Atmospheric calibration instrument data processing method using active and passive cooperative remote sensing

Abstract

The invention belongs to the technical field of remote sensing, and relates to an atmospheric calibration instrument data processing method using active and passive cooperative remote sensing, which comprises the following steps: S1, carrying a synchronous atmospheric calibration instrument integrating a multispectral polarization camera and a multi-beam laser radar on the same satellite with a high-resolution imaging load to realize complete coverage of an observation area and time-space synchronous observation, acquiring multispectral polarization data and laser radar data; s2, on the basis of laser radar data, atmospheric vertical profile parameters are inverted through a parameter inversion method; s3, based on the multispectral polarization data, combining with a ground surface polarization reflectivity model to invert aerosol related parameters; s4, performing space-time matching and uniformity judgment on the laser radar data and the multispectral polarization data, and screening effective synchronous observation points; s5, collaborative inversion and iterative optimization are carried out based on the active and passive data of the synchronous observation points, and geogas three-dimensional space-time distribution parameters are obtained; and S6, performing atmospheric correction on the remote sensing data of the high-resolution imaging load.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of remote sensing, and particularly relates to a data processing method of an atmospheric correction instrument using active and passive collaborative remote sensing. BACKGROUND

[0002] Atmospheric correction is a prerequisite for satellite remote sensing quantification, and currently used methods include: direct image information-based atmospheric correction method, radiation transmission calculation-based atmospheric correction method, image reflectivity data-based atmospheric correction method after atmospheric inversion, and synchronous atmospheric correction instrument-based method.

[0003] The first image information-based method does not focus on the physical mechanism of correction, improves image clarity through haze removal or defogging, and is only a relative correction; the radiation transmission-based method has high precision, and needs to separately input accurate physicochemical optical parameters of components such as aerosols, such as other satellite inversion products or ground-based observation results, but these parameters are difficult to obtain completely; the image reflectivity data-based method first obtains aerosol parameters through inversion, but is easily affected by various factors such as cloud recognition, matching and radiation calibration, and cannot guarantee the precision of parameter inversion and atmospheric correction.

[0004] The synchronous atmospheric correction instrument is used in cooperation with the main load, accurately obtains the instantaneous atmospheric parameters of the main load remote sensing through hardware synchronization, can realize accurate atmospheric correction, has become an important development direction of high spatial resolution satellite atmospheric correction, and has broad application prospects.

[0005] However, the current method based on the synchronous atmospheric correction instrument still has problems to be solved.

[0006] From the load index, the low spatial resolution and the preset three-dimensional space-time distribution cannot accurately describe the dynamic changes of the satellite remote sensing instantaneous pixel level, and it is difficult to guarantee the effectiveness of the atmospheric parameters; from the algorithm model, the non-uniformity of the atmosphere and the ground surface parameters is not considered, the radiation transmission simulation under the conditions of polluted weather and complex ground surface has the problems of low precision or cannot be simulated, and the atmospheric parameter inversion precision is affected; from the processing method, the atmospheric correction needs to provide detailed atmospheric parameters, but the current atmospheric inversion cannot guarantee high precision and full coverage, causes the partial absence of the ground-air parameters, and restricts the atmospheric correction precision and applicability. SUMMARY

[0007] The application provides a data processing method of an atmospheric correction instrument using active and passive collaborative remote sensing.

[0008] The data processing method of the atmospheric correction instrument using active and passive collaborative remote sensing comprises the following steps: S1: A synchronous atmospheric correction instrument integrating a multi-spectral polarization camera and a multi-beam lidar is carried on the same satellite as a high-resolution imaging payload to achieve complete coverage and time-space synchronous observation of the observation area, and to obtain multi-spectral polarization data and lidar data; S2: Based on the lidar data, the atmospheric vertical profile parameters are retrieved by a parameter inversion method; S3: Based on the multi-spectral polarization data, combined with the surface polarization reflectance model and the three-dimensional profile constraints provided by the lidar, the atmospheric aerosol parameter lookup table constructed by the three-dimensional radiative transfer model is traversed to retrieve the combination of the minimum error of the ground-to-air parameters and the sensor data, and to retrieve related parameters such as aerosol optical thickness and coarse and fine particle ratio; S4: Time-space matching and uniformity judgment are performed on the lidar data and multi-spectral polarization data, and effective synchronous observation points are selected; S5: Collaborative inversion is performed based on the active and passive data of the synchronous observation points, the retrieval results are further corrected by iterative optimization of the inversion parameters, the optimization inversion of the ground-to-air three-dimensional space-time distribution is realized, and high-precision ground-to-air three-dimensional space-time distribution parameters are obtained; S6: The ground-to-air three-dimensional space-time distribution parameters are used to perform atmospheric correction on the remote sensing data of the high-resolution imaging payload.

[0009] As a preferred, the multi-spectral polarization camera of the synchronous atmospheric correction instrument in S1 has several observation spectral bands, some of which have polarization detection capability, used to obtain total reflectance and polarization reflectance data of the observation target during on-orbit synchronous observation.

[0010] As a preferred, the lidar of the synchronous atmospheric correction instrument in S1 adopts a double-beam double-wavelength detection method, which is used to retrieve the aerosol and cloud backscattering coefficient and extinction coefficient profile.

[0011] As a preferred, the lidar data retrieval of S2 adopts Fernald double-component algorithm, which is adjusted and optimized by the initial parameters and subsequent collaborative inversion results.

[0012] As a preferred, in the multi-spectral polarization data inversion process of S3, the atmospheric aerosol parameter lookup table is constructed by a three-dimensional radiative transfer model, and during the traversal process, a cost function is set to quantify the error of the ground-to-air parameter and sensor data combination, and finally the combination with the minimum error is selected for high-precision inversion of the aerosol optical thickness and coarse and fine particle ratio.

[0013] As a preferred, the surface polarization reflectance model in S3 includes: using the Nadal & Bréon semi-empirical model, combining the land cover classification and NDVI value to determine the model parameters, and realizing the separation of the surface polarization contribution and the atmospheric aerosol contribution.

[0014] Preferably, the spatiotemporal matching in S4 includes: determining the best matching pixel by calculating the latitude and longitude deviation between the pixels of the multi-beam lidar and the multispectral polarization camera, and selecting the corresponding deviation calculation method according to the dimensionality.

[0015] Preferably, the uniformity judgment in S4 is evaluated by the cloud ratio of the passive instrument and the standard deviation of the cloud top height deviation and cloud bottom height deviation of the lidar profile.

[0016] Preferably, the three-dimensional profile constraint in the active and passive data collaborative inversion of S5 includes constraining the range of atmospheric parameters under cloud cover and high aerosol concentration conditions to ensure the effectiveness of the inversion results. On this basis, the aerosol optical thickness inverted from multispectral polarization data is used as a reference to adjust the lidar ratio until the aerosol optical thickness inversion error of the two types of data meets the preset range.

[0017] Preferably, the three-dimensional spatiotemporal distribution parameters of the Earth and atmosphere in S6 include the vertical profile of atmospheric composition, aerosol optical thickness, coarse and fine particle ratio, column water vapor content, and thin cirrus cloud-related parameters. These three-dimensional information of key Earth and atmosphere parameters for high-precision adaptation support atmospheric correction of high spatial resolution remote sensing data such as airborne main payloads and satellite payloads, thereby improving the quality of remote sensing images and the accuracy of quantitative parameters.

[0018] The present invention has the following beneficial effects: This invention utilizes a combined active and passive architecture of a multispectral polarization camera and a multibeam lidar to accurately invert atmospheric vertical profiles, providing three-dimensional constraints for multispectral polarization data inversion and effectively solving the problem of underdetermined parameters in traditional methods with "one equation and two unknowns".

[0019] 2. This invention uses a three-dimensional profile constraint mechanism to directly define the range of atmospheric parameters during the inversion process, and combines uniformity judgment to screen effective data, which significantly reduces the impact of cloud and aerosol interference on parameter inversion.

[0020] 3. By mounting the synchronous atmospheric correction instrument and the high-resolution imaging payload on the same satellite, this invention ensures complete coverage of the observation area and spatiotemporal synchronization, and the acquired ground-atmosphere parameters can accurately match the instantaneous pixel-level dynamic changes of the high-resolution imaging payload. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of a synchronous atmospheric correction instrument according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the inversion of backscattering coefficient and extinction coefficient of the Mie scattering lidar according to an embodiment of the present invention; Figure 3 This is a schematic diagram of multispectral polarization camera inversion according to an embodiment of the present invention; Figure 4 The polarization reflectance value of a typical surface in this embodiment of the invention is shown in the Nadal model. Figure 5 This is a schematic diagram of active trajectory and passive pixel matching in an embodiment of the present invention. Detailed Implementation

[0022] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions in the embodiments of this invention will be clearly described below in conjunction with the examples.

[0023] See Figure 1 The synchronous atmospheric corrector integrates a multispectral polarization camera and a multibeam lidar as active and passive collaborative payloads, enhancing the ability to measure spatial coverage parameters of clouds and atmospheric aerosols.

[0024] Develop corresponding parameter inversion methods to obtain the spatiotemporal distribution of atmospheric composition vertical profiles, and improve the accuracy of parameter inversion and identification such as aerosol fine modes, total optical thickness, column water vapor content, and thin cirrus clouds.

[0025] The developed synchronous atmospheric corrector is carried on the same satellite in conjunction with other high-resolution imaging payloads. The observation area of ​​the synchronous atmospheric corrector completely covers the target area observed by the high-resolution imaging payloads. The atmospheric corrector can synchronously obtain the three-dimensional spatiotemporal distribution of the atmosphere in each pixel during high-resolution imaging, simulating a radiative transfer process closer to the real scene. This is expected to overcome the limitations of traditional correction methods, thereby achieving better imaging quality.

[0026] The key to atmospheric correction is obtaining accurate aerosol parameters. Currently, the mainstream method is based on passive spectral data, using reflectance or polarized reflectance. However, passive data inversion has significant drawbacks: first, surface contributions are difficult to accurately eliminate, especially in urban areas with numerous buildings where ground reflectance is high, and polarized reflectance is difficult to model with a simple model of complex terrain features; second, aerosol models contain a large number of physical parameters, and it is difficult to accurately calculate these parameters from passively observed data; and third, it lacks vertical detection capabilities, relying mainly on prior knowledge.

[0027] like Figures 1 to 5 As shown in the figure, this invention proposes a data processing method for an atmospheric correction instrument that utilizes active and passive collaborative remote sensing. S1: By integrating a multispectral polarization camera and a multibeam lidar into a synchronous atmospheric correction instrument, and carrying it on the same satellite as the high-resolution imaging payload, the observation area is fully covered and spatiotemporally synchronized, and multispectral polarization data and lidar data are acquired. Among them, spaceborne active sensors can obtain vertical profile data and provide single or multiple profile data along the orbit at a fixed angle. The high complementarity between active and passive remote sensing data has made the integration of active and passive data a hot topic in recent years.

[0028] For a given spatiotemporally synchronized observation point, the advantages of both active multibeam lidar and passive data are comprehensively utilized to collaboratively invert and obtain high-precision three-dimensional atmospheric information. For multi-beam lidar, the Fernald method is used to invert the profiles of aerosol and cloud backscattering coefficients and extinction coefficients.

[0029] However, it faces the underdetermined problem of "one equation and two unknowns". It is usually assumed that the lidar ratio of aerosols and clouds is used as a priori input, and its rationality needs to be constrained by passive remote sensing. By supplementing multiple polarization band data, passive remote sensing data provides accurate fine modes and multi-band aerosol optical thickness, which is used to optimize the selection of lidar ratio, thereby iteratively achieving high-precision inversion.

[0030] Referring to Table 1, in one feasible embodiment of this application, the lidar adopts a dual-beam dual-wavelength (532 / 1064nm) detection method, and the multispectral polarization camera adopts the following spectral band design, wherein the 865 / 1640nm bands have polarization detection capability.

[0031] Table 1. Band design of the multispectral polarization camera for the synchronous atmospheric corrector. S2: Based on lidar data, atmospheric vertical profile parameters are retrieved using parameter inversion methods; See Figure 2 The core of cloud and aerosol optical parameter vertical profile inversion is solving the lidar equations. This application employs the Fernald two-component algorithm.

[0032] When a lidar emits a laser beam that interacts with atmospheric substances, the detection range increases. At this location, the energy of the atmospheric backscattered signal corresponding to the system. The lidar equation is as follows: (1) The atmospheric extinction coefficient and backscattering coefficient can be expressed as: (2) In the formula, the subscript mol represents atmospheric molecules and the subscript par represents aerosols.

[0033] The bicomponent Fernald method consists of two parts: forward and backward, expressed as follows: Backward: (3) Forward: (4) Where A(i) is It is the molecular backscattering coefficient determined according to the American Standard Atmospheric Model. The ratio of lidar to molecules, For aerosol radar ratio. This represents the minimum range resolution of the lidar. The Fernald algorithm is an iterative solution process; therefore, two parameters are required for inversion: the lidar ratio of the aerosol and the boundary value, and the calibrated aerosol backscattering coefficient. This is a constant, taken as 50 sr. This value can also be adjusted during subsequent fusion based on the aerosol model retrieved from the multispectral polarization camera. The calibration height c is generally chosen to be around 10 km, and it is assumed that... 1.02 or 1.05 Invert the measured data.

[0034] S3: Based on multispectral polarization data and combined with the surface polarization reflectance model, aerosol-related parameters are retrieved. See Figure 3 To develop inversion algorithms for aerosol optical thickness (AOD) and coarse-fine particle ratio (FMF) that couple polarization and non-polarization information, we will use an established atmospheric aerosol parameter lookup table to simulate the differences in reflectivity under different AOD and FMF conditions by setting different AOD values, FMF values, observation zenith angle, solar zenith angle, relative azimuth angle, and spectral bands, thereby determining whether high-precision inversion of AOD and FMF can be achieved.

[0035] Based on the sensitivity analysis results of AOD and FMF inversion, an AOD and FMF inversion algorithm (Equation 5) is constructed by setting the parameters of FMF. During the inversion process, the final inversion is achieved by setting the cost function (Equations 6 and 7). (5) in, Represents the ratio of coarse to fine particles. Reflectance and polarized reflectance simulated by an atmospheric aerosol parameter lookup table. and These represent the aerosol reflectance and polarized reflectance of fine and coarse particles, respectively. Observe the relevant parameters of the zenith angle, T, and the atmospheric transport coefficient.

[0036] (6) (7) in, It refers to the number of bands. These are the observations from each angle. and They are 0.490 respectively. Calculated and observed values ​​of total reflectance at the location; and They represent 0.675 respectively. and 0.870 Calculated and observed values ​​of polarization reflectance at the point of polarization; The cosine value of the solar zenith angle. It is the cosine value of the observed zenith angle. This refers to the relative azimuth angle.

[0037] S4: Perform spatiotemporal matching and uniformity assessment on lidar data and multispectral polarization data to select effective synchronous observation points; Cloud masks were identified using the 1380 and 1640 nm wavelengths from the Synchronous Atmospheric Corrector Multispectral Polarization Camera; the relevant algorithms are not detailed here. Due to the lack of a 2200 nm band, the closest unpolarized and polarized bands of 1640 nm were used, combined with the NVDI values ​​calculated from the 650 nm and 865 nm unpolarized bands, as the basis for calculating surface reflectivity and polarized reflectivity.

[0038] Nadal and Bréon (1999) performed cloud detection, atmospheric absorption correction, and atmospheric molecular and aerosol scattering correction on POLDER observation data from November 1996 and June 1997, and then analyzed the relationship between polarization reflectance and its characteristics. They found that polarization reflectance is related to… The correspondence between them is compared with that between them. The results show a better match, and the polarization reflectivity does not continuously increase at small scattering angles, but rather approaches saturation. Based on these findings, they propose a semi-empirical polarization reflectivity model: (8) in, The Fresnel coefficients of polarized light. and It is an empirical coefficient, jointly determined by the land cover classification model provided by the International Geosphere and Biosphere Project (IGBP) and the NDVI of the land surface.

[0039] See Table 2. The IGBP includes 17 different land surface types. After a series of attempts, Nadal and Bréon merged all the types in the IGBP into four land cover types: desert, shrubland, forest, and sparse vegetation.

[0040] In addition to the three categories of deserts, they are further subdivided into three categories based on NDVI values: 0 ≤ NDVI < 0.15, 0.15 ≤ NDVI < 0.3, and 0.3 ≤ NDVI; deserts are subdivided into two categories: 0 ≤ NDVI < 0.15 and 0.15 ≤ NDVI. This formula uses a logarithmic form, therefore, when the scattering angle is small, the polarization reflectivity will not increase indefinitely, but will tend to saturate.

[0041] Referring to Table 3, based on the data simulation experiment, the results of this semi-empirical model are closer to the actual observed values. See Figure 4 The values ​​are polarization reflectance values ​​simulated based on the Nadal model parameters of typical surfaces in Table 3. They are the same as the simulation data mentioned above. The observation location is on the main plane, and observations are conducted every 5 degrees.

[0042] As can be seen from the figure, even with the same NDVI, the polarization reflectance varies among different land features. Forest types have the lowest polarization reflectance values, while deserts have higher values. Compared to the Bréon model, the Nadal & Bréon model has higher fitting accuracy for actual measured multi-angle polarization data. This model is used by the POLDER land aerosol inversion algorithm. Therefore, this project intends to use the Nadal & Bréon model to calculate the surface polarization reflectance and achieve the separation of surface polarization contribution from atmospheric aerosol contribution.

[0043] S5: Based on the active and passive data from synchronous observation points, perform collaborative inversion, iteratively optimize inversion parameters, and obtain the three-dimensional spatiotemporal distribution parameters of the Earth and atmosphere; By leveraging the observational advantages of both active lidar and passive data, joint inversion of active and passive data is performed to provide more accurate active and passive data products for subsequent atmospheric correction. The Fernald method is used to invert the profiles of aerosol and cloud backscattering coefficients and extinction coefficients.

[0044] However, this process faces an underdetermined problem of "one equation, two unknowns," requiring the assumption of a lidar ratio between aerosols and clouds as prior input. The rationality of this lidar ratio assumption significantly impacts the retrieved optical characteristics; incorrect assumptions can introduce substantial inversion errors, severely affecting the accuracy of the retrieved aerosol properties. Passive remote sensing data can provide accurate aerosol optical thicknesses across multiple bands, which can serve as validation conditions for lidar inversion products, optimizing the selection of the lidar ratio and thus achieving high-precision data inversion.

[0045] The specific steps for combining AOD results from passive remote sensing to assist in retrieving the backscattering coefficient and extinction coefficient of lidar are as follows: (1) The threshold method is used to perform hierarchical identification of the lidar echo signal, and the cloud and aerosol classification and aerosol subclass identification are performed by combining polarization measurement information and ground scene information. Based on the identified cloud and aerosol types, a typical lidar ratio is preset as the initial value.

[0046] (2) Assuming the backscattering ratio at the reference height of the Fernald algorithm, the extinction coefficient and backscattering coefficient profiles of aerosols and clouds are inverted by combining the forward and backward Fernald algorithms.

[0047] (3) The aerosol extinction coefficient obtained by inversion is used to calculate AOD and compare it with the passive data of the corresponding pixel.

[0048] (4) Using the aerosol AOD value corresponding to the passive remote sensing wavelength as the true value, fine-tune the lidar ratio value according to the difference in column AOD and repeat steps (2)-(3) until the AOD inversion result obtained by lidar and the AOD inversion error of passive data reach the acceptable error range.

[0049] The observation results from the multi-beam lidar are matched to the corresponding passive pixels. These pixels, which simultaneously possess both active and passive measured data, are called synchronous observation points. This process utilizes the latitude and longitude distances between the multi-beam lidar observation points and the passive pixels for spatiotemporal matching.

[0050] For each pixel on the multi-beam lidar trajectory, the pixel of the best-matched passive instrument can be determined by the squared deviation. Or cost function Sure: (9) (10) in, This represents the sum of squares of the absolute latitude and longitude errors of the multi-beam lidar and passive pixel points. It is expressed as the sum of squares of relative errors. and These represent the latitude and longitude of the passive pixel, respectively. and This represents the latitude, longitude, and dimension of the active pixel.

[0051] See Figure 5 Since the relationship between radian difference and latitude difference is not linear across different dimensions, Equation 9 is used for judgment when the matching dimension is below the dimension threshold. The pixel corresponding to the minimum value is the best-matching pixel on the trajectory. When the matching dimension exceeds the dimension threshold, Equation 10 is used for judgment. The pixel with the smallest value that is less than the threshold of 10⁻⁸ is the best-matching pixel on the trajectory. The boundary condition, i.e., the choice of the dimensional threshold, can be obtained by converting the dimensional difference into radian difference and then verifying it through comparison.

[0052] Due to the difference in the actual observation range (spatial resolution) between active and passive data, multiple active pixels may exist within a single passive grid. Therefore, the uniformity of the passive data must be assessed. If the uniformity of the passive data is poor, that point should be removed from the synchronous observation points and not used as a candidate input for subsequent inversion analysis. The uniformity assessment can be performed using the cloud coverage rate provided by the passive instrument and the variation amplitude of multiple vertical profiles from the active multibeam lidar falling within the passive grid point.

[0053] To analyze the homogeneity or similarity of cloud types, the cloud top height (CTH) and cloud base height (CBH) of the active multibeam radar profiles can be calculated, and the standard deviations of the cloud top height deviation (CTHD) and cloud base height deviation (CBHD) can be evaluated. If the variations in the above parameters are small among different lidar profiles within the same passive data grid, it indicates that the passive data is homogeneous and that the lidar and passive data can be well matched.

[0054] S6: Using the aforementioned three-dimensional spatiotemporal distribution parameters of the Earth and atmosphere, perform atmospheric correction on the remote sensing data of the high-resolution imaging payload.

[0055] In one feasible embodiment, the three-dimensional spatiotemporal distribution parameters of the Earth and atmosphere obtained by active-passive collaborative inversion are used to accurately describe the three-dimensional atmospheric structure corresponding to each pixel at the instant of high-resolution imaging, including aerosols in the vertical direction, cloud distribution, and parameter uniformity in the horizontal direction, thereby simulating a radiative transfer process that is closer to the real scene. Based on the simulation results of real radiative transfer, the radiative interference of the atmosphere on high-resolution imaging remote sensing data, such as aerosol scattering, water vapor absorption, and cloud effects, is removed. This achieves decoupling of the Earth-atmosphere system, separates the atmospheric radiation contribution from the true surface radiation information, and ultimately breaks through the limitations of traditional correction methods.

[0056] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A data processing method for an atmospheric correction instrument utilizing active-passive coordinated remote sensing, characterized in that, Includes the following steps: S1: By integrating a multispectral polarization camera and a multibeam lidar into a synchronous atmospheric correction instrument, and carrying it on the same satellite as the high-resolution imaging payload, the observation area is fully covered and spatiotemporally synchronized, and multispectral polarization data and lidar data are acquired. S2: Based on lidar data, atmospheric vertical profile parameters are retrieved using parameter inversion methods; S3: Based on multispectral polarization data, combined with the surface polarization reflectance model and the three-dimensional profile constraints provided by lidar, the atmospheric aerosol parameter lookup table constructed by the three-dimensional radiative transfer model is traversed to retrieve the combination of ground-atmosphere parameters and sensor data with the smallest error, and the aerosol optical thickness and coarse-fine particle ratio related parameters are inverted. S4: Perform spatiotemporal matching and uniformity assessment on lidar data and multispectral polarization data to select effective synchronous observation points; S5: Based on the active and passive data of synchronous observation points, a collaborative inversion is performed. By iteratively optimizing the inversion parameters, further corrected retrieval results are obtained to achieve the optimal inversion of the three-dimensional spatiotemporal distribution of the Earth and atmosphere, and to obtain high-precision three-dimensional spatiotemporal distribution parameters of the Earth and atmosphere. S6: Using the aforementioned three-dimensional spatiotemporal distribution parameters of the Earth and atmosphere, perform atmospheric correction on the remote sensing data of the high-resolution imaging payload.

2. The atmospheric correction instrument data processing method using active-passive coordinated remote sensing according to claim 1, characterized in that, The multispectral polarization camera of the synchronous atmospheric corrector in S1 has several observation spectral bands, some of which have polarization detection capabilities, used to acquire the total reflectance and polarization reflectance data of the observed target during on-orbit synchronous observation.

3. The atmospheric correction instrument data processing method using active-passive coordinated remote sensing according to claim 1, characterized in that, The lidar of the synchronous atmospheric corrector in S1 adopts a dual-beam, dual-wavelength detection method to invert the profiles of aerosol and cloud backscattering coefficients and extinction coefficients.

4. The atmospheric correction instrument data processing method using active-passive coordinated remote sensing according to claim 1, characterized in that, The lidar data inversion of S2 adopts the Fernald two-component algorithm, which is adjusted and optimized by preset initial parameters and combined with subsequent collaborative inversion results.

5. The atmospheric correction instrument data processing method using active-passive coordinated remote sensing according to claim 1, characterized in that, In the multispectral polarization data inversion process of S3, the atmospheric aerosol parameter lookup table is constructed by a three-dimensional radiative transfer model. During the traversal process, the error of the combination of ground and atmospheric parameters and sensor data is quantified by setting a cost function, and finally the combination with the smallest error is selected to achieve high-precision inversion of aerosol optical thickness and coarse and fine particle ratio.

6. The atmospheric correction instrument data processing method using active-passive coordinated remote sensing according to claim 1, characterized in that, The surface polarization reflectance model in S3 includes: using the Nadal & Bréon semi-empirical model, combining land cover classification and NDVI values ​​to determine model parameters, thereby achieving the separation of surface polarization contribution from atmospheric aerosol contribution.

7. The atmospheric correction instrument data processing method using active-passive coordinated remote sensing according to claim 1, characterized in that, The spatiotemporal matching in S4 includes: determining the best matching pixel by calculating the latitude and longitude deviation between the pixels of the multi-beam lidar and the multispectral polarization camera, and selecting the corresponding deviation calculation method according to the dimensionality.

8. The atmospheric correction instrument data processing method using active-passive coordinated remote sensing according to claim 1, characterized in that, The uniformity judgment in S4 is evaluated by the cloud ratio of the passive instrument and the standard deviation of the cloud top height deviation and cloud bottom height deviation of the lidar profile.

9. The atmospheric correction instrument data processing method using active-passive coordinated remote sensing according to claim 1, characterized in that, The three-dimensional profile constraints in the active and passive data collaborative inversion of S5 include constraining the range of atmospheric parameters under cloud cover and high aerosol concentration conditions to ensure the effectiveness of the inversion results. Based on this, the aerosol optical thickness inverted from multispectral polarization data is used as a reference to adjust the lidar ratio until the aerosol optical thickness inversion error of the two types of data meets the preset range.

10. A method for processing atmospheric correction instrument data using active-passive coordinated remote sensing according to claim 1, characterized in that, The three-dimensional spatiotemporal distribution parameters of the Earth and atmosphere in S6 include the vertical profile of atmospheric composition, aerosol optical thickness, coarse and fine particle ratio, column water vapor content, and thin cirrus cloud-related parameters. As key three-dimensional information of Earth and atmosphere parameters for high-precision adaptation, these parameters support atmospheric correction of high spatial resolution remote sensing data such as airborne main payloads and satellite payloads, thereby improving the quality of remote sensing images and the accuracy of quantitative parameters.

Images