Space-borne laser radar matching verification method based on spatial variability and flow field

Abstract

The invention discloses a space-borne laser radar matching verification method based on spatial variability and a flow field. The method comprises the following steps: calculating the spatial variability of water color remote sensing data of each month by adopting a multi-scale grid; traversing the grid points, and searching the maximum grid size of which the spatial variability is smaller than a threshold value; calculating a space matching window and setting a time matching window according to the size of the maximum grid; measuring time and position according to the in-situ data, searching satellite-borne laser radar data by adopting a maximum space matching window and a set time matching window, and correcting the position of the in-situ data hour by hour according to the corresponding ocean current speed; searching a corresponding space matching window according to the corrected position, and matching the space matching window with satellite-borne laser radar data; and performing quality control on the matched spaceborne data to form a spaceborne laser radar-in-situ data matching pair. According to the method, the matching of the satellite-borne data and the in-situ data can be realized, and the reasonability of space matching window selection is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of laser radar, and in particular relates to a satellite-borne laser radar matching and verification method based on spatial variability and flow field. Background Art

[0002] Spaceborne LiDAR is an active remote sensing instrument with advantages such as continuous observation day and night and global-scale detection, and plays an important role in ocean observation. Spaceborne LiDAR can invert ocean optical parameters such as particle backscattering coefficient and attenuation coefficient, and then estimate parameters such as chlorophyll concentration, phytoplankton biomass and particulate organic carbon content, which is of great significance for the study of marine ecosystems. In-situ data usually refers to precise data obtained through field measurements, which has a high degree of accuracy and reliability and can be used to verify remote sensing data, but its temporal and spatial distribution is uneven and its coverage is limited.

[0003] The Chinese patent with publication number CN103020478A discloses a method for verifying the authenticity of ocean color remote sensing products, which uses multiple spatial matching window sizes to calculate spatial variability and classify the matching results, but does not involve how to select the optimal spatial matching window based on spatial variability. The Chinese patent with publication number CN112699204A discloses a method and device for determining a spatial matching window, which uses a preset database to calculate a relative error sequence and determine a spatial matching window, but does not involve the calculation of spatial variability and data matching methods. Therefore, the method in the above invention cannot be directly used for matching satellite-borne laser radar with in-situ data.

[0004] In the existing satellite-borne lidar-in-situ data matching method, a fixed time and space matching window is usually used. The satellite data falling within the matching window is collected with the in-situ measurement as the center, and the average value of the in-situ data and the satellite data is used as a pair of matching data to complete the data matching. However, there are significant differences in the spatial variability of data in different regions. The fixed matching window cannot be applied to all regions, which may result in too few matching data pairs or significant errors. In addition, due to the time difference between the satellite-borne lidar data and the in-situ data, the ocean currents will cause the measurement points to move, resulting in certain uncertainties. Summary of the invention

[0005] The present invention provides a satellite-borne laser radar matching verification method based on spatial variability and flow field. The spatial variability of ocean parameters is evaluated by using water color remote sensing data, and then the spatial matching window size is obtained. The longitude and latitude of the in-situ data are corrected according to the ocean current speed to achieve matching between the satellite-borne laser radar and the in-situ data.

[0006] A space-borne laser radar matching verification method based on spatial variability and flow field comprises the following steps:

[0007] S1: Use multiple grid sizes to calculate the spatial variability of water color remote sensing data for each month;

[0008] S2: Traverse each grid point and search for the maximum grid size corresponding to a spatial variation that is less than the set threshold;

[0009] S3: Calculate the spatial matching window matrix according to the maximum grid size and set the temporal matching window to a fixed value;

[0010] S4: According to the measurement time and longitude and latitude of the in-situ data, the largest spatial matching window and time matching window are used to search the satellite-borne lidar data, find the corresponding hourly ocean current speed, and gradually correct the position of the in-situ data;

[0011] S5: according to the corrected measurement time and longitude and latitude of the in-situ data, find the corresponding spatial matching window, and match the satellite-borne lidar data with the in-situ data according to the spatial matching window;

[0012] S6: Perform quality control on the matched spaceborne lidar data to form a spaceborne lidar-in-situ data matching pair.

[0013] The specific process of step S1 is:

[0014] Get the monthly average water color remote sensing grid data, and for each grid point (x, y), select n nearby i ×n i Grid data, n i The values ​​of are recorded as n from small to large 1 ,n 2 ,…,n q ; Calculate the spatial variability of each grid point:

[0015]

[0016] In the formula, m is month and std is n i ×n i The standard deviation of all data in the grid, mean is n i ×n i The average value of all data in the grid.

[0017] The specific process of step S2 is:

[0018] For month m 0 and the grid points (x 0 ,y 0 ), the corresponding spatial variability is CV(x 0 ,y 0 ,m 0 ,n i), set a fixed threshold, and find the value that satisfies CV(x 0 ,y 0 ,m 0 ,n i ) is less than the maximum subscript i of threshold, denoted as i 0 , the corresponding maximum grid size is According to the above rules, all grid points of each month are traversed to obtain the maximum grid size matrix N(x, y, m).

[0019] The specific process of step S3 is as follows:

[0020] According to the obtained maximum grid size matrix N(x,y,m), calculate the spatial matching window of all grid points in each month:

[0021]

[0022] Where d is the spatial resolution of water color remote sensing grid data;

[0023] Set the time matching window size to t.

[0024] The specific process of step S4 is as follows:

[0025] The measurement time t for obtaining the in-situ data insitu and longitude and latitude (lat insitu ,lon insitu );

[0026] Set the spatial matching window s and use the haversine algorithm to calculate the longitude and latitude (lat insitu ,lon insitu ) and the surface spot position of the satellite-borne laser radar (lat lidar ,lon lidar ) 0 , find the 0 <s of space-borne lidar data;

[0027] According to the time matching window t and the satellite-borne laser radar measurement time t lidar , find the data satisfying |t in the data satisfying the spatial matching window s insitu -t lidar |<t of space-borne lidar data;

[0028] According to the measurement time t of the in-situ data insitu and longitude and latitude (lat insitu ,lon insitu ), find the corresponding ocean current speed

[0029] If insitu<t lidar , set the timestamp t 0 ,t insitu ,t 1 ,t 2 ,…,t p ,t lidar ,t p+1 , where t 1 ,t 2 ,…,t p Between t insitu and t lidar All consecutive integer hours between 0 is less than t insitu The maximum integer hour moment, t p+1 is greater than t lidar The minimum whole hour of the moment; t u-1 Time to t u Time, longitude and latitude (lat u ,lon u )(u=1,2…,p) at the average ocean current velocity;

[0030] If insitu >t lidar , set the timestamp t 0 ,t lidar ,t 1 ,t 2 ,…,t p ,t insitu ,t p+1 , where t 1 ,t 2 ,…,t p Between t insitu and t lidar All consecutive integer hours between 0 is less than t lidar The maximum integer hour moment, t p+1 is greater than t insitu The minimum whole hour of the moment; t u-1 Time to t u Time, longitude and latitude (lat u ,lon u ) at the average ocean current velocity;

[0031] If insitu =t lidar , directly output the latitude and longitude of the original data;

[0032] According to the ocean current speed and time difference, the displacement distance is calculated, and then according to the direction of the ocean current speed, the longitude and latitude of the original data are updated to (latu ,lon u ), where: (lat 1 ,lon 1 )=(lat insitu ,lon insitu );

[0033] Finally, the output latitude and longitude of the original data is: (lat correct ,lon correct )=(lat p ,lon p );

[0034] If the k laser radar tracks are within the time matching window, according to the average measurement time t of each track, lidarb And longitude and latitude (lat lidarb ,lon lidarb ), correct the k orbits respectively to obtain multiple corrected in-situ data points: (lat correctb ,lon correctb )(b=1,2,…,k).

[0035] The specific process of step S5 is as follows:

[0036] According to the measurement time t of the corrected in-situ data insitu and longitude and latitude (lat correctb ,lon correctb ) in the moment

[0037] Find the spatial matching window r of the corresponding month and the nearest grid point in the matrix R(x,y,m) b (b=1,2,…,k);

[0038] Further, according to the spatial matching window r b , using the haversine algorithm to calculate (lat correctb ,lon correctb ) and the surface spot position of the satellite-borne laser radar (lat lidarb ,lon lidarb ) b (b=1,2,…,k), find the b <r b Space-borne lidar data.

[0039] The specific process of step S6 is as follows:

[0040] According to the quality control mark, the space-borne lidar data with signal saturation, cloud and land conditions are excluded; if more than 50% of the data are excluded, the matching of this data is skipped;

[0041] If the amount of data remaining after screening according to the quality control markers is not less than 50%, calculate the mean value a and standard deviation σ of the remaining data, and retain the data between a-1.5σ and a+1.5σ;

[0042] Calculate the average value a of the remaining satellite-borne lidar data again f and standard deviation σ f , calculate the coefficient of variation:

[0043]

[0044] If CV f >0.15, skip this data matching; if CV f <0.15, the original data measurement value a insitu Compared with the average value of satellite-borne lidar data f As a pair of matching data, this matching is completed;

[0045] If the satellite-borne lidar data of multiple orbits all meet CV f <0.15, then calculate their total data average a f , the in-situ data measurement value a insitu Compared with the average value of satellite-borne lidar data f As a pair of matching data, this matching is completed.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] The present invention provides an improved satellite-borne lidar matching verification method based on spatial variability and flow field. Different spatial matching window sizes are selected based on spatial variability, which solves the problem that most existing methods use a fixed matching window, resulting in inconsistent matching results in different areas, and takes into account the movement of measurement points caused by ocean currents. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 The present invention is a flow chart of a satellite-borne laser radar matching verification method based on spatial variability and flow field.

[0049] Figure 2 This is a statistical distribution diagram of the spatial variability of MODIS data calculated using multiple grid sizes in an example of the present invention. DETAILED DESCRIPTION

[0050] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be pointed out that the embodiments described below are intended to facilitate the understanding of the present invention and do not have any limiting effect on the present invention.

[0051] The present invention takes the matching between the particle backscattering coefficient measured by the satellite-borne laser radar CALIOP and the in-situ float Argo as an example. Figure 1 , specifically including the following steps:

[0052] Step 1: Use multiple grid sizes to calculate the spatial variability of water color remote sensing data for each month.

[0053] In this example, the monthly average MODIS particle backscattering coefficient grid data is used. For each grid point (x, y), the nearby n i ×n i Grid data, n i The values ​​of are n 1 =3, n 2 =5, n 3 =7, n 4 =11, n 5 =21 and n 6 =41. Calculate the spatial variability of each grid point:

[0054]

[0055] In the formula, m is month and std is n i ×n i The standard deviation of all data in the grid, mean is n i ×n i The average value of all data in the grid. Figure 2 As shown in the figure, the spatial variability calculated using the MODIS particle backscattering coefficient data from March to May 2019 is statistically analyzed. The spatial variability increases with the increase of grid size.

[0056] Step 2: traverse each grid point and search for the maximum grid size corresponding to a spatial variability that is less than a set threshold.

[0057] For month m 0 and the grid points (x 0 ,y 0 ), the corresponding spatial variability is CV(x 0 ,y 0 ,m 0 ,n i ). Set the fixed threshold to threshold=0.02 and find the value that satisfies CV(x 0 ,y 0 ,m 0 ,n i ) is less than the maximum subscript i of threshold, denoted as i 0 , the corresponding maximum grid size is According to the above rules, all grid points of each month are traversed to obtain the maximum grid size matrix N(x, y, m).

[0058] Step 3: Calculate the spatial matching window matrix based on the maximum grid size and set the temporal matching window to a fixed value.

[0059] According to the obtained maximum grid size matrix N(x,y,m), calculate the spatial matching window of all grid points in each month:

[0060]

[0061] Where d is the spatial resolution of the water color remote sensing grid data. In this example, the spatial resolution of MODIS data is 4 kilometers.

[0062] Set the time matching window size to a fixed value t, which can be 3 hours, 12 hours, 24 hours, or 48 hours.

[0063] Step 4: Based on the measurement time and longitude and latitude of the in-situ data, use the largest spatial matching window and time matching window to search the satellite-borne lidar data, find the corresponding hourly ocean current speed, and gradually correct the position of the in-situ data.

[0064] The measurement time t for obtaining Argo in-situ data insitu and longitude and latitude (lat insitu ,lon insitu ).

[0065] Set the spatial matching window s and use the haversine algorithm to calculate the longitude and latitude (lat insitu ,lon insitu ) and the surface spot position of the satellite-borne laser radar (lat lidar ,lon lidar ) 0 , find the 0 <s of space-borne lidar data;

[0066] According to the time matching window t and the satellite-borne laser radar measurement time t lidar , find the data satisfying |t in the data satisfying the spatial matching window s insitu -t lidar |<t of space-borne lidar data;

[0067] According to the measurement time t of the in-situ data insitu and longitude and latitude (lat insitu ,lon insitu ), find the corresponding ocean current speed

[0068] If insitu<t lidar , set the timestamp t 0 ,t insitu ,t 1 ,t 2 ,…,t p ,t lidar ,t p+1 , where t 1 ,t 2 ,…,t p Between t insitu and t lidar All consecutive integer hours between 0 is less than t insitu The maximum integer hour moment, t p+1 is greater than t lidar The minimum whole hour of the moment; t u-1 Time to t u Time, longitude and latitude (lat u ,lon u )(u=1,2…,p) at the average ocean current velocity;

[0069] If insitu >t lidar , set the timestamp t 0 ,t lidar ,t 1 ,t 2 ,…,t p ,t insitu ,t p+1 , where t 1 ,t 2 ,…,t p Between t insitu and t lidar All consecutive integer hours between 0 is less than t lidar The maximum integer hour moment, t p+1 is greater than t insitu The minimum whole hour of the moment; t u-1 Time to t u Time, longitude and latitude (lat u ,lon u ) at the average ocean current velocity;

[0070] If insitu =t lidar , directly output the latitude and longitude of the original data;

[0071] According to the ocean current speed and time difference, the displacement distance is calculated, and then according to the direction of the ocean current speed, the longitude and latitude of the original data are updated to (latu ,lon u ), where: (lat 1 ,lon 1 )=(lat insitu ,lon insitu );

[0072] Finally, the output latitude and longitude of the original data is: (lat correct ,lon correct )=(lat p ,lon p );

[0073] If the k laser radar tracks are within the time matching window, according to the average measurement time t of each track, lidarb And longitude and latitude (lat lidarb ,lon lidarb ), correct the k orbits respectively to obtain multiple corrected in-situ data points: (lat correctb ,lon correctb )(b=1,2,…,k).

[0074] Step 5: Find the corresponding spatial matching window according to the corrected measurement time and longitude and latitude of the in-situ data, and match the satellite-borne lidar data with the in-situ data according to the spatial matching window.

[0075] According to the measurement time t of the corrected in-situ data insitu and longitude and latitude (lat correctb ,lon correctb ) in the moment

[0076] Find the spatial matching window r of the corresponding month and the nearest grid point in the matrix R(x,y,m) b (b=1,2,…,k);

[0077] Further, according to the spatial matching window r b , using the haversine algorithm to calculate (lat correctb ,lon correctb ) and the surface spot position of the satellite-borne laser radar (lat lidarb ,lon lidarb ) b (b=1,2,…,k), find the b <r b Space-borne lidar data.

[0078] Step 6: Perform quality control on the matched satellite-borne lidar data to form a satellite-borne lidar-in-situ data matching pair.

[0079] Based on the quality control flags, space-borne lidar CALIOP data with signal saturation, cloud, land, etc. are excluded. If more than 50% of the data are excluded, the matching of this data is skipped.

[0080] If the amount of data remaining after screening according to the quality control markers is not less than 50%, calculate the mean value a and standard deviation σ of the remaining data, and retain the data between a-1.5σ and a+1.5σ.

[0081] Calculate the average value a of the remaining satellite-borne lidar data again f and standard deviation σ f , calculate the coefficient of variation:

[0082]

[0083] If CV f >0.15, skip this data matching; if CV f <0.15, the original data measurement value a insitu Compared with the average value of satellite-borne lidar data f As a pair of matching data, this matching is completed.

[0084] If the satellite-borne lidar data of multiple orbits all meet CV f <0.15, then calculate their total data average a f , the in-situ data measurement value a insitu Compared with the average value of satellite-borne lidar data f As a pair of matching data, this matching is completed.

[0085] The embodiments described above provide a detailed description of the technical solutions and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A space-borne laser radar matching verification method based on spatial variability and flow field, characterized in that: The following steps are involved: S1: Use multiple grid sizes to calculate the spatial variability of water color remote sensing data for each month; S2: Traverse each grid point and search for the maximum grid size corresponding to a spatial variation that is less than the set threshold; S3: Calculate the spatial matching window matrix according to the maximum grid size and set the temporal matching window to a fixed value; S4: According to the measurement time and longitude and latitude of the in-situ data, the largest spatial matching window and time matching window are used to search the satellite-borne lidar data, find the corresponding hourly ocean current speed, and gradually correct the position of the in-situ data; S5: according to the corrected measurement time and longitude and latitude of the in-situ data, find the corresponding spatial matching window, and match the satellite-borne lidar data with the in-situ data according to the spatial matching window; S6: Perform quality control on the matched spaceborne lidar data to form a spaceborne lidar-in-situ data matching pair.

2. The spaceborne laser radar-in-situ data matching method based on spatial variability according to claim 1 is characterized in that: The specific process of step S1 is: Get the monthly average water color remote sensing grid data, and for each grid point (x, y), select n nearby i ×n i Grid data, n i The values ​​of are recorded from small to large as n1, n2, ..., n q ; Calculate the spatial variability of each grid point: In the formula, m is month and std is n i ×n i The standard deviation of all data in the grid, mean is n i ×n i The average value of all data in the grid.

3. The spaceborne laser radar-in-situ data matching method based on spatial variability according to claim 1 is characterized in that: The specific process of step S2 is as follows: For month m0 and grid point (x0, y0), the corresponding spatial variability is CV(x0, y0, m0, n i ), set a fixed threshold, and find the value that satisfies CV(x0,y0,m0,n i ) is less than the maximum subscript i of the threshold, denoted as i0, and the corresponding maximum grid size is n i0 ; According to the above rules, all grid points of each month are traversed to obtain the maximum grid size matrix N(x,y,m).

4. The spaceborne laser radar-in-situ data matching method based on spatial variability according to claim 1 is characterized in that: The specific process of step S3 is as follows: According to the obtained maximum grid size matrix N(x,y,m), calculate the spatial matching window of all grid points in each month: Where d is the spatial resolution of water color remote sensing grid data; Set the time matching window size to t.

5. The spaceborne laser radar-in-situ data matching method based on spatial variability according to claim 1 is characterized in that: The specific process of step S4 is as follows: The measurement time t for obtaining the in-situ data insitu and longitude and latitude (lat insitu ,lon insitu ); Set the spatial matching window s and use the haversine algorithm to calculate the longitude and latitude (lat insitu ,lon insitu ) and the surface spot position of the satellite-borne laser radar (lat lidar ,lon lidar ) and find the space-borne laser radar data satisfying D0<s; According to the time matching window t and the satellite-borne laser radar measurement time t lidar , find the data satisfying |t in the data satisfying the spatial matching window s insitu -t lidar |<t of space-borne lidar data; According to the measurement time t of the in-situ data insitu and longitude and latitude (lat insitu ,lon insitu ), find the corresponding ocean current speed If insitu <t lidar , set timestamp t0,t insitu ,t1,t2,…,t p ,t lidar ,t p+1 , where t1, t2, …, t p Between t insitu and t lidar All consecutive integer hours between t0 and t1, t2 is less than t insitu The maximum integer hour moment, t p+1 is greater than t lidar The minimum whole hour of the moment; t u-1 Time to t u Time, longitude and latitude (lat u ,lon u ) is the average ocean current velocity at , u=1,2…,p; If insitu >t lidar , set timestamp t0,t lidar ,t1,t2,…,t p ,t insitu ,t p+1 , where t1, t2, …, t p Between t insitu and t lidar All consecutive integer hours between t0 and t1, t2 is less than t lidar The maximum integer hour moment, t p+1 is greater than t insitu The minimum whole hour of the moment; t u-1 Time to t u Time, longitude and latitude (lat u ,lon u ) at the average ocean current velocity; If insitu =t lidar , directly output the latitude and longitude of the original data; According to the ocean current speed and time difference, the displacement distance is calculated, and then according to the direction of the ocean current speed, the longitude and latitude of the original data are updated to (lat u ,lon u ), where: (lat1,lon1)=(lat insitu ,lon insitu ); Finally, the output latitude and longitude of the original data is: (lat correct ,lon correct )=(lat p ,lon p ); If the k laser radar tracks are within the time matching window, according to the average measurement time t of each track, lidarb And longitude and latitude (lat lidarb ,lon lidarb ), correct the k orbits respectively and obtain multiple corrected in-situ data points: (lat correctb ,lon correctb )b=1,2,…,k.

6. The space-borne laser radar-in-situ data matching method based on spatial variability according to claim 1 is characterized in that: The specific process of step S5 is as follows: According to the measurement time t of the corrected in-situ data insitu and longitude and latitude (lat correctb ,lon correctb ) Find the spatial matching window r of the corresponding month and the nearest grid point in the matrix R(x,y,m) b , b=1,2,…,k; Further, according to the spatial matching window r b , calculated using the haversine algorithm (lat correctb ,lon correctb ) and the surface spot position of the satellite-borne laser radar (lat lidarb ,lon lidarb ) b , find the b <r b Space-borne lidar data.

7. The space-borne laser radar-in-situ data matching method based on spatial variability according to claim 1 is characterized in that: The specific process of step S6 is as follows: According to the quality control mark, the space-borne lidar data with signal saturation, cloud and land conditions are excluded; if more than 50% of the data are excluded, the matching of this data is skipped; If the amount of data remaining after screening according to the quality control markers is not less than 50%, calculate the mean value a and standard deviation σ of the remaining data, and retain the data between a-1.5σ and a+1.5σ; Calculate the average value a of the remaining satellite-borne lidar data again f and standard deviation σ f , calculate the coefficient of variation: If CV f >0.15, skip this data matching; if CV f <0.15, the original data measurement value a insitu Compared with the average value of satellite-borne lidar data f As a pair of matching data, this matching is completed; If the satellite-borne lidar data of multiple orbits all meet CV f <0.15, then calculate their total data average a f , the in-situ data measurement value a insitu Compared with the average value of satellite-borne lidar data f As a pair of matching data, this matching is completed.

Images