Satellite-borne polarization SMAP-R data gap filling method based on multi-polarization SAR data

Abstract

The invention discloses a space-borne polarization SMAP-R data gap filling method based on multi-polarization SAR data. The method comprises the following steps: acquiring SMAP-R, SMAP, GF-3, Sentinel-1, MODIS and SR TM 30m DEM data; preprocessing the data to obtain a total intensity reflectivity, a normalized Stokes parameter and a backscattering coefficient; unifying data spatial resolution and dividing data sets; building a data supplement method model of the multi-polarization SAR back scattering modeling SMAP-R polarization parameters based on integrated learning; verifying the parameter inversion performance by using an ANN model; and finally filling the missing grid data. According to the method, modeling is carried out on SMAP-R polarization data through SAR back scattering data, secondary inversion verification is carried out, SMAP-R spatial data missing is effectively supplemented, and the spatial resolution of inversion earth surface parameters is improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of space downscaling of spaceborne polarimetric GNSS-R technology, and particularly relates to supplementing and enhancing SMAP-R data using multi-source data fusion, and designing different model schemes for different secondary inversion requirements. BACKGROUND

[0002] The SMAP mission provides complementary L-band measurements by radiometer and radar to obtain geophysical measurements with fine spatial resolution. However, in 2015, the radar transmitter stopped working due to a fault, and by switching the band-pass center frequency of its radar receiver to the GPS L2C band, the SMAP-R data has become the largest polarimetric GNSS-R data set after more than seven years of continuous measurement. SMAP can receive reflected signals from the ground surface through the V channel and the H channel, and it is the only data set that provides polarimetric measurements. Compared with other GNSS-R missions, the unique feature of SMAP-R is the high-gain antenna and its hybrid compact polarimetric (HCP) capability. The SMAP-R data is collected in the in-phase and quadrature (I / Q) samples, and after processing, the forward scattering measurements generated by GPS reflection are extracted in the form of four Stokes parameters (S0, S1, S2, S3) and total intensity reflectivity (Γ0) calibrated by S0. The SMAP-R data set has great potential in the inversion research of surface parameters (vegetation, soil moisture, polar region), although SMAP-R has the advantage of high temporal resolution, due to the weak reflection signal, uneven distribution of reflection point position, and non-coherent accumulation time, and the observation points are sparse track points, resulting in many "gaps" or incomplete areas in space. The average spatial resolution of SMAP-R is 2.8 kilometers for wetlands, 5.3 kilometers for arid lands, 12.1 kilometers for low vegetation, and 26.6 kilometers for very dense vegetation. Therefore, in order to obtain higher spatial resolution of surface parameter inversion results, we need to supplement the SMAP-R data with other data. SAR data can provide high-resolution images, and there is a certain relationship between the brightness temperature of the L-band radiometer and SAR. In order to fully utilize the ability of SMAP-R to invert earth parameters and ensure its spatial resolution, it is necessary to supplement the sparse data of SMAP-R, so as to realize the inversion of soil moisture, vegetation optical thickness, vegetation water content, polar region parameters and other high temporal and spatial resolution. Therefore, by modeling the backscattering (SAR) of the forward scattering (SMAP-R), the SMAP-R data is supplemented to solve the problem of the lack of polarimetric SMAP-R data at the medium (or small) scale, which is the technical solution proposed in the present application.

[0003] In view of this, the present application is proposed. SUMMARY

[0004] The present invention aims to propose a method for filling gaps in spaceborne polarimetric SMAP-R data based on multi-polarimetric SAR data. This method uses SAR backscatter data to model the SMAP-R polarimetric data, and then verifies the performance of this data through secondary inversion, thereby supplementing the missing data in SMAP-R space and improving the spatial resolution of SMAP-R inverted surface parameters.

[0005] The present invention provides the following technical solutions:

[0006] A method for filling gaps in spaceborne polarimetric SMAP-R data based on multi-polarimetric SAR data comprises the following steps:

[0007] Step S1: Obtain SMAP-R, SMAP, GF-3, Sentinel-1, MODIS, and SRTM 30m DEM data.

[0008] Step S2: preprocess the SMAP-R, SAR data and auxiliary data to obtain the total intensity reflectivity and normalized Stokes parameters as well as dual-polarization (Sentinel-1) and full-polarization (GF-3) backscatter coefficients.

[0009] Step S3: unify the spatial resolution of all data and perform data division.

[0010] Step S4: constructing a data supplementation method model for SMAP-R polarization parameters of multi-polarization SAR backscatter modeling based on ensemble learning.

[0011] Step S5: Verify the parameter inversion performance based on the ANN model.

[0012] Step S6: filling in missing grid data.

[0013] Preferably, the data preparation in step S1 includes the following:

[0014] Download SMAP-R, SMAP, GF-3, Sentinel-1, MODIS, and SRT M 30m DEM data from official websites and institutions.

[0015] Preferably, step S2 pre-processes the SMAP-R, SAR data, and auxiliary data to obtain total intensity reflectivity, normalized Stokes parameters, and dual-polarization (Sentinel-1) and full-polarization (GF-3) backscatter coefficients. The process includes the following sub-steps:

[0016] In step S2.1, the SAR data (GF-3, Sentine-1) were first radiometrically corrected, filtered, geometrically corrected, and georeferenced using QGIS and SNAP / PolSARpro combined with DEM data.

[0017] In step S2.2, Freeman–Durden and Cloude–Pottier decompositions are performed on the GF-3 full polarimetric data scattering matrix and eigenvalues.

[0018] The Freeman-Durden method is based on the incoherent decomposition of the model. It models the polarization coherence matrix as the contribution of three scattering mechanisms: surface scattering, double-bounce scattering, and volume scattering. In the SAR image, each pixel is represented by a 3x 3 coherence matrix T

[0019]

[0020] The superscript * indicates conjugation.

[0021] Freeman expresses the coherence matrix as:

[0022] T=P s T sueface +P d T double +P v T volume (2)

[0023] Among them, P s 、P d and P v Corresponding to the power of each scattered component, T sueface , T double , T volume They are surface scattering, double-bounce scattering and volume scattering, and their sum is equal to Span:

[0024] Span=P s +P d +P v =T 11 +T 22 +T 33 (3)

[0025] Surface scattering is modeled by a first-order Bragg surface scatterer as follows:

[0026]

[0027] Where β is the ratio of horizontally polarized waves to vertically polarized waves in the polarization scattering matrix component under surface scattering.

[0028] Double-bounce scattering is modeled as follows:

[0029]

[0030] Where χ is the ratio of horizontally polarized waves to vertically polarized waves in the polarization scattering matrix component under double-reflection scattering.

[0031] Volumetric scattering is modeled as follows:

[0032]

[0033] The Cloude–Pottier method is based on the eigenvalue analysis of the polarization coherence matrix, and the decomposition results are angle (α), entropy (H), and anisotropy (A).

[0034] α=p1a1+p2a2+p3a3 (7)

[0035]

[0036] Where a i and p i (i = 1, 2, 3) represents the type and probability of the three scattering mechanisms. α, H, and A can characterize the complexity and dominance of the scattering process.

[0037] Step S2.3, obtain the total intensity reflectivity (Γ0) and the normalized first second third Stokes parameter, calculated as:

[0038]

[0039] Where σ(τ, f) is the bistatic radar cross section (BRCS) calibrated by the first Stokes parameter, R T is the distance from the emitter to the mirror reflection point, R R is the distance from the mirror reflection point to the receiver; (τ, f) represents the delay Doppler frequency. The above formula calculates the integral within the 5 delay and ±1.8kHz frame.

[0040] Step S2.4, extract other auxiliary data parameters: soil moisture (SM), surface roughness (Roughness), vegetation water content (VWC), vegetation optical depth (VOD), and surface temperature (RC) from SMAP data; and enhanced vegetation index (EVI) and normalized difference vegetation index (NDVI) from MODIS data.

[0041] Preferably, the step S3 of unifying the spatial resolution of all data and performing data division includes the following:

[0042] The selected area's SMAP-R, SMAP, MODIS, GF-3, and Sentine-1 data were resampled using bilinear interpolation and gridded onto an EASE Grid 2.09 km x 9 km grid, generating a dataset containing the same number of grid points as the target area. A grid search was used to average SMAP-R reflection points that fell within the same grid, and grids without SMAP-R reflection points were labeled. The dataset was randomly partitioned: 60% of the data was used for training, 25% for testing, and 15% for validation.

[0043] Preferably, the data supplementation method model for multi-polarization SAR backscatter modeling SMAP-R polarization parameters based on ensemble learning in step S4 includes the following:

[0044] The gradient boosting-based integrated machine learning CatBoost algorithm uses the full polarization (GF-3) and dual polarization (Sentinel-1) backscattering coefficients and auxiliary data to calculate the SMAP-R polarimetric parameters (total intensity reflectivity Γ0 and normalized three Stokes parameters) ) for modeling. The specific steps are: take all filtered and normalized data including (GF-3 and Sentinel-1 backscatter coefficients; SMAP SM, RC, VWC, VOD, Roughness; MODIS NDVI, EVI) as input, and use SMAP-R polarization parameters As output, different input features are combined according to different secondary inversion requirements. After training is completed, SMAP-R parameters with different secondary inversion capabilities are obtained from the inversion of SAR backscatter coefficients.

[0045] Preferably, the parameter inversion performance verification based on the ANN model in step S5 includes the following contents:

[0046] The parameters with different secondary inversion capabilities inverted in step S4 The corresponding auxiliary parameters are input into the ANN model for training and the results are compared and verified with the SMAP data. The selected accuracy verification indicators are correlation coefficient (CC) and root mean square error (RMSE).

[0047] Preferably, the missing mesh data filling in step S6 includes the following: after the modeling is completed, the missing meshes in the selected area are filled with the inverted data according to the previously marked missing meshes. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings of the implementation cases.

[0049] Figure 1 Schematic diagram of a flow chart of an embodiment of the present invention.

[0050] Figure 2 Schematic diagram of the structure of the ANN model in an embodiment of the present invention.

[0051] Figure 3 This is a comparison result between the vegetation optical depth inverted by the technical solution proposed in the embodiment of the present invention and the reference data. This figure uses a color figure to more intuitively reflect the comparison result.

[0052] Figure 4 This is a comparison result between the vegetation moisture content inverted by the technical solution proposed in the embodiment of the present invention and the reference data. This figure uses a color figure to more intuitively reflect the comparison result.

[0053] Figure 5 This is a comparison result between the soil moisture inverted by the technical solution proposed in the embodiment of the present invention and the reference data. This figure uses a color figure to more intuitively reflect the comparison result. DETAILED DESCRIPTION

[0054] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0055] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0056] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a single or selective embodiment that is mutually exclusive of other embodiments. The present invention provides the following embodiments.

[0057] Example 1

[0058] To verify the feasibility and reliability of the SMAP-R data inverted by the present invention, the present invention first completed the reconstruction of SMAP-R parameters from SAR data through ensemble learning, and then used the reconstructed data to invert vegetation optical depth, vegetation water content, and soil moisture. The specific implementation case is as follows: download SMAP-R data from the relevant public website; SMAP data (SM, Roughness, RC, VOD, VWC); MODIS data (MOD13C2 NDVI, EVI), the data time is: January 1, 2017 to December 31, 2020. Among them, SMAP data is used as reference data. GF-3 fully polarized surface scattering, double-bounce scattering, volume scattering coefficient, angle, entropy, anisotropy; Sentinel-1 vertical polarization and cross-polarization backscattering coefficients; SMAP SM, Roughness, VWC, RC, VOD; MODIS EVI and NDVI data are used as input data for model inversion; SRTM 30mDEM is used for geometric correction of SAR data processing. The temporal and spatial resolution information of the six types of data used in the present invention and the parameters used are shown in Table 1:

[0059] Table 1 Time and space resolution information of six types of data used in the present invention

[0060]

[0061] A method for filling gaps in spaceborne polarimetric SMAP-R data based on multi-polarimetric SAR data comprises the following steps:

[0062] Step S1: Obtain SMAP-R, SMAP, GF-3, Sentinel-1, MODIS, and SRTM 30m DEM data.

[0063] Step S2: preprocess the SMAP-R, SAR data and auxiliary data to obtain the total intensity reflectivity and normalized Stokes parameters as well as dual-polarization (Sentinel-1) and full-polarization (GF-3) backscatter coefficients.

[0064] Step S3: unify the spatial resolution of all data and perform data division.

[0065] Step S4: constructing a data supplementation method model for SMAP-R polarization parameters of multi-polarization SAR backscatter modeling based on ensemble learning.

[0066] Step S5: filling in missing grid data.

[0067] Step S6: Verify the parameter inversion performance based on the ANN model.

[0068] As an implementation of this embodiment, step S2 includes the following sub-steps:

[0069] Step S2.1: Preprocess the SAR data (GF-3, Sentine-1). Import the Sentinel-1 data into SNAP. Use Calibration to output the backscatter coefficients of the VV and VH channels. Then, use the Refined Lee filter to despeckle and filter the data. Select a 5 x 5 window. Then, input the spliced ​​DEM for geometric correction, and select the WGS84 coordinate system. Import the GF-3 data into PolSARpro. Select Calibration for polarization channel amplitude correction and select normalized output. Next, filter the covariance matrix using the Refined Lee filter. Input the filtered data into SNAP, perform geometric correction using the DEM, and select WGS84 as the output coordinate system.

[0070] In step S2.2, Freeman–Durden and Cloude–Pottier decompositions are performed on the GF-3 full polarimetric data scattering matrix and eigenvalues.

[0071] The Freeman-Durden method is based on the incoherent decomposition of the model. It models the polarization coherence matrix as the contribution of three scattering mechanisms: surface scattering, double-bounce scattering, and volume scattering. In the SAR image, each pixel is represented by a 3x 3 coherence matrix T

[0072]

[0073] The superscript * indicates conjugation.

[0074] Freeman expresses the coherence matrix as:

[0075] T=P s T sueface +P d T double +P v T volume (2)

[0076] Among them, P s 、P d and P v Corresponding to the power of each scattered component, T sueface , T double , T volume They are surface scattering, double-bounce scattering and volume scattering, and their sum is equal to Span:

[0077] Span=P s +P d +P v =T11 +T 22 +T 33 (3)

[0078] Surface scattering is modeled by a first-order Bragg surface scatterer as follows:

[0079]

[0080] Where β is the ratio of horizontally polarized waves to vertically polarized waves in the polarization scattering matrix component under surface scattering.

[0081] Double-bounce scattering is modeled as follows:

[0082]

[0083] Where χ is the ratio of horizontally polarized waves to vertically polarized waves in the polarization scattering matrix component under double-reflection scattering.

[0084] Volumetric scattering is modeled as follows:

[0085]

[0086] The Cloude–Pottier method is based on the eigenvalue analysis of the polarization coherence matrix, and the decomposition results are angle (α), entropy (H), and anisotropy (A).

[0087] α=p1a1+p2a2+p3a3 (7)

[0088]

[0089] Where a i and p i (i = 1, 2, 3) represents the type and probability of the three scattering mechanisms. α, H, and A can characterize the complexity and dominance of the scattering process.

[0090] Step S2.3, obtain the total intensity reflectivity (Γ0) and the normalized first second third Stokes parameter, calculated as:

[0091]

[0092] Where σ(τ, f) is the bistatic radar cross section (BRCS) calibrated by the first Stokes parameter, R T is the distance from the emitter to the mirror reflection point, R R is the distance from the mirror reflection point to the receiver; (τ, f) represents the delay Doppler frequency. The above formula calculates the integral within the 5 delay and ±1.8kHz frame.

[0093] Step S2.4: Preprocess other auxiliary data parameters: Extract SM, Roughness, VWC, VOD, and RC from the SMAP data; and EVI and NDVI from the MODIS data. Perform quality control on these data: assign NAN to SM values ​​less than 0.02 and equal to -9999, assign NAN to VWC, VOD, Roughness, and RC values ​​equal to -9999, and use the SMAP data's inherent quality flag to assign NAN to values ​​with quality flags of 0 and 8. Assign NAN to NDVI values ​​less than -2000 or greater than 10000. Use the MODIS data's inherent quality flag to filter out data with bits 0 and 1 not set to 0.

[0094] As an implementation of this embodiment, step S3 includes the following contents:

[0095] The SMAP-R, SMAP, MODIS, GF-3, and Sentinel-1 data for the selected area were resampled using bilinear interpolation and gridded onto an EASE Grid 2.09 km x 9 km grid, generating a dataset containing the same number of grid points as the target area. A grid search method was used to average the SMAP-R reflection points that fell within the same grid, and grids without SMAP-R reflection points were labeled. The dataset was randomly partitioned: 60% of the data was used for training, 25% for testing, and 15% for validation.

[0096] As an implementation of this embodiment, step S4 includes the following contents:

[0097] The SMAP-R polarimetric parameters (total intensity reflectance and normalized three Stokes parameters) were modeled using full polarization (GF-3) and dual polarization (Sentinel-1) backscatter coefficients and auxiliary data. The selected ensemble learning method was CatBoost. The specific training process was as follows: all filtered and normalized data including (GF-3 and Sentinel-1 backscatter coefficients; SMAP SM, Roughness, VWC, VOD, RC; MODIS NDVI, EVI) were used as input, and SMAP-R as output.

[0098] According to the different requirements of secondary inversion of filled data and to prevent overfitting, the input scheme is designed as follows:

[0099]

[0100] As an implementation of this embodiment, step S5 includes the following sub-steps:

[0101] Step S5.1: The SMAP-R parameters with different secondary inversion capabilities inverted from the SAR backscatter coefficients after training in step S4 are The data are input into the ANN network for secondary inversion, with the target parameters being VWC, VOD and SM.

[0102] Step S5.2, for the ANN model, as shown in the attached Figure 2 As shown in the figure, the model consists of 8 fully connected layers, with the number of neurons in each layer being 32, 64, 64, 32, 64, 128, 128, and 64, respectively. The activation function between layers 1-6 is uniformly ReLU function, and no activation function is used in layers 7-8. Dropout mechanism is added in the middle to suppress overfitting, and the Dropout ratio is 0.1. The output layer uses Linear function to enhance output stability. The optimization algorithm uses Adam optimizer, and the initial learning rate is 0.001. The loss function uses mean square error (MSE), which is defined as follows:

[0103]

[0104] Where y i represents the observed value of the i-th sample, represents the predicted value of the model, and n is the total number of samples.

[0105] In step S5.3, the results are compared and validated with the SMAP data. During training, if the validation set error does not significantly decrease over 10 consecutive iterations, training is terminated early. After constructing a complete regional training sample, the overall model is applied to predict missing grids, achieving a complete reconstruction of the spatial distribution of SMAP-R inversion parameters.

[0106] In step S5.,4, the selected accuracy indicators are correlation coefficient (CC) and root mean square error (RMSE).

[0107]

[0108] Where n is the number of data samples, is the true value of the four parameters of SMAP-R, y i is the predicted value of the model, They are and y i The average value of .

[0109] As an implementation of this embodiment, step S6 includes the following: after completing the modeling and verification, the missing meshes in the selected area are filled with the inverted data according to the previously marked missing meshes.

[0110] Finally, the results of secondary inversion using ANN are as follows Figure 3-Figure 5 As shown in the figure, the scatter density maps of vegetation optical thickness, vegetation water content and soil moisture inverted from SMAP-R data are given, as well as the correlation coefficient and root mean square error.

[0111] As can be seen from the figure, there is a strong correlation between the three inversion results and the validation data (inverted vegetation optical depth CC = 0.89, vegetation water content CC = 0.88, soil moisture CC = 0.83), and the inverted soil moisture shows the smallest RMSE = 0.054cm 3 / cm 3 , RMSE of vegetation optical depth = 0.076, RMSE of vegetation water content = 0.645 kg / m 2 The above results prove that SMAP-R data modeled by SAR backscatter data can successfully invert surface parameters with good accuracy and can be used as a substitute for missing data, thereby filling the gaps in the SMAP-R inversion parameter space and improving spatial resolution.

[0112] The above content is a further detailed description of the present invention in conjunction with specific implementation methods. It cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, they can make several simple deductions or substitutions without departing from the concept of the present invention, which should be regarded as falling within the scope of protection determined by the claims submitted for the present invention.

Claims

1. A method for filling gaps in spaceborne polarimetric SMAP-R data based on multi-polarimetric SAR data, characterized in that: The following steps are involved: Step S1, acquiring SMAP-R data, SMAP data, GF-3 fully polarized SAR data, Sentinel-1 dual-polarized SAR data, MODIS vegetation index data, and SRTM 30m DEM terrain data; Step S2: Preprocess the SMAP-R data, SMAP data, GF-3 fully polarized SAR data, Sentinel-1 dual-polarized SAR data, MODIS vegetation index data, and SRTM 30m DEM terrain data to extract the total intensity reflectance and normalized Stokes parameters of SMAP-R, obtain the surface scattering, double-bounce scattering, and volume scattering components and entropy, anisotropy, and angle of the GF-3 fully polarized data, extract the vertical polarization backscattering coefficient and cross-polarization backscattering coefficient of Sentinel-1, and integrate the soil moisture (SM), surface roughness (Roughness), vegetation water content (VWC), vegetation optical depth (VOD), and land surface temperature (RC) of SMAP and the enhanced vegetation index (EVI) and normalized difference vegetation index (NDVI) of MODIS; Step S3: Resample the SMAP-R, SMAP, MODIS, GF-3, and Sentine-1 data to the EASE Grid 2.09 km × 9 km standard grid and divide them into training, test, and validation sets; Step S4: Based on the ensemble learning model, the inversion model of SMAP-R polarization parameters is constructed using the SAR backscatter coefficient, SMAP SM, surface roughness, VWC, VOD, RC, and MODIS EVI and NDVI; Step S5: Secondary inversion of vegetation parameters and soil moisture through artificial neural network (ANN) to verify data validity; Step S6: Fill in the spatially missing grids of the SMAP-R data based on the model output.

2. The method for filling gaps in spaceborne polarimetric SMAP-R data based on multi-polarimetric SAR data according to claim 1, characterized in that: The pre-processing of step S2 includes: Freeman-Durden decomposition was performed on the GF-3 data to obtain surface scattering, double-bounce scattering, and volume scattering components; entropy H, anisotropy A, and scattering angle α were extracted through Cloude-Pottier decomposition; normalized Stokes parameters were calculated for the SMAP-R data; and the Sentinel-1 data were radiometrically corrected and geometrically aligned before extraction of vertically and cross-polarized backscattering coefficients.

3. The method for filling gaps in spaceborne polarimetric SMAP-R data based on multi-polarimetric SAR data according to claim 1, characterized in that: The data division ratio in step S3 is: The training set accounts for 60%, the test set accounts for 25%, and the validation set accounts for 15%; the grids with missing SMAP-R data are marked with null values.

4. The method for filling gaps in spaceborne polarimetric SMAP-R data based on multi-polarimetric SAR data according to claim 1, characterized in that: The ensemble learning model in step S4 is the CatBoost algorithm, and the input features include: Full polarization mode: T surface 、T double 、T volume ,α,H,A,SM,Roughness,RC,VOD,VWC,EVI,NDVI; Dual polarization mode: σ VV , σ VH , SM, Roughness, VWC, VOD, RC, EVI, NDVI.

5. The method for filling gaps in spaceborne polarimetric SMAP-R data based on multi-polarimetric SAR data according to claim 1, characterized in that: The ANN model in step S5 includes an 8-layer fully connected network, the activation function uses ReLU, the loss function is the mean square error MSE, the optimization algorithm uses the Adam optimizer, and the output layer inversion parameters include VOD, VWC and SM.

6. The method for filling gaps in spaceborne polarimetric SMAP-R data based on multi-polarimetric SAR data according to claim 1, characterized in that: The data filling in step S6 is specifically as follows: The inverted Γ0, The parameters are mapped to the EASE Grid, and the marked missing grids are filled by interpolation to generate a spatially continuous SMAP-R polarimetric dataset.

7. A method for filling gaps in spaceborne polarimetric SMAP-R data based on multi-polarimetric SAR data according to claims 1-6, characterized in that: The method is suitable for improving the spatial resolution of SMAP-R data in vegetation monitoring, soil moisture inversion and polar parameter extraction, and filling the spatial gaps caused by the sparse distribution of the original data.

Images