Method and device for correcting nonlinear deviation of backscattering coefficient

Abstract

The invention discloses a method and a device for correcting nonlinear deviation of a backscattering coefficient, relates to the technical field of active microwave remote sensing, and aims to solve the problem of poor measurement precision of the backscattering coefficient of a microwave scatterometer in the prior art. The method comprises the following steps: acquiring measurement data and truth value data of the satellite microwave scatterometer; performing data preprocessing on the measurement data and the truth value data to obtain an initial sample data set; constructing an azimuth angle uniform distribution calibration model data set by using a dynamic uniform distribution construction method and a random error elimination method based on the initial sample data set; correcting the back scattering coefficient nonlinear deviation of the satellite microwave scatterometer based on the azimuth angle uniform distribution calibration model data set; based on this, accurate correction of the non-linear deviation of the backscattering coefficient of the satellite microwave scatterometer is realized, and the measurement accuracy of the backscattering coefficient of the satellite microwave scatterometer is improved.

Description

Technical Field

[0001] The invention relates to the technical field of active microwave remote sensing, in particular to a method and a device for correcting nonlinear deviation of a backscattering coefficient. Background Art

[0002] Satellite microwave scatterometers are mainly used to measure the backscatter coefficient of the ocean surface, and then invert the wind speed and wind direction information on the sea surface. Therefore, the accuracy of the products of microwave scatterometers inverting sea surface wind speed and wind direction directly depends on the calibration accuracy of their backscatter coefficients. In order to ensure the quality and accuracy of satellite microwave scatterometer observation data, accurate calibration of its backscatter coefficient is a very necessary and indispensable business process. The existing on-orbit calibration technology of microwave scatterometers mainly achieves systematic deviation correction of microwave scatterometer measurements by constructing special ground-based microwave scatterometer active calibrator equipment and typical natural objects with known backscattering characteristics (including open oceans, tropical rainforests, polar ice, etc.).

[0003] In order to eliminate the influence of surface background noise, the signal forwarded by the microwave scatterometer active calibrator is generally higher than the scattering intensity of the land surface, and its simulated surface backscattering intensity is about 10dB. The backscattering intensity of the land surface is generally between -15dB and 5dB, the average value of the grassland surface is about -12dB, the average value of the Amazon rainforest is about -7dB, and the average value of the backscattering of polar ice is about -5dB. The backscattering coefficient of the ocean surface is related to the wind speed, and its value ranges from about -45dB to -10dB, and the average value is about -22dB. Therefore, the backscatter coefficient simulated by the active calibrator of the microwave scatterometer far exceeds the measurement range of the backscatter coefficient of the ocean surface; the backscatter characteristics of tropical rainforests and polar ice are relatively uniform and relatively stable, but the numerical values ​​of their backscatter coefficients are approximately in a small range between -10dB and 0dB. Although the numerical values ​​of their backscatter coefficients are more stable and more conducive to calibration analysis, the dynamic range of their own numerical values ​​of the backscatter coefficients is small and cannot cover the measurement range of the backscatter coefficient of the ocean surface. Microwave scatterometers are mainly used to observe the sea surface wind field. The backscatter analysis directly based on the ocean surface can just cover the dynamic range of the backscatter coefficient required for nonlinear deviation analysis. This is an advantage that the backscatter characteristics of active calibrators, rainforests and other targets do not have. Therefore, backscattering based on the ocean surface is more conducive to ensuring the validity and availability of nonlinear deviation analysis results. However, the backscatter coefficient of the ocean surface changes dynamically, and the random error of its measurement is significantly larger than the error of active calibrators and typical stable land natural objects. This is the biggest difficulty in calibration analysis and processing. Therefore, there is currently no technology for using the ocean surface to carry out nonlinear bias correction of microwave scatterometer observations.

[0004] In view of this, there is an urgent need to design a more advanced method for correcting the nonlinear deviation of the backscattering coefficient to solve the problem of poor measurement accuracy of the backscattering coefficient of the microwave scatterometer in the prior art. Summary of the invention

[0005] The present invention aims to provide a method and device for correcting the nonlinear deviation of backscattering coefficient. The method and device are based on the symmetrical characteristics of ocean surface backscattering, utilize the statistical principle of dynamic equalization elimination of random errors of large samples that are far beyond statistical significance, take uniform distribution of sample data as the principle, and based on dynamic uniform distribution technology, remove redundant sample data from the initial matching sample data set to construct a calibration model data set with uniform distribution in azimuth angle. Then, based on the calibration model data set with uniform distribution in azimuth angle, accurate correction of the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer is achieved, and the measurement accuracy of the backscattering coefficient of the satellite microwave scatterometer is improved. The problem of poor measurement accuracy of the backscattering coefficient of the microwave scatterometer in the prior art is solved.

[0006] In order to achieve the above object, the present invention provides the following technical solutions: In a first aspect, the present invention provides a method for correcting a nonlinear deviation of a backscatter coefficient, which is applied to a satellite microwave scatterometer and may include: Acquire measurement data and true value data for a satellite microwave scatterometer, wherein the measurement data is the measurement data of the satellite microwave scatterometer itself during its on-orbit operation, and the true value data is the true value data of the satellite microwave scatterometer during its on-orbit operation obtained by a preset data acquisition method; Performing data preprocessing on the measurement data and the true value data to obtain an initial sample data set; the data preprocessing at least includes geographic time-space matching and eliminating observation data under rainfall conditions; Based on the initial sample data set, a dynamic uniform distribution construction method and a random error elimination method are used to construct an azimuth uniform distribution calibration model data set; Based on the azimuthally uniformly distributed calibration model data set, the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer is corrected.

[0007] Preferably, the constructing of the azimuth uniform distribution calibration model data set based on the initial sample data set by using a dynamic uniform distribution construction method and a random error elimination method may include: The initial sample data set is preprocessed using a dynamic uniform distribution construction method to obtain a plurality of target azimuth interval calibration data sets with uniform distribution of azimuth intervals; the dynamic uniform distribution construction method at least includes a target wind speed interval division method and a target azimuth interval division method; The invalid wind speed data in a plurality of the target azimuth interval calibration data sets are eliminated by using a random error elimination method to obtain the azimuth uniform distribution calibration model data set.

[0008] Preferably, the method of using a dynamic uniform distribution construction method to perform data preprocessing on the initial sample data set to obtain a plurality of target azimuth interval calibration data sets uniformly distributed in azimuth intervals may include: Using the target wind speed interval division method to divide the sample data in the initial sample data set into wind speed intervals, to obtain a plurality of target wind speed interval data sets with wind speed intervals evenly distributed; The target azimuth interval division method is used to divide the multiple target wind speed interval data sets into azimuth intervals, so as to obtain multiple target azimuth interval calibration data sets with the azimuth intervals evenly distributed.

[0009] Preferably, the method of dividing the sample data in the initial sample data set into wind speed intervals by using the target wind speed interval division method to obtain a plurality of target wind speed interval data sets with wind speed intervals uniformly distributed may include: The sample data in the initial sample data set are divided into wind speed intervals according to a preset wind speed interval interval, so as to obtain a plurality of target wind speed interval data sets with the wind speed interval intervals being evenly distributed; wherein the preset wind speed interval interval is 2 m / s.

[0010] Preferably, the step of dividing the target wind speed interval data sets into azimuth intervals by the target azimuth interval division method to obtain a plurality of target azimuth interval calibration data sets evenly distributed in the azimuth intervals may include: The multiple target wind speed interval data sets are divided into wind speed intervals at preset azimuth intervals to obtain multiple target azimuth interval data sets that are evenly distributed in the azimuth intervals; wherein the minimum number of data in the multiple target azimuth interval data sets is greater than the preset number.

[0011] Preferably, the method of eliminating invalid wind speed data in a plurality of target azimuth interval calibration data sets by using a random error elimination method to obtain the azimuth uniform distribution calibration model data set may include: Taking the target azimuth interval calibration data set with the least data volume as a benchmark, a random error elimination method is used to randomly remove redundant data in other target azimuth interval calibration data sets, so that the data volumes in the multiple target azimuth interval calibration data sets are evenly distributed, thereby obtaining the azimuth evenly distributed calibration model data set.

[0012] Preferably, the correction of the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer based on the azimuthally uniformly distributed calibration model data set may include: Based on the azimuth uniform distribution calibration model data set, the backscatter coefficient of each azimuth corresponding interval data is averaged to obtain multiple target feature data; Based on the plurality of target characteristic data, a least squares method is used to fit the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer; The nonlinear deviation of the backscatter coefficient of the satellite microwave scatterometer is obtained according to the fitting, and the nonlinear deviation of the backscatter coefficient of the satellite microwave scatterometer is corrected.

[0013] Preferably, the measurement data may include at least a first backscatter coefficient, time, longitude, latitude, polarization beam, incident angle, azimuth, and quality indicator; The true value data at least includes a second backscatter coefficient, sea surface wind speed, sea surface wind direction, atmospheric cloud liquid water content, time, longitude and latitude; The method of obtaining the true value data of the satellite microwave scatterometer during the on-orbit operation by using a preset data acquisition method may include: Using an airborne microwave scatterometer with the same frequency as the satellite microwave scatterometer to carry out satellite-ground synchronous measurement to obtain a measured backscatter coefficient for the sea surface, and using the measured backscatter coefficient as the second backscatter coefficient; Alternatively, the sea surface wind speed and wind direction data of the sea surface observed by satellite are obtained from numerical model data, and combined with a preset geophysical model function, a calculated backscatter coefficient of the sea surface matching the satellite observation is simulated and calculated, and the calculated backscatter coefficient is used as the second backscatter coefficient.

[0014] Preferably, the performing data preprocessing on the measurement data and the true value data to obtain an initial sample data set may include: Performing geographic and temporal spatial matching on the measurement data and the true value data to obtain an intermediate sample data set; The measurement data under rainfall conditions in the intermediate sample data set are eliminated according to the atmospheric cloud liquid water content to obtain an initial sample data set.

[0015] In a second aspect, the present invention provides a device for correcting nonlinear deviation of backscatter coefficients, which is applied to a satellite microwave scatterometer and may include: An acquisition module, the acquisition module is used to acquire measurement data and true value data for a satellite microwave scatterometer, the measurement data is the measurement data of the satellite microwave scatterometer itself during its on-orbit operation, and the true value data is the true value data of the satellite microwave scatterometer during its on-orbit operation obtained by a preset data acquisition method; A preprocessing module, the preprocessing module is used to perform data preprocessing on the measurement data and the true value data to obtain an initial sample data set; the data preprocessing at least includes geographic time-space matching and eliminating observation data under rainfall conditions; A construction module, wherein the construction module is used to construct an azimuth uniform distribution calibration model data set based on the initial sample data set by using a dynamic uniform distribution construction method and a random error elimination method; A correction module is used to correct the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer based on the azimuth uniform distribution calibration model data set.

[0016] Compared with the prior art, the present invention provides a method for correcting the nonlinear deviation of a backscattering coefficient, which is applied to a satellite microwave scatterometer. The method obtains measurement data and true value data of the satellite microwave scatterometer, wherein the measurement data is the measurement data of the satellite microwave scatterometer itself during its on-orbit operation, and the true value data is the true value data of the satellite microwave scatterometer during its on-orbit operation obtained by a preset data acquisition method. The measurement data and the true value data are further subjected to data preprocessing to obtain an initial sample data set. The data preprocessing at least includes geographic time-space matching and elimination of observation data under rainfall conditions. Based on the initial sample data set, an azimuth uniform distribution calibration model data set is constructed by using a dynamic uniform distribution construction method and a random error elimination method. Finally, based on the azimuth uniform distribution calibration model data set, the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer is corrected. Based on this, the present invention fully considers the data characteristics of the ocean surface, constructs an initial sample data set with a large sample data volume, and uses a dynamic uniform distribution method and a random error elimination method to construct an azimuth uniform distribution calibration model data set, eliminating the influence of differences in wind speed probability density distribution, thereby achieving a more accurate evaluation of the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer, and more accurately correcting the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer, thereby improving the measurement accuracy of the backscattering coefficient of the satellite microwave scatterometer. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings: Figure 1 A schematic diagram of the main process of a method for correcting nonlinear deviation of backscattering coefficient provided by the present invention; Figure 2The diagram is a diagram of the variation curve between the backscattering coefficient and the relative azimuth angle at the Ku-band HH polarization under the condition of 46° incident angle drawn using the QSCAT-1 model function; Figure 3 The schematic diagram of the change curve between the backscattering coefficient and wind speed under the Ku-band HH polarization 46° incident angle condition drawn using the QSCAT-1 model function; Figure 4 The schematic diagram of the backscattering coefficient error variation curve caused by different wind direction deviations under different wind speed conditions, drawn according to the QSCAT-1 model function under the Ku-band HH polarization 46° incident angle condition; Figure 5 It is a schematic diagram of the probability density distribution of the sea surface wind speed and sea surface wind direction in the selected open ocean within a certain period of time; Figure 6 Based on Figure 5 The probability density distribution diagram of the relative azimuth (wind direction) of two wind speed intervals is selected from the data; Figure 7 A main logic diagram of wind speed and azimuth interval division in a method for correcting nonlinear deviation of backscatter coefficient provided by the present invention; Figure 8 Schematic diagram of data of an azimuthally uniformly distributed calibration model data set for a method for correcting nonlinear deviation of backscattering coefficient provided by the present invention; Fig. 9 A schematic structural diagram of a device for correcting nonlinear deviation of backscattering coefficient provided by the present invention. DETAILED DESCRIPTION

[0018] In order to facilitate the clear description of the technical solutions of the embodiments of the present invention, in the embodiments of the present invention, the words "first", "second" and the like are used to distinguish the same items or similar items with substantially the same functions and effects. For example, the first threshold and the second threshold are only used to distinguish different thresholds, and do not limit their order of precedence. Those skilled in the art can understand that the words "first", "second" and the like do not limit the quantity and execution order, and the words "first", "second" and the like do not necessarily limit them to be different.

[0019] It should be noted that, in the present invention, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0020] In the present invention, "at least one" means one or more, and "plurality" means two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the associated objects in the previous time are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can mean: a, b, c, the combination of a and b, the combination of a and c, the combination of b and c, or the combination of a, b and c, where a, b, c can be single or multiple.

[0021] At present, microwave scatterometers are mainly used for wind field observation on the sea surface. Nonlinear deviations will lead to inconsistent error deviations in the scatterometer inversion of different wind speeds, especially for high wind speeds above 20m / s and typhoon monitoring. Although the satellite microwave scatterometer will calibrate the energy response characteristics of its internal receivers, traveling wave tube amplifiers and other key microwave devices in the laboratory calibration before launch to ensure that the microwave scatterometer has the same or similar measurement deviations at different energy response levels, that is, there is no nonlinear deviation. However, due to the influence of factors such as the near-field effect of ground microwave antenna radiation transmission, multipath effect, and long-term performance attenuation of remote sensors, the laboratory calibration before launch cannot completely and long-term solve the problem of nonlinear deviation correction of microwave scatterometers. At the same time, during the on-orbit operation, the existing on-orbit calibration technology cannot detect and identify the nonlinear deviation of the backscatter coefficient related to the observed energy response. It can only rely on and believe the system energy response calibration results before launch, assuming that there is no nonlinear deviation in the entire measurement system of the microwave scatterometer, and then only correct its systematic measurement deviation.

[0022] Therefore, the present invention proposes a method and device for correcting the nonlinear deviation of the backscattering coefficient, which are applied to a satellite microwave scatterometer. Based on the measurement data of the ocean surface observed by the satellite microwave scatterometer and the true value data of the observed ocean surface during the on-orbit operation of the satellite microwave scatterometer obtained by a preset data acquisition method, a large sample initial sample data set is constructed, and an azimuth uniform distribution calibration model data set is constructed by using a dynamic uniform distribution construction method and a random error elimination method, thereby eliminating random errors in the sample data set; finally, based on the azimuth uniform distribution calibration model data set, the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer is corrected, thereby improving the measurement accuracy of the backscattering coefficient of the satellite microwave scatterometer; and there is no need to build expensive specific equipment or conduct a large number of field observation tests, and the method has the advantages of being simple, easy to operate, stable and reliable.

[0023] Next, the technical solution of the present invention is described in detail with reference to the accompanying drawings: In a first aspect, the present invention provides a method for correcting nonlinear deviation of backscatter coefficients, which is applied to satellite microwave scatterometers; see Figure 1 , Figure 1 A schematic diagram of the main process of a method for correcting the nonlinear deviation of a backscattering coefficient provided by the present invention; its executor is a server or terminal that adopts the correction method provided by the present invention, such as a satellite operation monitoring platform, a calibration and inspection data processing platform or a satellite microwave scatterometer, etc.

[0024] In FIG. 1 , the method may include: Step 110: Acquire measurement data and true value data for a satellite microwave scatterometer, wherein the measurement data is the measurement data of the satellite microwave scatterometer itself during its on-orbit operation, and the true value data is the true value data of the satellite microwave scatterometer during its on-orbit operation obtained by a preset data acquisition method.

[0025] Step 120: performing data preprocessing on the measurement data and the true value data to obtain an initial sample data set; the data preprocessing at least includes geographic spatiotemporal matching and eliminating observation data under rainfall conditions.

[0026] In step 110 to step 120, since the satellite microwave scatterometer is mainly used to measure the backscatter coefficient of the ocean surface, and then invert the wind speed and wind direction information of the sea surface; therefore, the measurement data of the satellite microwave scatterometer is the measurement data of the ocean surface, and the ocean surface data (true value data) synchronously observed by the satellite microwave scatterometer during the on-orbit operation is obtained by using a preset data acquisition method; the method for obtaining the true value data can be a method of using an airborne microwave scatterometer with the same frequency as the satellite microwave scatterometer to carry out satellite-ground synchronous measurement or obtaining the sea surface wind speed and wind direction data of the sea surface observed by the satellite from numerical model data, and obtaining the backscatter coefficient by a method of simulating and calculating according to a preset geophysical model function, and obtaining other relevant data such as time, longitude, latitude, sea surface wind speed, etc. by combining the numerical model data. By performing data preprocessing on the above measurement data and true value data, including at least geographic time-space matching and eliminating observation data under rainfall conditions, an initial sample data set with a large sample size is constructed; compared with the prior art, the initial sample data set has a larger data volume, which is more conducive to the analysis and processing of dynamically uniformly distributed data, as well as the dynamic elimination of random errors.

[0027] Step 130: Based on the initial sample data set, an azimuth uniform distribution calibration model data set is constructed using a dynamic uniform distribution construction method and a random error elimination method.

[0028] Step 140: Based on the azimuthally uniformly distributed calibration model data set, correct the nonlinear deviation of the backscatter coefficient of the satellite microwave scatterometer.

[0029] In step 130 to step 140, firstly, based on the target sample set data, the target sample set data is divided into a plurality of data sets with uniform distribution in wind speed intervals by using a dynamic uniform distribution construction method, and then each wind speed interval is divided into a wind direction angle interval with uniform distribution, so as to obtain a plurality of target azimuth interval calibration data sets with uniform distribution in azimuth intervals; finally, the invalid wind speed data in the plurality of target azimuth interval calibration data sets are eliminated by using a random error elimination method, so as to obtain an azimuth uniform distribution calibration model data set; the random error of the initial sample data set is eliminated, so that the azimuth uniform distribution calibration model data set can be used to accurately correct the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer, so as to improve the accuracy of the correction of the nonlinear deviation of the backscattering coefficient, thereby improving the measurement accuracy of the backscattering coefficient of the satellite microwave scatterometer.

[0030] Compared with the prior art, the present invention provides a method for correcting the nonlinear deviation of a backscattering coefficient. The method obtains measurement data and true value data of a satellite microwave scatterometer, performs data preprocessing on the measurement data and the true value data, and obtains an initial sample data set. Then, based on the initial sample data set, an azimuth uniform distribution calibration model data set is constructed by using a dynamic uniform distribution construction method and a random error elimination method. Finally, based on the azimuth uniform distribution calibration model data set, the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer is corrected. Based on this, since an initial sample data set with a large sample data volume compared with the prior art is constructed based on the measurement data and the true value data, and an azimuth uniform distribution calibration model data set is constructed by using a dynamic uniform distribution construction method and a random error elimination method, the influence of the difference in wind speed probability density distribution is eliminated, and nonlinear deviation estimation is performed with data that approximates the characteristics of wind speed uniform distribution. Therefore, based on the azimuth uniform distribution calibration model data set, the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer can be more accurately evaluated, so as to realize accurate correction of the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer and improve the measurement accuracy of the backscattering coefficient of the satellite microwave scatterometer.

[0031] It should be noted that the applicant has conducted a large number of research and experiments on the correction method of the nonlinear deviation of the backscatter coefficient. The research shows that the premise of nonlinear deviation correction is to accurately measure the "true value" or "reference value" of the backscatter coefficient that can cover the required measurement range, and then compare it with the satellite scatterometer measurement value, so as to evaluate the nonlinear deviation of the satellite microwave scatterometer backscatter coefficient.

[0032] Backscatter coefficient of ocean surface It can be simplified as: (1) in, A, a, b and γ It can be determined by empirical fitting of data, and its value is related to the incident angle, polarization and frequency of the electromagnetic wave; U represents the sea surface wind speed; ψ Represents the relative azimuth between the sea surface wind direction and the incident electromagnetic wave; since the scatterometer observation azimuth can be accurately determined with an accuracy of 0.01°, and the current measurement accuracy of the sea surface wind direction is about 10° to 20°, the error of the relative azimuth ψ related to the backscatter coefficient is mainly affected by the sea surface wind direction. At present, domestic and foreign researchers have developed different empirical models of backscatter coefficients and electromagnetic wave signals (frequency, polarization, incident angle, azimuth), and sea surface environmental parameters (wind speed, wind direction) according to different operating frequencies and specific scatterometers, namely geophysical model functions. The empirical models of Ku-band geophysical model functions commonly used internationally include SASS-II, NSCAT-1, NSCAT-2, NSCAT-4 and QSCAT-1 model functions. Please refer to Figures 2 to 3 , Figure 2 The diagram is a diagram of the variation curve between the backscattering coefficient and the relative azimuth angle at the Ku-band HH polarization under the condition of 46° incident angle drawn using the QSCAT-1 model function; Figure 3 This is a schematic diagram of the change curve between the backscattering coefficient and wind speed under the Ku-band HH polarization 46° incident angle condition drawn using the QSCAT-1 model function; it should be noted that Figure 2 The approximate double cosine variation characteristics between the ocean surface backscatter coefficient and the relative azimuth (the angle between the satellite observation azimuth and the sea surface wind direction) are shown; Figure 3 The variation characteristics of the ocean surface backscatter coefficient with the sea surface wind speed are shown.

[0033] from Figure 2 and Figure 3 It can be concluded without a doubt that under the same wind speed condition, the difference in the backscatter coefficient between different relative azimuths can reach 5dB; and the greater the wind speed, the greater the backscatter coefficient; the change of the backscatter coefficient with the wind speed is not linear, the lower the wind speed, the greater the rate of change of the backscatter coefficient. The size of the sea surface backscatter coefficient is closely related to the sea surface wind speed and wind direction, and the dynamic range varies greatly. The value range of the backscatter coefficient is about -45dB to -10dB.

[0034] In summary, the noise (random error) of the backscatter coefficient measured by the satellite microwave scatterometer is related to the size of the sea surface backscatter coefficient. The larger the sea surface backscatter coefficient, the stronger the backscatter energy, the greater the signal-to-noise ratio of the microwave scattering and the measurement signal, and thus the higher the accuracy of the backscatter coefficient measurement, and the smaller the random error; conversely, the greater the random error of the backscatter coefficient measurement. Therefore, the accuracy of the backscatter coefficient measurement can be improved by eliminating the random error of the microwave scatterometer in measuring the backscatter coefficient. Based on this, the present invention establishes a large sample data set by acquiring sample data from various aspects, so that a large sample (far exceeding the number of samples required for estimating the sample average value agreed by statistics) data averaging can be used to reduce the impact of random measurement errors, thereby improving the accuracy of the systematic deviation and nonlinear deviation estimation of the satellite microwave scatterometer.

[0035] In step 110, preferably, the measurement data may include at least a first backscatter coefficient, time, longitude, latitude, polarization beam, incident angle, azimuth and quality indicator; the true value data may include at least a second backscatter coefficient, sea surface wind speed, sea surface wind direction, atmospheric cloud liquid water content, time, longitude and latitude.

[0036] Specifically, the measurement data parameters of the satellite microwave scatterometer itself do not need to be measured and obtained, but only need to be read from the satellite data file; while the true value data needs to be obtained by conducting satellite-ground synchronous observation with the satellite, for example, the following two preset data acquisition methods are used to obtain the true value data of the satellite microwave scatterometer during its on-orbit operation. Because the large sample dynamic uniform distribution technology used later can effectively eliminate the measurement error of the ocean surface backscatter coefficient, both methods of obtaining the sea surface backscatter coefficient are acceptable. The second method of simulating the sea surface wind speed and direction has a lower economic cost, and the present invention preferably uses method two.

[0037] Method 1: Use an airborne microwave scatterometer with the same frequency as the satellite microwave scatterometer to carry out satellite-ground synchronous measurement to obtain the measured backscatter coefficient for the sea surface, and use the measured backscatter coefficient as the second backscatter coefficient.

[0038] Method 2: The sea surface wind speed and wind direction data of the satellite-observed sea surface are obtained from the numerical model data, and combined with the preset geophysical model function, the calculated backscatter coefficient of the sea surface matching the satellite observation is simulated and calculated, and the calculated backscatter coefficient is used as the second backscatter coefficient.

[0039] It should be noted that the sea surface wind speed, sea surface wind direction and atmospheric cloud liquid water content in the true data can be obtained through numerical model data, and then matched with the backscattering coefficient through geographic space-time matching.

[0040] In step 120, preferably, data preprocessing is performed on the measured data and the true value data to obtain an initial sample data set, which may include: firstly performing geographic and temporal matching on the measured data and the true value data to obtain an intermediate sample data set; and then eliminating the measured data under rainfall conditions in the intermediate sample data set according to the atmospheric cloud liquid water content to obtain the initial sample data set.

[0041] For details, please refer to Figure 4 , Figure 4 The following is a schematic diagram of the backscatter coefficient error curve caused by different wind direction deviations under Ku-band HH polarization 46° incident angle conditions and different wind speed conditions, drawn according to the QSCAT-1 model function. This figure shows the numerical value of the backscatter coefficient error caused by the common relative azimuth error (the satellite observation azimuth accuracy can reach 0.03°, mainly caused by wind direction error) under the current wind field measurement technology level.

[0042] exist Figure 4 In the figure, the magnitude of the sea surface backscatter coefficient deviation caused by different wind direction accuracy deviations at 5m / s and 10m / s is given. Figure 4 It can be undoubtedly obtained that the error size and strength of the backscatter coefficient caused by different wind speeds and different relative azimuth conditions are different; at the same time, the wind speed and wind direction on the ocean surface are dynamically changing. Therefore, only using a large sample (far exceeding the number of samples required for estimating the sample mean agreed by statistics) data averaging method cannot guarantee the complete elimination of the influence of random measurement errors, which leads to inaccurate estimates of systematic deviations and nonlinear deviations. Considering these factors, when using the open ocean to carry out nonlinear error correction analysis of satellite microwave scatterometer backscatter coefficients, in addition to obtaining the sea surface backscatter coefficient, it is also necessary to obtain sea surface wind speed and wind direction information, and through the unique backscattering characteristics of the ocean surface and the probability distribution characteristics of the wind field, to minimize the influence of the random error of the second backscatter coefficient itself. Because electromagnetic waves need to pass through the atmosphere between the satellite microwave scatterometer and the sea surface transmission, under rainfall conditions, raindrops will not only interfere with the backscattering law of the sea surface, but also attenuate and scatter the electromagnetic waves. These factors cannot be accurately simulated and calculated, resulting in the ocean surface backscattering under rainfall conditions no longer having Figure 2 The backscattering characteristics displayed will lead to the failure of the dynamic elimination technology of random errors based on uniformly distributed samples. Therefore, it is necessary to obtain the cloud liquid water information of the atmosphere, eliminate the observation data of rainfall conditions, ensure that the second backscattering coefficient sample data have the same backscattering characteristic law, and further reduce the impact of random measurement errors.

[0043] In summary, the present invention obtains the corrected scatterometer data (measurement data) and the correction source data (true value data) respectively through steps 110 to 120, further completes the geographic spatiotemporal matching of the two, and then eliminates the rainfall condition observation data according to the atmospheric cloud liquid water content to realize the initial sample data set for subsequent processing, which is conducive to eliminating the influence of random measurement errors to the greatest extent.

[0044] In step 130, preferably, based on the initial sample data set, using a dynamic uniform distribution construction method and a random error elimination method, constructing an azimuth uniform distribution calibration model data set may include: Firstly, the initial sample data set is preprocessed by using a dynamic uniform distribution construction method to obtain multiple target azimuth interval calibration data sets with uniform distribution in azimuth intervals; the dynamic uniform distribution construction method at least includes a target wind speed interval division method and a target azimuth interval division method.

[0045] Then, the random error elimination method is used to eliminate invalid wind speed data in multiple target azimuth interval calibration data sets to obtain the azimuth uniform distribution calibration model data set.

[0046] Specifically, because the backscattering characteristics of the sea surface are dynamically changing in time and space, and no matter it is the actual measured sea surface backscattering coefficient, or the simulated calculation sea surface backscattering coefficient, the sea surface backscattering coefficient obtained will inevitably have a measurement random error error. Based on the current measurement technology level, the error of the actual measured backscattering coefficient is generally greater than 1.5dB, and the simulated calculation backscattering coefficient error introduced by wind speed and wind direction will also reach several dB errors, and the size of this random error is generally significantly higher than the systematic error. Therefore, it is necessary to eliminate the random error in the measurement as much as possible by the average processing of a large number of samples. The sea surface backscattering coefficient has a variation relationship of approximate double cosine with the relative azimuth angle related to the wind direction, therefore, by means of the symmetrical relationship of double cosine, and the probability density distribution of different wind speeds and wind directions, the random error of the sea surface backscattering coefficient obtained can be greatly eliminated.

[0047] Furthermore, from the relationship between the sea surface backscatter coefficient and the wind field environmental variables established by formula (1), it can be seen that when the relative azimuth between the wind direction and the scatterometer observation direction is uniformly distributed in the range of (0°, 360°), the dependence on the relative azimuth can be eliminated by integrating and averaging along the relative azimuth. The influence of the remaining random error is only related to the wind speed. However, in reality, the probability density distribution of the relative azimuth between satellite observations and sea surface wind direction is not uniformly distributed. Please refer to Figure 5 to Figure 6 , Figure 5 It is a schematic diagram of the probability density distribution of the sea surface wind speed and sea surface wind direction in the selected open ocean within a certain period of time; Figure 5The wind field distribution characteristics of common ocean surfaces are shown; Figure 6 Based on Figure 5 The probability density distribution diagram of the relative azimuth (wind direction) of two wind speed intervals is selected from the data; Figure 5 It can be clearly seen that the wind direction is not evenly distributed. Figure 6 It can be clearly seen that they are not evenly distributed, and the probability distribution of wind directions in different wind speed ranges is also obviously different.

[0048] Based on this, the dynamic uniform distribution construction method for constructing a calibration model data set with uniform azimuth distribution proposed in step 130 of the present invention can be used to divide the initial sample data set with a large sample data set into small wind speed interval data sets; then, for each wind speed interval data set, it is further divided into small azimuth interval data sets according to the wind direction; finally, based on the number of data in the minimum azimuth interval, the redundant data in other azimuth intervals are randomly eliminated, so that the data in the wind speed interval is uniformly distributed; thus, formula (1) can be used to perform integral averaging along the relative azimuth angle to eliminate the dependence on the relative azimuth angle, and the influence of the remaining random error is only related to the wind speed, thereby reducing the random error of the acquired sea surface backscatter coefficient.

[0049] It should be noted that the method for dividing the wind speed intervals and the azimuth intervals in the present invention may be a step-by-step division, that is, firstly dividing the wind speed intervals, and then dividing the azimuth intervals for each wind speed interval.

[0050] Preferably, the initial sample data set is preprocessed using a dynamic uniform distribution construction method to obtain multiple target azimuth interval calibration data sets with uniformly distributed azimuth intervals, which may include: dividing the sample data in the initial sample data set into wind speed intervals using a target wind speed interval division method to obtain multiple target wind speed interval data sets with uniformly distributed wind speed intervals; and then dividing the multiple target wind speed interval data sets into azimuth intervals using a target azimuth interval division method to obtain multiple target azimuth interval calibration data sets with uniformly distributed azimuth intervals.

[0051] Preferably, the target wind speed interval division method is used to divide the sample data in the initial sample data set into wind speed intervals to obtain multiple target wind speed interval data sets with uniformly distributed wind speed intervals. This may include: dividing the sample data in the initial sample data set into wind speed intervals at preset wind speed intervals to obtain multiple target wind speed interval data sets with uniformly distributed wind speed intervals; wherein the preset wind speed interval is 2 m / s.

[0052] Furthermore, a target azimuth interval division method is used to divide multiple target wind speed interval data sets into azimuth intervals to obtain multiple target azimuth interval calibration data sets with uniform distribution of azimuth intervals, which may include: dividing multiple target wind speed interval data sets into wind speed intervals at preset azimuth intervals to obtain multiple target azimuth interval data sets with uniform distribution of azimuth intervals; wherein the minimum number of data in the multiple target azimuth interval data sets is greater than the preset number, and the azimuth intervals with the number of data less than the preset number are eliminated; the azimuth interval interval may be 1°, 2° or other positive integers.

[0053] Preferably, a random error elimination method is used to eliminate invalid wind speed data in multiple target azimuth interval calibration data sets to obtain an azimuth uniformly distributed calibration model data set, which may include: taking the target azimuth interval calibration data set with the least data volume as a benchmark, and using a random error elimination method to randomly eliminate redundant data in other target azimuth interval calibration data sets, so that the data volume in the multiple target azimuth interval calibration data sets is evenly distributed, thereby obtaining an azimuth uniformly distributed calibration model data set.

[0054] Specifically, in order to construct a calibration model data set with uniform azimuth distribution, we divide the preset large sample data set into small wind speed interval data sets, that is, the data set with a wind speed interval of 2m / s as the small wind speed interval; then, for each wind speed interval data set, we divide it into small azimuth interval data sets according to the wind direction, and randomly remove the redundant data in other azimuth intervals based on the number of data in the minimum azimuth interval, so that the data in the wind speed interval is evenly distributed. Please refer to Figure 7 , Figure 7 A main logic diagram of wind speed and azimuth interval division in a method for correcting nonlinear deviation of backscatter coefficient provided by the present invention.

[0055] exist Figure 7 In the example, the initial sample data set of a certain polarization beam of the preset scatterometer can be divided into U1 to U n The data in each wind speed interval is further divided into ψ1 to ψ n azimuth intervals, thereby obtaining multiple target azimuth interval calibration data sets.

[0056] For example, for the principle of dividing the azimuth interval size, in order to ensure the effectiveness of uniform distribution, the interval interval should be less than 2°, and the number of samples in each interval should be statistically significant, generally requiring at least greater than "interval interval / 0.1". That is, if the azimuth interval is set to 2°, the minimum number of data in the azimuth interval should be greater than 20 (preset number); if the azimuth interval is set to 1°, the minimum number of data in the azimuth interval should be greater than 10.

[0057] Please refer to further Figure 3 and Figure 8 The present invention provides a method for correcting nonlinear deviation of backscatter coefficients provided by the present invention and shows schematic diagrams of data sets of an azimuthally uniformly distributed calibration model.

[0058] By adopting the random error elimination and control technology of the high dynamic change sea surface backscatter coefficient provided by the present invention, the following can be obtained: Figure 8 The azimuthally uniformly distributed calibration model data set is shown. Based on this data set, the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer can be corrected.

[0059] Further, combined with Figure 3 It can be concluded that the probability distribution of global sea surface wind speed has a certain degree of regularity, and generally presents the characteristics of Weibull distribution. If the data is eliminated according to the uniform distribution standard, most of the data exceeding 80% will be eliminated. Considering the difficulty of obtaining the sample data set required for calibration analysis, this method is not worth the loss.

[0060] Therefore, in step 140, preferably, based on the azimuth uniform distribution calibration model data set, the nonlinear deviation of the backscatter coefficient of the satellite microwave scatterometer is corrected, which can include: based on the azimuth uniform distribution calibration model data set, the average value of the backscatter coefficient of each azimuth corresponding interval data is calculated to obtain multiple target feature data; based on the multiple target feature data, the nonlinear deviation of the backscatter coefficient of the satellite microwave scatterometer is obtained by fitting using the least squares method; according to the nonlinear deviation of the backscatter coefficient of the satellite microwave scatterometer obtained by fitting, the nonlinear deviation of the backscatter coefficient of the satellite microwave scatterometer is corrected; the problem of using the uniform distribution standard to eliminate data, resulting in the elimination of most of the data exceeding 80%, is avoided.

[0061] Specifically, the backscattering coefficients in the interval data corresponding to each azimuth angle in the azimuth uniform distribution calibration model data set can be extracted, and the average value of the extracted backscattering coefficients can be calculated to obtain multiple target feature data; that is, the multiple numerical values ​​corresponding to the multiple target feature data are a data set of the average values ​​of the backscattering coefficients in the interval corresponding to each azimuth angle in the azimuth uniform distribution calibration model data set; based on the multiple target feature data, the least squares method is used to fit the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer; based on the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer obtained by fitting, the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer is corrected; thereby achieving a more accurate correction of the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer.

[0062] It should be noted that since the wind speed on the ocean surface presents a probability density distribution characteristic that is close to the Rayleigh distribution, that is, the probability of medium wind speed of about 7m / s is the highest, and the probability of sea conditions with higher and lower wind speeds gradually decreases. Therefore, if the correction data set is directly used for nonlinear deviation estimation, the data with a larger wind speed probability density distribution will have a greater weight on the error estimation; conversely, the contribution weight will be smaller; therefore, the error size of sample data with different contribution weights is different, which leads to inaccurate estimation of nonlinear deviation. In order to eliminate the influence of the difference in wind speed probability density distribution, the nonlinear deviation estimation is performed with data that approximates the uniform distribution characteristics of wind speed, and the result will be more accurate. Therefore, we first calculate the average value of the backscatter coefficient of the samples in each azimuth interval as the target feature data of the interval, and then estimate the nonlinear deviation based on the average value, and the result will be more accurate.

[0063] Based on this, a method for correcting the nonlinear deviation of a backscattering coefficient provided by the present invention can be applied to a satellite microwave scatterometer, and the azimuth uniform distribution technology is used to eliminate random errors and ensure the accuracy of subsequent nonlinear deviation estimation. In the nonlinear deviation estimation, the original data of the calibration model data set uniformly distributed in azimuth is not directly used, but the deviation is estimated based on the target characteristic data of the average value of the interval samples, so that the dynamic range characteristics of the change of the backscattering coefficient of the ocean surface under different wind speed and sea conditions can be retained, and the contribution difference of the sample weights under different wind speed and sea conditions can be reduced, so as to achieve more accurate estimation of the nonlinear deviation; the method provided by the present invention is applied to a satellite microwave scatterometer, and the ability to accurately identify and correct the inconsistency of the nonlinear deviation of the microwave scatterometer under different energy response conditions during the on-orbit operation of the satellite is achieved, and the accuracy of the backscattering coefficient of the microwave scatterometer is more refined, thereby improving the inversion accuracy of the sea surface wind speed and direction products, especially improving the wind speed inversion accuracy under conditions above 20m / s; and the microwave scatterometer will have better system performance, enhance the application ability of its data products under extremely harsh sea conditions, and provide valuable technical reference for the subsequent payload design optimization of the satellite microwave scatterometer. At the same time, the method provided by the present invention does not require the use of expensive active calibrator specific equipment, nor does it require a large number of field observation tests to achieve the estimation and correction of the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer, and has the advantages of being simple, easy to operate and low cost.

[0064] In a second aspect, the present invention provides a device for correcting nonlinear deviation of backscatter coefficients, which is applied to satellite microwave scatterometers. Fig. 9 , Fig. 9 A schematic structural diagram of a device for correcting nonlinear deviation of backscattering coefficient provided by the present invention.

[0065] exist Fig. 9 In the embodiment, the device may include: An acquisition module 910 is used to acquire measurement data and true value data for a satellite microwave scatterometer, wherein the measurement data is the measurement data of the satellite microwave scatterometer itself during its on-orbit operation, and the true value data is the true value data of the satellite microwave scatterometer during its on-orbit operation obtained by a preset data acquisition method; A preprocessing module 920 is used to perform data preprocessing on the measurement data and the true value data to obtain an initial sample data set; the data preprocessing at least includes geographic time-space matching and eliminating observation data under rainfall conditions; A construction module 930 is used to construct an azimuth uniform distribution calibration model data set based on the initial sample data set by using a dynamic uniform distribution construction method and a random error elimination method; The correction module 940 is used to correct the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer based on the azimuth uniform distribution calibration model data set.

[0066] Based on this, the present invention provides a device for correcting the nonlinear deviation of the backscatter coefficient, which obtains the measurement data and true value data of the satellite microwave scatterometer through the acquisition module 910, wherein the measurement data is the measurement data of the satellite microwave scatterometer itself during the on-orbit operation, and the true value data is the true value data of the satellite microwave scatterometer during the on-orbit operation obtained by a preset data acquisition method; and then uses the preprocessing module 920 to perform data preprocessing on the measurement data and the true value data to obtain an initial sample data set; the data preprocessing at least includes geographic time-space matching and eliminating observation data under rainfall conditions; and then The construction module 930 is used to construct an azimuth uniform distribution calibration model data set based on the initial sample data set, using a dynamic uniform distribution construction method and a random error elimination method; finally, the correction module 940 is used to correct the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer based on the azimuth uniform distribution calibration model data set; thereby, the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer can be more accurately evaluated, the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer can be more accurately corrected, and the measurement accuracy of the backscattering coefficient of the satellite microwave scatterometer is improved.

[0067] Although the present invention is described herein in conjunction with various embodiments, in the process of implementing the claimed invention, those skilled in the art may understand and implement other variations of the disclosed embodiments by viewing the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other components or steps, and "one" or "an" does not exclude multiple situations. A single processor or other unit may implement several functions listed in a claim. Certain measures are recorded in different dependent claims, but this does not mean that these measures cannot be combined to produce good results.

[0068] Although the present invention has been described in conjunction with specific features and embodiments thereof, it is apparent that various modifications and combinations may be made thereto without departing from the spirit and scope of the present invention. Accordingly, this specification and the accompanying drawings are merely exemplary illustrations of the present invention as defined by the appended claims and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present invention. Obviously, those skilled in the art may make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, the present invention is intended to include such modifications and variations if they fall within the scope of the claims of the present invention and their equivalents.

Claims

1. A method for correcting nonlinear deviation of backscatter coefficient, applied to satellite microwave scatterometer, characterized in that: include: Acquire measurement data and true value data for a satellite microwave scatterometer, wherein the measurement data is the measurement data of the satellite microwave scatterometer itself during its on-orbit operation, and the true value data is the true value data of the satellite microwave scatterometer during its on-orbit operation obtained by a preset data acquisition method; Performing data preprocessing on the measurement data and the true value data to obtain an initial sample data set; The data preprocessing at least includes geographic time-space matching and eliminating observation data under rainfall conditions; Based on the initial sample data set, a dynamic uniform distribution construction method and a random error elimination method are used to construct an azimuth uniform distribution calibration model data set; Based on the azimuthally uniformly distributed calibration model data set, the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer is corrected.

2. The method for correcting the nonlinear deviation of the backscattering coefficient according to claim 1, characterized in that: The method of constructing an azimuth uniform distribution calibration model data set based on the initial sample data set by using a dynamic uniform distribution construction method and a random error elimination method includes: The initial sample data set is preprocessed using a dynamic uniform distribution construction method to obtain a plurality of target azimuth interval calibration data sets with uniform distribution of azimuth intervals; the dynamic uniform distribution construction method at least includes a target wind speed interval division method and a target azimuth interval division method; The invalid wind speed data in a plurality of the target azimuth interval calibration data sets are eliminated by using a random error elimination method to obtain the azimuth uniform distribution calibration model data set.

3. The method for correcting the nonlinear deviation of the backscatter coefficient according to claim 2, characterized in that: The method of using the dynamic uniform distribution construction method to preprocess the initial sample data set to obtain a plurality of target azimuth interval calibration data sets uniformly distributed in azimuth intervals includes: Using the target wind speed interval division method to divide the sample data in the initial sample data set into wind speed intervals, to obtain a plurality of target wind speed interval data sets with wind speed intervals evenly distributed; The target azimuth interval division method is used to divide the multiple target wind speed interval data sets into azimuth intervals, so as to obtain multiple target azimuth interval calibration data sets with the azimuth intervals evenly distributed.

4. The method for correcting the nonlinear deviation of the backscatter coefficient according to claim 3, characterized in that: The method of dividing the sample data in the initial sample data set into wind speed intervals by using the target wind speed interval division method to obtain a plurality of target wind speed interval data sets with wind speed intervals uniformly distributed, including: The sample data in the initial sample data set are divided into wind speed intervals according to a preset wind speed interval interval, so as to obtain a plurality of target wind speed interval data sets with the wind speed interval intervals being evenly distributed; wherein the preset wind speed interval interval is 2 m / s.

5. The method for correcting the nonlinear deviation of backscatter coefficient according to claim 3, characterized in that: The method of dividing the target azimuth interval into azimuth intervals for the plurality of target wind speed interval data sets using the target azimuth interval division method to obtain a plurality of target azimuth interval calibration data sets evenly distributed in the azimuth intervals comprises: The multiple target wind speed interval data sets are divided into wind speed intervals at preset azimuth intervals to obtain multiple target azimuth interval calibration data sets that are evenly distributed in the azimuth intervals; wherein the minimum number of data in the multiple target azimuth interval data sets is greater than the preset number.

6. The method for correcting the nonlinear deviation of backscatter coefficient according to claim 2, characterized in that: The method of eliminating invalid wind speed data in a plurality of target azimuth interval calibration data sets by using a random error elimination method to obtain the azimuth uniform distribution calibration model data set includes: Taking the target azimuth interval calibration data set with the least data volume as a benchmark, a random error elimination method is used to randomly remove redundant data in other target azimuth interval calibration data sets, so that the data volumes in the multiple target azimuth interval calibration data sets are evenly distributed, thereby obtaining the azimuth evenly distributed calibration model data set.

7. The method for correcting the nonlinear deviation of backscatter coefficient according to claim 1, characterized in that: The method of correcting the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer based on the azimuthally uniformly distributed calibration model data set includes: Based on the azimuth uniform distribution calibration model data set, the backscatter coefficient of each azimuth corresponding interval data is averaged to obtain multiple target feature data; Based on the plurality of target characteristic data, a least squares method is used to fit the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer; The nonlinear deviation of the backscatter coefficient of the satellite microwave scatterometer is obtained according to the fitting, and the nonlinear deviation of the backscatter coefficient of the satellite microwave scatterometer is corrected.

8. The method for correcting the nonlinear deviation of backscatter coefficient according to claim 1, characterized in that: The measurement data at least includes a first backscatter coefficient, time, longitude, latitude, polarization beam, incident angle, azimuth, and quality indicator; The true value data at least includes a second backscatter coefficient, sea surface wind speed, sea surface wind direction, atmospheric cloud liquid water content, time, longitude and latitude; The method of obtaining the true value data of the satellite microwave scatterometer during the on-orbit operation by using a preset data acquisition method includes: Using an airborne microwave scatterometer with the same frequency as the satellite microwave scatterometer to carry out satellite-ground synchronous measurement to obtain a measured backscatter coefficient for the sea surface, and using the measured backscatter coefficient as the second backscatter coefficient; Alternatively, the sea surface wind speed and wind direction data of the sea surface observed by satellite are obtained from numerical model data, and combined with a preset geophysical model function, a calculated backscatter coefficient of the sea surface matching the satellite observation is simulated and calculated, and the calculated backscatter coefficient is used as the second backscatter coefficient.

9. The method for correcting the nonlinear deviation of backscatter coefficient according to claim 1, characterized in that: The performing data preprocessing on the measurement data and the true value data to obtain an initial sample data set includes: Performing geographic and temporal spatial matching on the measurement data and the true value data to obtain an intermediate sample data set; The measurement data under rainfall conditions in the intermediate sample data set are eliminated according to the atmospheric cloud liquid water content to obtain an initial sample data set.

10. A device for correcting nonlinear deviation of backscatter coefficient, applied to satellite microwave scatterometer, characterized in that: include: An acquisition module, the acquisition module is used to acquire measurement data and true value data for a satellite microwave scatterometer, the measurement data is the measurement data of the satellite microwave scatterometer itself during its on-orbit operation, and the true value data is the true value data of the satellite microwave scatterometer during its on-orbit operation obtained by a preset data acquisition method; A preprocessing module, the preprocessing module is used to perform data preprocessing on the measurement data and the true value data to obtain an initial sample data set; The data preprocessing at least includes geographic time-space matching and eliminating observation data under rainfall conditions; A construction module, wherein the construction module is used to construct an azimuth uniform distribution calibration model data set based on the initial sample data set by using a dynamic uniform distribution construction method and a random error elimination method; A correction module is used to correct the nonlinear deviation of the backscattering coefficient of the satellite microwave scatterometer based on the azimuth uniform distribution calibration model data set.

Images