A quantitative method of migrating birds based on dual-polarization weather radar and bird detection radar

Abstract

The application discloses a kind of quantitative methods of migratory birds based on dual-polarized weather radar and bird detection radar.The application is jointly observed by dual-polarized weather radar and bird detection radar, carefully selects the original observation data of the two radars, eliminates the false alarm and clutter track of the bird detection radar, eliminates the meteorological echo and insect echo of the polarization weather radar, then constructs effective quantitative relationship using the data in the airspace monitored jointly, the present method carefully considers the observation airspace of different radars, has higher quantitative accuracy, and is more suitable for China.

Description

Technical Field

[0001] This invention relates to the field of weather radar technology, specifically to a quantitative method for migratory birds based on dual-polarization weather radar and bird-detecting radar. Background Technology

[0002] Large-scale migratory bird migration is an important material and energy cycle process in nature. Monitoring the dynamics of migratory bird migration is of great value to natural ecological research and human society: it provides important data on migratory bird migration; it studies the impact of climate change on bird migration; it studies the link between migratory birds and influenza epidemics; and it provides early warning of bird strike risks, etc.

[0003] Radar, as an important means of air surveillance, has been applied to the monitoring of biological migration. Bird-detecting and insect-detecting radars can acquire important parameters such as the trajectory and wingbeat frequency of individual flying animals; however, their monitoring range is limited and their deployment numbers are insufficient, making it difficult to cope with large-scale migration monitoring. Weather radar not only has a wide observation range, but China has also established a comprehensive new-generation Doppler weather radar network, which can provide an important data source for large-scale migratory bird migration monitoring.

[0004] To achieve accurate quantitative analysis of migratory birds based on weather radar, it is necessary to establish a mapping relationship between bird density and weather radar echo intensity. Dokter constructed a relationship between bird density and echo intensity using observation data from Europe. However, the differences in bird size across different regions and the inconsistencies in the airspace observed by the two radars were not carefully considered. Furthermore, Dokter's observation data came from C-band weather radar, while most of my country's data comes from S-band weather radar. Therefore, the existing quantitative relationship is not applicable to my country. Summary of the Invention

[0005] In view of this, the present invention provides a quantitative method for migratory birds based on dual-polarization weather radar and bird detection radar, which can provide an effective quantitative model for weather radar monitoring of migratory birds in my country, and help to study the large-scale migration process of migratory birds, the impact of climate change on migratory birds, and early warning of large-scale migration.

[0006] The present invention provides a quantitative method for migratory birds based on dual-polarization weather radar and bird-detecting radar, comprising:

[0007] S1, equipped with bird-detecting radar and weather radar; the effective detection range of the weather radar covers the detection range of the bird-detecting radar; the bird-detecting radar adopts an alternating near-far waveform scanning mode;

[0008] S2, remove false targets from the bird detection radar tracks, and then calculate the bird density;

[0009] Non-bird echoes from weather radar observations are removed, including meteorological echoes and insect echoes;

[0010] S3. Correlation analysis was performed on the time series of bird density observed by bird detection radar and the time series of echo intensity at different locations of weather radar to determine the airspace monitored by both.

[0011] S4. Extract the average bird density of the bird-detecting radar in the common monitoring airspace and the average reflectivity intensity of the weather radar in the common monitoring airspace. Through regression analysis, obtain the quantitative relationship between bird density and weather radar reflectivity intensity.

[0012] Preferably, the process of eliminating false track targets in S2 includes: setting a signal-to-noise ratio threshold and filtering out targets with a signal-to-noise ratio lower than the threshold observed by the bird-detecting radar.

[0013] Preferably, the signal-to-noise ratio threshold is 18dB.

[0014] Preferably, the process of removing false track targets in S2 includes: calculating the curvature of the track and removing targets from the track targets observed by the bird detection radar whose curvature is greater than a set curvature threshold.

[0015] A preferred method for target elimination based on flight path curvature is as follows:

[0016] Calculate the tangent angle between two consecutive points in the trajectory; calculate the mean and standard deviation of the tangent angle between all two consecutive points in the trajectory; remove targets with a standard deviation greater than the set deflection threshold.

[0017] The preferred tangent angle between two consecutive points in track n is:

[0018]

[0019] Where, θ i,n In track n, starting from detection point s i To the detection point s i+1 Direction; (x) i ,y i ), (x i+1 ,y i+1 ) represent the detection points s i and detection point s i+1 Horizontal position;

[0020] Record the directions between all consecutive points on track n as Θ. n :

[0021]

[0022] Among them, L n Let n be the total number of detection points for track n; calculate the mean value in the direction of track n. as follows:

[0023]

[0024]

[0025] Calculate the standard deviation σ of the direction sequence of track n. n :

[0026]

[0027] Preferably, the bending degree threshold is set based on the statistical results of the flight track observation data.

[0028] A preferred method for calculating bird density is as follows:

[0029] Based on the maximum and minimum elevation angle information of the bird detection radar and the pre-defined height axis, the radar sampling volume corresponding to different height layers is calculated;

[0030] During omnidirectional scanning, at a certain height layer h i ~h i+1 The sampling volume is expressed as the difference between the volumes of two frustums; where the larger frustum has a radius r1 = h. i / tanθ min r2 = h i+1 / tanθ min The smaller frustum has the following radius: r³ = h i / tanθ max r4 = h i+1 / tanθ max ; where θ min and θ max These are the minimum and maximum elevation angles for observing all valid track points, respectively;

[0031] The circumferential sampling volume of this height layer is:

[0032]

[0033] Where Azi is the range of the actual scanning azimuth angle;

[0034] Based on the altitude information of points in the effective flight path, the number of target points at different altitude levels is counted. The corresponding density is obtained from the statistical results and the sampling volume. Then, the density is divided by the number of scan cycles of the bird detection radar within the statistical time period to obtain the bird density.

[0035] Preferably, in S2, meteorological echoes are eliminated based on the differential reflectivity and correlation coefficient of the target observed by the weather radar and based on the depolarization rate.

[0036] Preferably, in S2, insect echoes are eliminated by setting a threshold based on the error between the estimated radial velocity of the target observed by the weather radar and the observed value.

[0037] Beneficial effects:

[0038] Compared to existing quantitative methods, this method, through joint observations by dual-polarization weather radar and bird-hunting radar, carefully filters the raw observation data from both radars and constructs an effective quantitative relationship using data from the shared monitoring airspace. This method carefully considers the observation airspace of different radars, has higher quantitative accuracy, and is more suitable for my country.

[0039] This invention utilizes the signal-to-noise ratio to eliminate false alarms caused by the near-far wave transformation of bird-finding radar, and utilizes the curvature of the flight path to eliminate clutter tracks. It is simple, feasible, and highly reliable.

[0040] The method utilizes depolarization rate to eliminate meteorological echoes from polarimetric weather radar observations and uses the error between radial velocity estimates and observed values ​​to eliminate insect echoes, achieving high accuracy. Attached Figure Description

[0041] Figure 1 The spatial relationship between weather radar and bird detection radar.

[0042] Figure 2 The results of bird-detecting radar observations were used to filter flight paths.

[0043] Figure 3 The graphs are from observations by two radars.

[0044] Figure 4 This demonstrates the quantitative relationship between bird density and weather radar reflectivity.

[0045] Figure 5 This is a flowchart of the present invention. Detailed Implementation

[0046] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0047] This invention provides a quantitative method for migratory birds based on dual-polarization weather radar and bird-detecting radar.

[0048] Weather radar networks can perform spatiotemporal sampling of migratory bird migration processes, effectively recording their dynamic processes. However, weather radar observations do not directly reflect bird populations; therefore, an effective mapping function is needed to transform the observations into meaningful bird density, i.e.

[0049] ρ bird =F(Z) (1)

[0050] Where Z is the reflectivity factor of weather radar observation, ρbird F is the spatial density of birds, and F(·) is the quantitative relationship function.

[0051] The flowchart of the method of this invention is as follows Figure 5 As shown, the specific steps include the following:

[0052] Step 1: Within the effective coverage area of ​​the dual-polarization weather radar, install a bird-detecting radar with a scanning system and conduct long-term operational monitoring to obtain the actual bird density in the area.

[0053] like Figure 1 As shown, the bird-detecting radar is located at approximately 20° south-southeast of the weather radar, at a straight-line distance of approximately 24 km.

[0054] Dual-polarization weather radar operates in the S-band, simultaneously emitting horizontally and vertically polarized electromagnetic waves to detect the atmosphere. With a 3dB beamwidth of approximately 1° and a detection radius of 460km, it effectively covers the location of bird-detecting radars. Compared to weather radars, bird-detecting radars provide more detailed individual target observation information. It uses electronic scanning for elevation, enabling rapid coverage of multiple elevation angles, and mechanical scanning for azimuth, with a rotation speed of 60deg / s, completing a scan in 6 seconds. The range resolution is 15m, allowing for the measurement of three-dimensional tracks of individual targets. Balancing detection blind range and effectiveness, bird-detecting radars employ alternating near- and far-range waveforms to improve range coverage. Near-range waveforms cover from 150m to 2400m, while the starting distance for far-range waveforms is 1500m, resulting in a radar cross-section of 0.03m². 2 The drone has an effective detection range of 5km. Through operational observation, the two radars have accumulated valuable historical observation data.

[0055] Step two involves processing the original flight paths observed by the bird-detecting radar, including removing false flight paths, and then calculating the bird density.

[0056] False targets mainly fall into two categories: those caused by false alarms and those caused by clutter. For the first type, they often appear in the overlapping area of ​​near- and far-field waveforms, exhibiting a ring-shaped feature centered on the radar origin and possessing a low signal-to-noise ratio (SNR). Therefore, by setting an effective SNR threshold, targets with an SNR lower than the threshold observed by the bird-detecting radar can be filtered out. The SNR of a UAV at 5000m is approximately 17dB, and the threshold for this range can be estimated to be approximately 18dB using radar equations.

[0057] For the second scenario, the flight path exhibits spatial stability, and its geometry is often a polygonal shape clustered in a two-dimensional plane, significantly different from typical bird flight paths. Therefore, the degree of curvature of the flight path can be used to filter out this type of clutter. The three-dimensional spatial position of a detection point within a certain flight path is recorded as s. i =[x i ,y i ,z i ], where i represents the sequence number of the detection point, and the track is denoted as Where L n This represents the number of detection points in the track, where n represents the track number. The degree of curvature of the track is represented by the difference in directional change between two consecutive points on the two-dimensional projection plane, defined as...

[0058]

[0059] Where θ i This represents the direction from detection point i to detection point i+1, and the direction between all consecutive points on this track is recorded as Θ. n The following indicates

[0060]

[0061] Based on the definition of the mean of direction data, the mean of the track direction is calculated as follows:

[0062]

[0063] in This is the mean of the final calculated direction data. Geometrically, it can be understood as the vector synthesis of all direction vectors, with the final vector direction serving as the average direction. The standard deviation of the direction sequence for each track is calculated as follows:

[0064]

[0065] Where σ n σ is the standard deviation of the direction sequence of the nth track, and the filtering threshold is set to 50° based on the statistical results of the track observation data. n Quantities exceeding a certain threshold are considered noise. For example... Figure 2 The image shows the results of filtering out the tracks in the two different scenarios.

[0066] In density calculation, the key is to calculate the sampling volume. Based on the maximum and minimum elevation angle information and the pre-defined height axis, the radar sampling volume corresponding to different height layers is calculated. During omnidirectional scanning, the sampling volume at a certain height layer h... i ~h i+1 The sampling volume can be decomposed into the subtraction of the volumes of two frustums. For the larger frustum, the radius r1 = h. i / tanθ min r2 = h i+1 / tanθ min For a smaller frustum, the radius r3 = h i / tanθ max r4 = h i+1 / tanθ max , where θ min and θ max These are the minimum and maximum elevation angles, respectively, for observing all valid track points. Based on the formula for calculating the volume of a frustum, the circumferential sampling volume for this altitude level can be obtained as follows:

[0067]

[0068] Where Azi is the range of the actual scanning azimuth angle.

[0069] Based on the altitude information of points in the valid flight paths, the number of target points at different altitude levels is statistically analyzed. The corresponding density can be obtained from the statistical results and the sampling volume. However, considering that weather radar is the result of a single scan, while bird detection radar is the accumulated result of multiple scans, the spatial density obtained from bird detection radar is averaged over time, i.e., divided by the number of scan cycles of bird detection radar within the statistical time period.

[0070] Step 3: Process the echoes from polarimetric weather radar observations to remove non-bird echoes such as meteorological and insect echoes.

[0071] For meteorological echoes, depolarization rate can be used for discrimination, defined as follows:

[0072]

[0073] Z DR It is differential reflectivity, ρ HV This is the correlation coefficient. Echoes below the threshold are classified as meteorological, while those above the threshold are classified as biological echoes. On the biological and meteorological echo dataset built based on this weather radar observation data, the classification accuracy is above 90%.

[0074] For insect echoes, based on the characteristic of insects' weak autonomous flight ability, discrimination can be made using the error between the estimated and observed radial velocity values, defined as...

[0075]

[0076] Where n is the number of azimuth pulses, v i Weather radar operates at an azimuth angle θ iThe observed radial velocity, θ is the estimated migration direction, and v is the estimated horizontal migration velocity. A threshold of 2 m / s is set; if the velocity is less than the threshold, it indicates that the autonomous flight capability of organisms at that altitude is weak, with small insects dominating; conversely, if the velocity is greater than the threshold, birds are dominant.

[0077] Step four: Extract data from the airspace jointly monitored by the two radars and construct a quantitative relationship through regression analysis.

[0078] Correlation analysis was performed between the time series of bird density observed by bird-detecting radar and the time series of echo intensity from weather radar at different locations to determine the accurate common monitoring airspace for both, and the corresponding spatial observation data were extracted: the average density ρ of the bird-detecting radar at the corresponding altitude layer. bird The weather radar value is the average reflectivity intensity Z within the corresponding area.

[0079] Figure 3 These are the observation curves from two radars. Figure 4 The results are linear regressions of the corresponding observation points, from which a quantitative relationship can be obtained between bird density and weather radar reflectivity intensity.

[0080] ρ bird [dB] = Z[dB] - 19.609[dB] (9)

[0081] In summary, the above are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A quantitative method for migrating birds based on dual-polarimetric weather radar and bird-detecting radar, characterized by, include: S1, equipped with bird-detecting radar and weather radar; the effective detection range of the weather radar covers the detection range of the bird-detecting radar; the bird-detecting radar adopts an alternating near-far waveform scanning mode; S2, remove false targets from the bird detection radar tracks and then calculate the bird density; among which, removing false targets includes: removing targets based on track curvature; Non-bird echoes from weather radar observations are removed, including meteorological echoes and insect echoes; Specifically, target elimination based on track curvature involves: calculating the tangent angle between two consecutive points on the track observed by bird detection radar; calculating the mean and standard deviation of the tangent angles between all consecutive points on the track; and eliminating targets with a standard deviation greater than a set curvature threshold. S3. Correlation analysis was performed on the time series of bird density observed by bird detection radar and the time series of echo intensity at different locations of weather radar to determine the airspace monitored by both. S4. Extract the average bird density of the bird-detecting radar in the common monitoring airspace and the average reflectivity intensity of the weather radar in the common monitoring airspace. Through regression analysis, obtain the quantitative relationship between bird density and weather radar reflectivity intensity.

2. The quantitative method of migrating birds based on dual-polarimetric weather radar and bird-detecting radar as claimed in claim 1, wherein The process of eliminating false track targets in S2 includes: setting a signal-to-noise ratio threshold and filtering out targets with a signal-to-noise ratio lower than the threshold observed by the bird-detecting radar.

3. The method of claim 2, wherein the migration bird quantification method is based on a dual-polarimetric weather radar and a bird detection radar. The signal-to-noise ratio threshold is 18dB.

4. The method of claim 1, wherein the migration bird quantification method is based on a dual-polarimetric weather radar and a bird detection radar. track n The tangent angle between the two points is: wherein, represents a track n from a detection point s i to a detection point s i+1 ;( x i ,y i ),( x i+1 ,y i+1 ) are horizontal positions of the detection point s i and the detection point s i+1 , respectively. Record the direction between all consecutive two points as n Record the direction between all consecutive two points as : in, L n For the track n Total number of detection points; calculate flight path n Mean of direction as follows: Computing a track n of standard deviations of the direction sequence : 。 5. The method of claim 1 or 4, wherein the migration bird quantification method is based on a dual-polarimetric weather radar and a bird detection radar. The bending degree threshold is set based on the statistical results of the flight track observation data.

6. The method of claim 1, wherein the migration bird quantification method is based on dual-polarimetric weather radar and bird detection radar. The method for calculating bird density is as follows: Based on the maximum and minimum elevation angle information of the bird detection radar and the pre-defined height axis, the radar sampling volume corresponding to different height layers is calculated; For omnidirectional scanning, the sampling volume of a certain height layer is expressed as the subtraction of two circular cone volumes; where the larger circular cone has a radius , and the smaller one has a radius , ; where θ min and θ max are the minimum and maximum elevation angles of all the valid track points observed, respectively. The circumferential sampling volume of this height layer is: wherein is the range of actual scan azimuth angles; Based on the altitude information of points in the effective flight path, the number of target points at different altitude levels is counted. The corresponding density is obtained from the statistical results and the sampling volume. Then, the density is divided by the number of scan cycles of the bird detection radar within the statistical time period to obtain the bird density.

7. The method of claim 1, wherein the migration bird quantification method is based on dual-polarimetric weather radar and bird detection radar. In S2, meteorological echoes are eliminated based on the differential reflectivity and correlation coefficient of the target observed by the weather radar and based on the depolarization rate.

8. The method of claim 1, wherein the migration bird quantification method is based on dual-polarimetric weather radar and bird detection radar. In step S2, based on the error between the estimated radial velocity of the target observed by the weather radar and the observed value, insect echoes are eliminated by setting a threshold.

Images