A method for temporal and spatial matching between spaceborne SAR sea surface images and gridded sea surface wind field products

Abstract

The present invention provides a method for space-based SAR sea surface image and gridded sea surface wind field product time-space matching. The method includes: obtaining SAR single-view complex image and metadata file; calculating DN (Digital Number) value of SAR single-view complex image; calculating backscatter coefficient (sigma0) of SAR single-view complex image by calibration coefficient; masking land data by mask data; obtaining gridded sea surface wind products at two moments before and after SAR imaging time; parsing coordinates and time information of matched gridded sea surface wind field products before and after; interpolating gridded sea surface wind speed and wind direction to SAR observation time by interpolation method; calculating grid coordinates containing gridded sea surface wind field products in SAR image; calculating SAR image information with given facet size centered on grid coordinate point; repeating the above time-space matching process, completing matching work of all grid coordinates in SAR image one by one; and finally outputting time-space matching number sets of two types of products. The method makes the successfully matched gridded wind field product and SAR image product unique.

Description

Technical Field

[0001] The invention relates to the technical field of marine remote sensing data processing, and in particular to a method for temporal and spatial matching of satellite-borne SAR sea surface images and gridded sea surface wind field products. Background Art

[0002] As an active microwave sensor, spaceborne SAR (synthetic aperture radar) obtains high-resolution sea surface images based on the interaction mechanism between electromagnetic waves and sea surface microscale waves. This product has the advantages of high spatial resolution, large coverage and high observation timeliness. Gridded sea surface wind field products are usually based on multi-source observation data fusion technology or integrated forecasting system combined with assimilation technology. This product has the advantages of global coverage and uniform spatial and temporal distribution. The use of effective spatial and temporal matching technology in the above two products will be conducive to the evaluation of sea surface wind field product accuracy indicators and geophysical model modeling research. Summary of the invention

[0003] The purpose of the present invention is to provide a method for spatiotemporal matching of satellite-borne SAR (Synthetic Aperture Radar) sea surface image products and gridded sea surface wind field products. Instead of searching for gridded sea surface wind field products based on the center coordinates of SAR sub-image pixels, the SAR image data is searched based on the coordinates of the gridded sea surface wind field products. This is conducive to making the elements matched between the gridded sea surface wind field products and the matching SAR sub-image pixels unique.

[0004] To achieve the above object, the technical solution of the present invention is implemented by the following steps: A method for spatiotemporal matching of SAR sea surface image products and gridded sea surface wind field products comprises the following steps:

[0005] Step (1), obtaining regional SAR single-view complex images and metadata information;

[0006] Step (2), using the information obtained in step (1) to calculate the DN value of the SAR data for the SAR single-view complex image data;

[0007] Step (3), geo-locating the SAR DN image data using the RPC coefficients obtained from the metadata;

[0008] Step (4), calibrating the SAR DN value obtained in step (3) using the calibration coefficient obtained from the metadata to obtain calibrated SAR data;

[0009] Step (5), using the land mask data, performing sea and land mask processing on the SAR data calibrated in step (4), and marking sea and land identifiers;

[0010] Step (6), using the prepared gridded sea surface wind field product data set to obtain the gridded sea surface wind field products before and after the SAR imaging time;

[0011] Step (7), analyzing the coordinate information and time information of the front and rear gridded sea surface wind field products matched in step (6);

[0012] Step (8), interpolating the front and rear gridded sea surface wind field product information matched in step (7) to the SAR observation time using a linear interpolation method;

[0013] Step (9), obtaining the geographic coordinate information of the four corner points of the SAR image from the SAR image metadata, calculating the grid coordinates of the gridded sea surface wind field product that falls within the SAR image based on the geographic coordinates of the four corner points of the SAR image, extracting one of the grid point coordinate pairs, and searching for SAR image elements that meet the conditions along the longitudinal direction and the latitudinal direction respectively, taking the grid point coordinates as the center, combining the radar incident angle, the spatial resolution information of the SAR image, and the SAR sub-image pixel size conditions.

[0014] Step (10), repeat step 9 until all grid point coordinates are matched, and finally output the spatiotemporal matching data set of the two types of products.

[0015] Furthermore, in step (1), the image data of the SAR single-view complex image is parsed to obtain the real part and the imaginary part; the metadata is parsed to obtain the azimuth width, range width, azimuth resolution, range resolution, incident angle, calibration coefficient, imaging time, radar line of sight, four-corner coordinates, and RPC coefficient of the SAR image.

[0016] Furthermore, the specific calculation formula of step (2) is: Here, P and Q correspond to the real and imaginary values ​​of the SAR image, respectively.

[0017] Furthermore, the specific geographic positioning method in step (3) is: geocoding the SAR data using the RPC coefficients and the row and column information of the elements,

[0018] The RPC model (rational polynomial model) is applicable to different types of sensors and does not require imaging geometric parameters. It associates the geodetic coordinates of ground points with their corresponding pixel coordinates using a ratio polynomial.

[0019] Through the ratio polynomial, the RPC model constructs the relationship between the geodetic coordinates of the ground point (lat, lon, H) and its corresponding imaging point coordinates (sample, line), and enhances the stability of parameter solution by standardizing the ground coordinates and image coordinates to between -1 and 1. For regional spaceborne SAR images, the following ratio polynomial is defined:

[0020]

[0021] In the formula,

[0022] N L (P, L, H) = a1 + a2L + a3P + a4H + a5LP + a6LH + a7PH + a8L 2 + a9P 2 + a 10 H 2 + a 11 PLH + a 12 L 3 + a 13 LP 2 + a 14 LH 2 + a 15 L 2 P + a 16 P 3 + a 17 PH 2 + a 18 L 2 H + a 19 P 2 H + a 20 H 3

[0023] D L (P, L, H) = b1 + b2L + b3P + b4H + b5LP + b6LH + b7PH + b8L 2 + b9P 2 + b 10 H 2 + b 11 PLH + b 12 L 3 + b 13 LP 2 + b 14 LH 2 + b 15 L 2 P + b 16 P 3 + b 17 PH 2 + b 18 L 2 H + b 19 P 2 H + b 20 H 3

[0024] N S (P, L, H) = c1 + c2L + c3P + c4H + c5LP + c6LH + c7PH + c8L 2 + c9P 2+c 10 H 2 +c 11 PLH+c 12 L 3 +c 13 LP 2 +c 14 LH 2 +c 15 L 2 P+c 16 P 3 +c 17 PH 2 +c 18 L 2 H+c 19 P 2 H+c 20 H 3

[0025] D S (P,L,H)=d1+d2L+d3P+d4H+d5LP+d6LH+d7PH+d8L 2 +d9P 2 +d 10 h 2 +d 11 PLH+d 12 L 3 +d 13 LP 2 +d 14 LH 2 +d 15 L 2 P+d 16 P 3 +d 17 PH 2 +d 18 L 2 H+d 19 P 2 H+d 20 H 3

[0026] a i , b i 、c i d i are RPC model coefficients, where the sum is usually 1; (P, L, H) are the ground coordinates of (lat, lon, H) after standardized calculation, and (Y, X) are the image coordinates of (sample, line) after standardized calculation.

[0027] Furthermore, in step (4), the calibration coefficient is used to calculate the SAR image sigma0 (radar backscatter coefficient). The calibration formula is as follows:

[0028]

[0029] Where DN (Digital Number) is the calculated SAR image quantization value, qv and K are the image quantization maximum value and calibration constant, respectively.

[0030] Furthermore, the spatial resolution of the land mask data is 1.0 km × 1.0 km. The prepared land mask data with a spatial resolution of 1.0 km × 1.0 km is used to perform land and sea mask processing on the SAR data calibrated in step (4) to mark the land and sea identifications.

[0031] Furthermore, the specific implementation method of step (6) is: using the SAR imaging time, obtaining two gridded sea surface wind field product data before and after the SAR imaging time from the prepared gridded sea surface wind field data set.

[0032] Furthermore, the specific implementation method of step (8) is: interpolating and calculating the gridded wind field product of the SAR imaging time to obtain the wind speed component u value and the wind speed component v value at the observation time t.

[0033] Furthermore, given the wind speed component u0 at the grid point (lat, lon) at time t0 and the wind speed component u1 at time t1, the wind speed component u value at SAR observation time t in the interval [t0, t1] is calculated: the linear interpolation formula is as follows: The same method is used to calculate the wind speed component v value at the grid point (lat, lon) SAR observation time t.

[0034] Furthermore, in step (9), the grid point coordinates are taken as the center, and the radar incident angle, spatial resolution, SAR sub-image pixel size and other meta-information are combined to search for SAR image elements in the longitudinal direction and the latitudinal direction respectively.

[0035]

[0036] In the formula, row num is the total number of row elements, col num is the total number of column elements, cellsize y , cellsize x are the longitudinal and latitudinal sizes of the SAR sub-surface element, respectively, rg and reso az are the azimuth resolution and range resolution respectively, and θ is the radar incident angle.

[0037] Furthermore, the gridded sea surface wind field dataset described in step 6 includes:

[0038] Analytical wind field products, NCEP reanalysis wind field products, CCMP fused sea surface wind field products, forecast sea surface wind field products, etc.

[0039] The technical solution adopted by the present invention searches for SAR image data with the coordinates of the gridded sea surface wind field product as the center, so that the matching elements of the gridded sea surface wind field product and the matching SAR sub-image facets are unique. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 The figure is a schematic diagram of data processing flow of a specific embodiment of the present invention.

[0041] Figure 2 The figure is a schematic diagram of matching SAR image data of sea surface wind field grid point coordinates according to a specific embodiment of the present invention. DETAILED DESCRIPTION

[0042] The technical solution of the present invention is further explained below in conjunction with the accompanying drawings and specific embodiments.

[0043] The present invention provides a method for searching matching SAR image data based on the grid point coordinates of the gridded sea surface wind field product, such as Figure 1 As shown, the method of the present invention comprises the following steps:

[0044] Step 1: Obtain regional SAR single-view complex images and metadata information.

[0045] Specifically, the image data of the SAR single-view complex image is parsed to obtain the real part and the imaginary part; the metadata is parsed to obtain the azimuth width, range width, azimuth resolution, range resolution, incident angle, calibration coefficient, imaging time, radar line of sight, four-corner coordinates, and RPC coefficient of the SAR image.

[0046] Step 2, calculate the DN value of SAR (SAR image gray value), specifically, Here, P and Q correspond to the real and imaginary parts of the SAR image, respectively. Step 3: Use the RPC coefficients obtained from the metadata to geolocate the SAR DN image data.

[0047] Specifically, the SAR data is geocoded using the RPC coefficients and the row and column information of the elements.

[0048] The RPC model (rational polynomial model) is applicable to different types of sensors and does not require imaging geometric parameters. It associates the geodetic coordinates of ground points with their corresponding pixel coordinates using a ratio polynomial.

[0049] Through the ratio polynomial, the RPC model constructs the relationship between the geodetic coordinates of the ground point (lat, lon, H) and its corresponding imaging point coordinates (sample, line), and enhances the stability of parameter solution by standardizing the ground coordinates and image coordinates to between -1 and 1. For regional spaceborne SAR images, the following ratio polynomial is defined:

[0050]

[0051] In the formula,

[0052] N L (P,L,H)=a1+a2L+a3P+a4H+a5LP+a6LH+a7PH+a8L 2 +a9P 2 +a 10 H 2 +a 11 PLH+a 12 L 3 +a 13 LP 2 +a 14 LH 2 +a 15 L 2 P+a 16 P 3 +a 17 PH 2 +a 18 L 2 H+a 19 P 2 H+a 20 H 3

[0053] D L (P,L,H)=b1+b2L+b3P+b4H+b5LP+b6LH+b7PH+b8L 2 +b9P 2 +b 10 H 2 +b 11 PLH+b 12 L 3 +b 13 LP 2 +b 14 LH 2 +b 15 L 2 P+b 16 P 3 +b 17 PH 2 +b 18 L 2 H+b 19 P2 H+b 20 H 3

[0054] N S (P,L,H)=c1+c2L+c3P+c4H+c5LP+c6LH+c7PH+c8L 2 +c9P 2 +c 10 H 2 +c 11 PLH+c 12 L 3 +c 13 LP 2 +c 14 LH 2 +c 15 L 2 P+c 16 P 3 +c 17 PH 2 +c 18 L 2 H+c 19 P 2 H+c 20 H 3

[0055] D S (P,L,H)=d1+d2L+d3P+d4H+d5LP+d6LH+d7PH+d8L 2 +d9P 2 +d 10 H 2 +d 11 PLH+d 12 L 3 +d 13 LP 2 +d 14 LH 2 +d 15 L 2 P+d 16 P 3 +d 17 PH 2 +d 18 L 2 H+d 19 P 2 H+d 20 H 3

[0056] a i 、b i 、c i 、d iare RPC model coefficients, where the sum is usually 1; (P, L, H) are the ground coordinates of (lat, lon, H) after standardized calculation, and (Y, X) are the image coordinates of (sample, line) after standardized calculation.

[0057] Step 4, calibrating the SAR DN data obtained in step (3) using the calibration coefficient obtained from the metadata to obtain calibrated SAR data;

[0058] Specifically, the SAR image sigma0 is calculated using the calibration coefficient. The calibration formula is as follows:

[0059]

[0060] Where DN is the calculated intensity value, qv and K are the maximum image quantization value and calibration constant, respectively.

[0061] Step 5, using the prepared land mask data with a spatial resolution of 1.0 km × 1.0 km, perform sea and land mask processing on the SAR data calibrated in step (4) to mark sea and land identifications;

[0062] Step 6, using the prepared gridded sea surface wind field product dataset, obtain the gridded sea surface wind field products before and after the SAR imaging time;

[0063] Specifically, using the SAR imaging time, two gridded sea surface wind field product data before and after the SAR imaging time are obtained from the prepared gridded sea surface wind field data sets (ERA5 reanalysis wind field products, NCEP reanalysis wind field products, CCMP fused sea surface wind field products, forecast sea surface wind field products, etc.).

[0064] Step 7, analyzing the coordinate information and time information of the front and rear gridded sea surface wind field products matched in step (6);

[0065] Step (8), interpolate the front and rear gridded sea surface wind field product information matched in step (7) to the SAR observation time using the linear interpolation method

[0066] Specifically, given the wind speed component u0 at the grid point (lat, lon) at time t0 and the wind speed component u1 at time t1, calculate the wind speed component u at SAR observation time t in the interval [t0, t1]: The linear interpolation formula is as follows:

[0067]

[0068] The same method is used to calculate the wind speed component v value at the grid point (lat, lon) SAR observation time t.

[0069] Step 9: Match the grid point coordinates of the sea surface wind field product to the SAR image data.

[0070] Specifically, the coordinate information of the four corners of the SAR image is used to determine whether the coordinates of the sea surface wind field product of the grid point are contained in the SAR image. The geographic coordinate information of the four corners of the SAR image is obtained from the SAR image metadata. According to the geographic coordinates of the four corners of the SAR image, the grid coordinates of the gridded sea surface wind field product falling within the SAR image are calculated. One of the grid point coordinate pairs is extracted, and with the grid point coordinates as the center, combined with the radar incident angle, the spatial resolution information of the SAR image, and the SAR sub-image pixel size conditions, the SAR image elements that meet the conditions are searched in the longitudinal and latitudinal directions respectively.

[0071]

[0072]

[0073] In the formula, row num is the total number of row elements, col num is the total number of column elements, cellsize y , cellsize x are the longitudinal and latitudinal sizes of the SAR sub-surface element, respectively, rg and reso az are the azimuth resolution and range resolution respectively, and θ is the radar incident angle.

[0074] like Figure 2 As shown, the white dots 6 are schematic diagrams of grid coordinates of the gridded sea surface wind field product falling within the SAR image 2. The size of the white box 5 is a schematic diagram of the SAR image sub-surface. The white arrow 3 is the radial search direction, and the white arrow 4 is the latitudinal search direction. The intersection 1 of the vertical and horizontal lines is a schematic diagram of the grid coordinates of the gridded sea surface wind field.

[0075] Step (10), repeat the empty matching process of step (9) until the grid coordinates of all gridded sea surface wind field products in the SAR image are matched. It is considered that the spatial matching of this scene SAR data and the gridded sea surface wind field product is completed, and the matched data set is output.

Claims

1. A method for temporal and spatial matching of spaceborne SAR sea surface images and gridded sea surface wind field products, characterized by: The following steps are involved: Step (1), obtaining regional SAR single-view complex images and metadata information; Step (2), using the information obtained in step (1) to calculate the DN value of the SAR data for the SAR single-view complex image data; Step (3), using the RPC coefficients obtained from the metadata to geolocate the SAR DN value image data; Step (4), calibrating the SAR DN data obtained in step (3) using the calibration coefficient obtained from the metadata to obtain calibrated SAR data; Step (5), using the land mask data, performing sea and land mask processing on the SAR data calibrated in step (4), and marking sea and land identifiers; Step (6), using the prepared gridded sea surface wind field product data set to obtain the gridded sea surface wind field products before and after the SAR imaging time; Step (7), analyzing the coordinate information and time information of the front and rear gridded sea surface wind field products matched in step (6); Step (8), interpolating the front and rear gridded sea surface wind field product information matched in step (7) to the SAR observation time using a linear interpolation method; Step (9), obtaining geographic coordinate information of the four corner points of the SAR image from the SAR image metadata, calculating the grid coordinates of the gridded sea surface wind field product that falls within the SAR image based on the geographic coordinates of the four corner points of the SAR image, extracting one of the grid point coordinate pairs, and searching for SAR image elements that meet the conditions along the longitudinal direction and the latitudinal direction respectively, taking the grid point coordinates as the center, combining the radar incident angle, the spatial resolution information of the SAR image, and the SAR sub-image pixel size conditions; Step (10), repeat step 9 until all grid point coordinates are matched, and finally output the spatiotemporal matching data set of the two types of products.

2. The method for temporal and spatial matching of spaceborne SAR sea surface images and gridded sea surface wind field products according to claim 1, characterized in that: In step (1), the image data of the SAR single-view complex image is parsed to obtain the real part and the imaginary part; the metadata is parsed to obtain the azimuth width, range width, azimuth resolution, range resolution, incident angle, calibration coefficient, imaging time, radar line of sight, four-corner coordinates, and RPC coefficient of the SAR image.

3. The method for temporal and spatial matching of spaceborne SAR sea surface images and gridded sea surface wind field products according to claim 1, characterized in that: In step (4), the calibration coefficient is used to calculate the SAR image sigma0. The calibration formula is as follows: Where DN is the calculated intensity value, qv and K are the maximum image quantization value and calibration constant, respectively.

4. The method for temporal and spatial matching of spaceborne SAR sea surface images and gridded sea surface wind field products according to claim 1, characterized in that: The spatial resolution of the land mask data in step (5) is 1.0 km × 1.0 km.

5. The method for temporal and spatial matching of spaceborne SAR sea surface images and gridded sea surface wind field products according to claim 1, characterized in that: A more specific implementation of step (6) is: using the SAR imaging time, obtaining two gridded sea surface wind field product data before and after the SAR imaging time from the prepared gridded sea surface wind field data set.

6. The method for temporal and spatial matching of spaceborne SAR sea surface images and gridded sea surface wind field products according to claim 1, characterized in that: A more specific implementation method of step (8) is as follows: interpolation is performed to calculate the gridded wind field product of the SAR imaging time, and the wind speed component u value and the wind speed component v value at the observation time t are obtained.

7. The method for temporal and spatial matching of spaceborne SAR sea surface images and gridded sea surface wind field products as claimed in claim 6, characterized in that: The wind speed component u0 at the grid point (lat, lon) at time t0, the wind speed component u1 at time t1, calculate the wind speed component u value at SAR observation time t in the interval [t0, t1]: The linear interpolation formula is as follows: The same method is used to calculate the wind speed component v value at the grid point (lat, lon) SAR observation time t.

8. The method for temporal and spatial matching of spaceborne SAR sea surface images and gridded sea surface wind field products as claimed in claim 1, characterized in that: The gridded sea surface wind field data set described in step (6) includes: ERA5 reanalysis wind field products, NCEP reanalysis wind field products, CCMP fused sea surface wind field products, and forecast sea surface wind field products.

Images