and slope features, the scan line data containing interference is identified and replaced. The interference is judged by the gain coefficient and threshold, thereby eliminating internal interference.
This improved the consistency between the measurement results of the spaceborne microwave radiometer and the actual radiation characteristics, ensuring data quality and supporting the accuracy of subsequent scientific analysis.
Smart Images
Figure CN121521276A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of remote sensing information technology, specifically relating to a method for identifying and eliminating internal interference in a spaceborne microwave radiometer. Background Technology
[0002] A spaceborne microwave radiometer is a remote sensing device carried on a satellite that receives microwave radiation signals naturally emitted by the Earth and its atmosphere to retrieve physical parameters of the Earth's surface and atmosphere. Unaffected by day / night cycles, cloud cover, or other weather conditions, it can perform all-weather, all-time observations, and has important applications in meteorology, oceanography, hydrology, and agriculture. Taking the Fengyun-3E satellite as an example, its microwave thermometer (FY-3E / MWTS-III) is a spaceborne microwave radiometer operating in the 23.8 GHz, 31.4 GHz, and 50–60 GHz frequency bands. It can observe the vertical temperature distribution of the atmosphere and other global meteorological data, thereby enabling medium- and long-term numerical weather prediction and improving the accuracy of weather forecasts. The spaceborne microwave radiometer receives microwave radiation energy from the Earth's surface and atmosphere through a highly sensitive antenna and receiver, converting it into electrical signals and recording them. Interference during the reception and transmission of these electrical signals can distort the signals received by the microwave radiometer, severely reducing data quality and affecting the accurate retrieval of physical parameters of the Earth's surface and atmosphere, thus impacting subsequent scientific analysis and applications.
[0003] Existing methods for detecting and eliminating interference from spaceborne microwave radiometers target electromagnetic radiation from outside the satellite, including uplink interference from ground radar or communication equipment. Detection methods include time-domain methods, frequency-domain methods, and statistical methods.
[0004] The time-domain method is suitable for pulsed radiative interference. It compares each sampling point in the data stream with a threshold, and if the value exceeds the threshold, it is considered to be affected by interference. By selecting an appropriate threshold, it has a good detection effect on interference with large magnitude or long duration.
[0005] The frequency domain method utilizes the trend of brightness temperature variation with frequency in different channels to detect the intensity and range of interference through spectral difference. After identifying the interference signal, it corrects abnormal observations through linear fitting. The frequency domain method is affected by natural spectral difference fluctuations in different regions / seasons, requiring the use of brightness temperature observation data to fit the spectral difference threshold. This method can be applied to the C and X bands of AMSR-E.
[0006] Statistical methods include the mean and standard deviation method and the higher-order moment method. The mean and standard deviation method extends the spectral difference threshold of the frequency domain method, using the statistical mean and standard deviation over the entire year as the detection basis. It obtains a matrix by accumulating and statistically analyzing interference indices in a global grid; if the mean and standard deviation at a certain location are large, interference is considered to exist at that location. This method requires a large amount of data accumulation. The higher-order moment method detects interference based on the difference between the statistical characteristics of natural microwave radiation and interference signals. Natural microwave radiation typically follows a Gaussian distribution, and its statistical characteristics are stable. Interference signals disrupt these statistical characteristics, manifesting as changes in higher-order moments (skewness / kurtosis). By constructing a statistical test quantity and comparing it with a reference distribution in an interference-free scenario, data points deviating from normal statistical characteristics are identified. The advantage of the higher-order moment method is that it does not depend on the characteristics of a specific interference source; its limitation is that it is insensitive to weak interference: if the interference power is close to the level of natural radiated noise, the change in statistical characteristics is not significant.
[0007] The existing methods for detecting interference with spaceborne microwave radiometers all fall under the category of external interference detection. However, electromagnetic coupling between various payloads and equipment on the satellite can also introduce internal interference. Taking the Fengyun-3 05 satellite as an example, it carries more than ten payloads, including a global navigation satellite occultation detector, an infrared hyperspectral atmospheric sounder, a medium-resolution spectral imager, a microwave hygrometer, a microwave thermometer (FY-3E / MWTS-III), a multi-angle ionospheric photometer, and a wind field measurement radar. Based on their working principles, these payloads can be categorized into active payloads (actively emitting electromagnetic signals) and passive payloads (only receiving external electromagnetic signals). The wide frequency band coverage and significant differences in their operating modes directly contribute to a complex electromagnetic environment. If electromagnetic interference is not effectively controlled, it will directly lead to distortion of payload detection data, affecting the accuracy of meteorological observations. Existing external interference detection methods are difficult to apply to internal interference. For example, the microwave thermometer (FY-3E / MWTS-III) has detection frequency bands of 23.8 GHz, 31.4 GHz, and 50–60 GHz, making frequency domain methods and statistical methods using averages and standard deviations unsuitable. Since the data covers the land-sea boundary, the natural microwave radiation deviates from the Gaussian distribution, so the higher-order moment method in statistical methods is not applicable. Summary of the Invention
[0008] In order to overcome the shortcomings of the existing technology, the purpose of this invention is to provide a method for identifying and eliminating internal interference in a spaceborne microwave radiometer. This method is applied to the radiometric calibration product processing steps of a spaceborne microwave radiometer, thereby improving the consistency between the measurement results and the actual radiation characteristics.
[0009] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for identifying and eliminating internal interference in a spaceborne microwave radiometer includes the following steps; Step 1: Data preparation; Obtain the ground observation voltage value, heat source observation voltage value, and cold source observation voltage value of each channel of the microwave radiometer, and input the coefficient matrix; Step 2: Use the input coefficient matrix as input to perform scan line marking; Step 3: Identify the markings between scan lines to determine if there is any interference; Step 4: Replace the scan lines that are causing interference to alleviate the interference.
[0010] In step 1, the ground observation voltage value, heat source observation voltage value, and cold source observation voltage value are the original observation results from the microwave radiometer.
[0011] The input coefficient matrix includes gain, gain coefficient indicating rapid changes in a single-point comparison threshold, slope of the interference line (observation point / scan line), threshold for determining all present, and threshold for determining all absent.
[0012] Step 2 specifically involves: Step 2-1: Calculate the large and small signs Calculate the rate of change of the data: Change Change=abs(D(ii+3)+D(ii+2)-D(ii-3)-D(ii-2)) Where D(ii) represents the voltage value at the observation point on the ii-th scan line, and abs represents the absolute value; If Change < threshold S (S is obtained from the input coefficient matrix), according to the characteristics of the sawtooth pattern, if the value of a certain point is greater than the values of its four adjacent points and the points between adjacent points, then its large value is set to 1; if the value of a certain point is smaller than the values of its four adjacent points and the points between adjacent points, then its small value is set to 1. If Change ≥ threshold S (S is obtained from the input coefficient matrix), then only neighboring points are compared; Step 2-2: Mark the serrations: If a scan line is marked as a large mark and the next scan line is marked as a small mark, then the scan line is marked as a serrated line. Steps 2-3: Combined Identification Serrated Edges: If three consecutive zigzag marks are marked, then the combined zigzag mark of these three marks is marked, as are the combined zigzag marks of the two preceding and two following marks.
[0013] Step 3 specifically involves: Step 3-1: Accumulate diagonally: Based on the positional pattern of the interference cycle change, the cumulative number of sawtooth teeth on the diagonal line is SumY, and the number of points on each diagonal line is SumA. The slope of the diagonal line is obtained from the input coefficient matrix. The interference is generated by radar, and its period is twice the observation period of the microwave radiometer. Step 3-2: Determining the diagonal line: If the ratio of the number of serrations SumY to the number of points SumA on each diagonal line is higher than the threshold (the threshold for determining all lines, i.e., SumY / SumA), then the line is considered to have interference. Step 3-3: Diagonal marking: If the diagonal line is marked as interference, then all positions of the diagonal line are considered to contain interference.
[0014] Step 4 specifically involves: For the first scan line: if it is marked as interfering, use the data from the second scan line instead; for the last scan line: if it is marked as interfering, use the data from the second-to-last scan line instead; for scan lines 2 to the second-to-last: if it is marked as interfering, use the average value of the preceding and following scan lines instead.
[0015] The voltage values of the ground observation points after interference removal and the voltage values of the cold and heat sources are input into the calibration program (i.e., the process of voltage to brightness temperature correspondence) to obtain the observed brightness temperature value. The difference between the observed and simulated brightness temperature values is obtained by subtracting the observed and simulated brightness temperature values after interference removal.
[0016] Output Earth observation interference flag, heat source observation interference flag, cold source observation interference flag, Earth observation interference flag after interference removal, heat source observation interference flag after interference removal, and cold source observation interference flag after interference removal.
[0017] The beneficial effects of this invention are: This invention addresses the problem of internal interference affecting spaceborne microwave radiometers, which degrades data quality and makes it difficult to accurately retrieve Earth's surface and atmospheric physical parameters. The invention utilizes the sawtooth-like characteristics of interference between scan lines and employs two-dimensional features based on the scan lines and observation points (one dimension between scan lines and another between observation points) to identify and remove interference. This allows for the reconstruction of Earth observations and cold / hot source voltages after interference removal, which is of great significance for the calibration and data preprocessing of spaceborne microwave radiometers.
[0018] By utilizing the interference identification and elimination voltages of Earth observation and hot / cold source observations obtained in this invention, radiation calibration products are processed, ensuring the consistency between the measurement results of the spaceborne microwave radiometer and the actual radiation characteristics. Attached Figure Description
[0019] Figure 1 The simulated brightness temperature difference is shown in the image, which is affected by internal interference. The locations circled in red are two typical sources of interference.
[0020] Figure 2 This is a flowchart of the interference identification and removal method.
[0021] Figure 3 A schematic diagram for calculating the large marker (the high point of the serration) and the small marker (the low point of the serration).
[0022] Figure 4 This is a diagram illustrating the serrated edge (triangular FLAG).
[0023] Figure 5 The combined zigzag pattern is obtained after traversing all scan lines. Black represents marked interfering observation points, and white represents unmarked observation points.
[0024] Figure 6 These are interference markers after the observation points have been identified. White indicates marked interfering observation points, and black indicates unmarked observation points.
[0025] Figure 7 shows a comparison before and after interference removal. (a) Observed simulated brightness temperature difference before interference removal; (b) Observed simulated brightness temperature difference after interference removal; (c) The interference removed. Detailed Implementation
[0026] The present invention will now be described in further detail with reference to the accompanying drawings.
[0027] This invention is a method for identifying and eliminating internal interference in a spaceborne microwave radiometer based on two-dimensional features between scan lines and observation points. The method is illustrated using the FY-3E microwave thermometer as an example.
[0028] Figure 1 The simulated brightness-temperature difference caused by interference is shown. The horizontal axis represents observation points 1 to 98, and the vertical axis represents scan lines 1 to 1276. The location circled in red indicates a typical internal interference. The interference manifests as a relatively large value on one scan line and a relatively small value on the next scan line, alternating between them.
[0029] like Figure 2 As shown, a method for identifying and eliminating internal interference in a spaceborne microwave radiometer based on two-dimensional features between scan lines and observation points includes the following steps; Step 1: Data Preparation: Obtain the ground observation voltage value, heat source observation voltage value, and cold source observation voltage value of each channel of the microwave radiometer. Input the coefficient matrix, which includes the gain, the gain coefficient of the single-point comparison threshold, the gain coefficient of the rapidly changing gain coefficient, the slope of the interference line (observation point / scan line), the threshold for determining all presence, and the threshold for determining all absence.
[0030] The purpose of this step is to provide basic data and key parameters to support subsequent interference identification and processing. Its effect is to lay the data foundation for completing interference processing, ensuring that step 2 has a clear analysis object and judgment criteria, and providing a fundamental guarantee for the accuracy of interference identification.
[0031] Step 2, Sub-module 1: Scan Line Identification. The purpose of this step is to identify interference points with "sawtooth" characteristics by analyzing the voltage value change characteristics between scan lines. The effect is to initially locate sawtooth interference from the scan line dimension, providing a foundation for Step 3.
[0032] Step 2-1: Calculate the large and small signs: Calculate the rate of change of the data: Change Change=abs(D(ii+3)+D(ii+2)-D(ii-3)-D(ii-2)) Where D(ii) represents the voltage value at the observation point on the ii-th scan line, and abs represents the absolute value; If Change < threshold S (S is obtained from the input coefficient matrix), according to the characteristics of the sawtooth pattern, if the value of a certain point is greater than the values of its four adjacent points, the points between those adjacent points, and the points at intervals between those adjacent points, then its larger value flag is set to 1. For example... Figure 3 The image shows the voltage values of 250 scan lines, with the high point (*) of the sawtooth pattern marked as 1. If the value of a point is smaller than the values of its four adjacent points and the points between them, its low point is marked as 1, which corresponds to the low point (o) of the sawtooth pattern in the image.
[0033] If Change ≥ threshold S (S is obtained from the input coefficient matrix), then only neighboring points are compared.
[0034] Step 2-2: Mark the serrations: If a scan line is marked with a large mark and the next scan line is marked with a small mark, then that scan line is marked as jagged. Figure 4 A triangle.
[0035] Steps 2-3: Combined Identification Serrated Edges: If three consecutive zigzag marks are marked, then the combined zigzag mark of these three marks is marked, as are the combined zigzag marks of the two preceding and two following marks.
[0036] Figure 5 The combined zigzag pattern is obtained after traversing all scan lines. White represents marked interfering observation points, and black represents unmarked observation points.
[0037] Step 3: Submodule 2: Inter-observation point identification. Based on the periodic location pattern of interference, systematic interference with diagonal distribution characteristics is identified from a global perspective of the observation points. Effect: This supplements the shortcomings of local identification between scan lines, identifying periodically distributed diagonal interference from a global perspective, making interference labeling more comprehensive.
[0038] Step 3-1: Accumulate diagonally: Based on the positional pattern of the interference cycle, the cumulative number of sawtooth teeth on the diagonal lines is SumY, and the number of points on each diagonal line is SumA. The slope of the diagonal lines is obtained from the input coefficient matrix.
[0039] Step 3-2: Determining the diagonal line: If the ratio of the number of serrations (SumY) to the number of individual diagonal points (SumA) on a diagonal line is higher than a threshold (the threshold is obtained from the input coefficient matrix), then the line is considered to have interference. The core characteristic of interference is that interference points appear densely along a diagonal line with a certain slope, rather than existing in isolation. Through the dual constraints of "diagonal line distribution characteristics + proportional threshold," the identification upgrade from "local interference points" to "global systemic interference" is achieved—utilizing the periodic distribution pattern of interference (diagonal lines) and eliminating the influence of random noise through statistical proportions, ultimately accurately locating the systemic interference area that needs to be processed.
[0040] Step 3-3: Diagonal marking: If the diagonal line is marked as interference, then all positions of the diagonal line are considered to contain interference.
[0041] Step 4: Submodule 3: Interference Mitigation. The aim is to specifically eliminate identified interference and restore the continuity and accuracy of the data. The effect is that scan line data marked as interference are effectively replaced, and the processed data is closer to the true observations, providing reliable input for subsequent brightness temperature inversion.
[0042] For scan line 1: If it is marked as interfering, the data from scan line 2 is used instead. For the last scan line: If it is marked as interfering, the data from the second-to-last scan line is used instead. For scan lines 2 to the second-to-last: If it is marked as interfering, the average value of the preceding and following scan lines is used instead. The physical properties of the Earth's surface or atmosphere (such as temperature and humidity) do not change abruptly within a short distance (adjacent scan lines), and the valid data from adjacent scan lines can be used as a reasonable substitute for the interfering data. The voltage values of the Earth observation points and the voltage values of the cold and heat sources after removing the interference are substituted into the calibration program to obtain the observed brightness temperature value. The difference between the observed and simulated brightness temperature values is obtained by subtracting them from the simulated brightness temperature value after interference removal, as shown in Figure 7(b). The removed interference is shown in Figure 7(c). Figure 6 The horizontal axis represents 98 observation points, the vertical axis represents the scan lines, and white marks indicate interference locations.
[0043] Figure 7(a) shows the original observation data affected by interference, Figure 7(b) shows the observation data after interference removal, and Figure 7(c) shows the interference removed.
Claims
1. A method for identifying and eliminating internal interference in a spaceborne microwave radiometer, characterized in that, Includes the following steps; Step 1: Obtain the ground observation voltage value, heat source observation voltage value, and cold source observation voltage value of each channel of the microwave radiometer, and input the coefficient matrix; Step 2: Use the input coefficient matrix as input to perform scan line marking; Step 3: Identify the markings between scan lines to determine if there is any interference; Step 4: Replace the scan lines that are causing interference to alleviate the interference.
2. The method for identifying and eliminating internal interference in a spaceborne microwave radiometer according to claim 1, characterized in that, In step 1, the ground observation voltage value, heat source observation voltage value, and cold source observation voltage value are the original observation results from the microwave radiometer. The input coefficient matrix includes gain, gain coefficient indicating rapid changes in a single-point comparison threshold, slope of the interference line (observation point / scan line), threshold for determining all present, and threshold for determining all absent.
3. The method for identifying and eliminating internal interference in a spaceborne microwave radiometer according to claim 2, characterized in that, Step 2 specifically involves: Step 2-1: Calculate the rate of change of the data (Change). Change=abs(D(ii+3)+D(ii+2)-D(ii-3)-D(ii-2)) Where D(ii) represents the voltage value at the observation point on the ii-th scan line, and abs represents the absolute value; calculate the large and small flags; Step 2-2: If a scan line is marked as a large mark and the next scan line is marked as a small mark, then the scan line is marked as a serrated line. Steps 2-3: If three consecutively spaced zigzag marks are marked, then the combined zigzag mark of these three marks is marked, and the combined zigzag mark of the two preceding and two following intervals is also marked.
4. The method for identifying and eliminating internal interference in a spaceborne microwave radiometer according to claim 3, characterized in that, In step 2-1, if Change < threshold S (S is obtained from the input coefficient matrix), according to the characteristics of the sawtooth pattern, if the value of a certain point is greater than the values of four points (adjacent points and the interval point between adjacent points), then its large value is set to 1; if the value of a certain point is smaller than the values of four points (adjacent points and the interval point between adjacent points), then its small value is set to 1. S is obtained from the input coefficient matrix; If Change ≥ threshold S, then only the neighboring points are compared.
5. The method for identifying and eliminating internal interference in a spaceborne microwave radiometer according to claim 4, characterized in that, Step 3 specifically involves: Step 3-1: Based on the positional pattern of the interference cycle change, accumulate the number of sawtooth teeth on the diagonal line (SumY) and the number of points on each diagonal line (SumA). The slope of the diagonal line is obtained from the input coefficient matrix. Step 3-2: If the ratio of the number of serrations SumY to the number of points SumA on each diagonal line is higher than the threshold, the line is considered to have interference if the threshold is SumY / SumA. Step 3-3: If the diagonal line is marked as interference, then all positions of the diagonal line are considered to contain interference.
6. The method for identifying and eliminating internal interference in a spaceborne microwave radiometer according to claim 5, characterized in that, In step 3-1, the interference is generated by radar, and its period is twice the observation period of the microwave radiometer.
7. The method for identifying and eliminating internal interference in a spaceborne microwave radiometer according to claim 6, characterized in that, Step 4 specifically involves: For the first scan line: if it is marked as interfering, then use the data from the second scan line to replace it; for the last scan line: if it is marked as interfering, then use the data from the second to last scan line to replace it. For scan lines 2 to the second to last: if they are marked as interference, use the average value of the preceding and following scan lines instead.
8. The method for identifying and eliminating internal interference in a spaceborne microwave radiometer according to claim 7, characterized in that, The voltage values of the ground observation points after interference removal and the voltage values of the cold and heat sources are input into the calibration program (i.e., the process of voltage to brightness temperature correspondence) to obtain the observed brightness temperature value. The difference between the observed and simulated brightness temperature values is obtained by subtracting the observed and simulated brightness temperature values after interference removal.
9. The method for identifying and eliminating internal interference in a spaceborne microwave radiometer according to claim 7, characterized in that, Output Earth observation interference flag, heat source observation interference flag, cold source observation interference flag, Earth observation interference flag after interference removal, heat source observation interference flag after interference removal, and cold source observation interference flag after interference removal.