A multi-level tsunami early warning method based on combined two-dimensional L-array HFSWR and GNSS anomaly information detection

Abstract

The present invention relates to a combined two-dimensional L A multi-level tsunami early warning method for detecting HFSWR and GNSS anomaly information, the method comprising: L The HFSWR and GNSS observation data were combined to achieve spatial matching of the two information, and the time series of HFSWR flow field, ionospheric disturbance information and GNSS ionospheric disturbance information were extracted respectively. The Mann-Kendall test method and slope detection method were used to detect anomalies in the ionospheric information obtained based on GNSS and HFSWR respectively to determine whether it entered the "primary warning" state. If it entered the "primary warning" state, further use was made to Q A factor test method detects anomalies in HFSWR flow field information to determine whether to enter an "emergency warning" state. This invention integrates HFSWR and GNSS technologies to detect anomalies in ocean flow fields and ionospheric disturbances, effectively identifying and grading tsunami-induced anomalies, and providing technical support for improving tsunami warning decision-making support capabilities.

Description

Technical Field

[0001] The present invention relates to the technical field of tsunami monitoring and early warning, and in particular to a multi-level tsunami early warning method combining two-dimensional L-array high frequency surface wave radar (HFSWR) and global navigation satellite system (GNSS) anomaly information detection. Background Art

[0002] Tsunamis are natural disasters that severely impact marine and coastal safety. Real-time monitoring and effective early warning are crucial for reducing disaster losses. Currently, real-time tsunami observation relies primarily on inverting tsunami wave heights from seafloor pressure gauges or observing tsunami waves in coastal areas. These point-based methods have limited spatial coverage, making them difficult to implement for large-scale, long-distance tsunami monitoring and early warning.

[0003] Currently, the main technologies applicable to large-scale tsunami remote sensing monitoring include HFSWR and GNSS. GNSS, by calculating the total electron content (TEC), can detect the dynamic characteristics of the ionosphere during a tsunami, offering advantages in ionospheric information monitoring. HFSWR, on the other hand, can identify and assess the magnitude of tsunamis by observing changes in ocean currents caused by tsunami waves. By strategically deploying a two-dimensional L-array receiving antenna, HFSWR can simultaneously acquire both azimuth and elevation angle information of the target, thereby improving the spatial resolution of tsunami-induced ionospheric disturbances.

[0004] Although GNSS technology has been widely used to monitor ionospheric anomalies and boasts global coverage, it struggles to directly reflect changes in ocean dynamic parameters. Two-dimensional L-array HFSWRs can monitor large-scale ocean surface currents beyond the horizon and acquire certain ionospheric data, but they still have limitations in identifying and analyzing the propagation speed and direction of traveling ionospheric disturbances (TIDs). Therefore, current tsunami warning methods based on a single monitoring method struggle to simultaneously account for both surface and ionospheric disturbance information, resulting in insufficient warning reliability. Further integration of multiple observation technologies is urgently needed to enhance the spatial awareness and timeliness of tsunami warnings. Summary of the Invention

[0005] (1) Technical issues to be resolved

[0006] The present invention aims to provide a multi-level tsunami early warning method that combines two-dimensional L-array HFSWR and GNSS to achieve multi-dimensional detection and hierarchical response of tsunami anomaly information, thereby improving the spatial coverage and accuracy of the warning and reducing the risk of false alarms.

[0007] (2) Technical solution

[0008] The present invention comprises the following steps:

[0009] Step 1: Spatial matching.

[0010] The ionospheric distance information obtained by the two-dimensional L-array HFSWR during oblique detection is converted into horizontal distance information to achieve precise matching with the GNSS penetration point position information, including: establishing a spatial positioning model based on the two-dimensional L-array HFSWR detection range. The model uses the two-dimensional L-array HFSWR operating frequency as a parameter to define the detection range of the ocean flow field and the ionosphere; and calculating and obtaining the position coordinates of the ionospheric puncture point and the corresponding sub-satellite point based on GNSS observation data.

[0011] Step 2: Data preprocessing.

[0012] Based on the RD spectrum obtained by the two-dimensional L-array HRSWR, the disturbance time series is extracted, including: extracting the ocean echo according to the RD spectrum and inverting it to obtain flow velocity information; after extracting and removing the ocean echo according to the RD spectrum, extracting the ionospheric echo and obtaining its time-frequency distribution result.

[0013] Based on the total electron content data obtained by GNSS, the ionospheric disturbance time series is extracted, including: based on the GNSS dual-frequency observation data, using the carrier phase smoothing pseudorange method to obtain the slant path total electron content data within the monitoring range; converting the slant path total electron content into the vertical total electron content through the ionospheric thin layer model; using the second-order numerical difference method to process the vertical total electron content data to obtain the ionospheric disturbance time series.

[0014] Step 3: Detection of ionospheric information anomalies.

[0015] The Mann-Kendall test method is used to detect abnormal information in the ionospheric disturbance time series extracted by GNSS. If an abnormal value is detected, the ionospheric information obtained by the two-dimensional L-array HFSWR is further determined to be abnormal; otherwise, the monitoring area is considered to be unaffected by the tsunami.

[0016] The slope detection method is used to detect abnormal information in the ionospheric information obtained by the two-dimensional L-array HFSWR. If an abnormal value is detected, the "primary warning" stage is entered.

[0017] Step 4: Detect flow field information anomaly.

[0018] The velocity data obtained from the two-dimensional L-array HFSWR is tested for abnormal information using the Q-factor test method. If an abnormal value is detected, the "emergency warning" phase is initiated; otherwise, the monitored area is considered to be unaffected by the tsunami and everything is normal.

[0019] (3) Beneficial effects

[0020] The advantages of the present invention are:

[0021] The present invention innovatively combines the two-dimensional L-array HFSWR and GNSS technologies to comprehensively detect ocean flow field and ionospheric disturbance anomalies, achieving effective identification and graded response to abnormal information caused by tsunamis, and providing technical support for improving the coverage, accuracy and decision-making support capabilities of tsunami warnings. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a basic flow chart of the present invention.

[0023] Figure 2 This is the two-dimensional L-array signal receiving model of the present invention.

[0024] Figure 3 This is the GNSS ionosphere detection result of the present invention.

[0025] Figure 4 This is the ionospheric detection result of the two-dimensional L-array HFSWR of the present invention.

[0026] Figure 5 This is the flow field detection result of the two-dimensional L-array HFSWR of the present invention. DETAILED DESCRIPTION

[0027] To make the purpose, content and advantages of the present invention clearer, the following Figure 1 , the specific embodiments of the present invention are further described in detail.

[0028] Step 1: Spatial matching.

[0029] Based on the azimuth and elevation resolution capabilities of the two-dimensional L-array HFSWR, a spatial positioning model based on the detection range of the two-dimensional L-array HFSWR is established. The model uses the operating frequency of the two-dimensional L-array HFSWR as a parameter to define the detection range of the ocean flow field and the ionosphere.

[0030] Specifically, the signal receiving model of the two-dimensional L-array HFSWR is as follows: Figure 2As shown, the blue circles represent array elements, and the origin is the reference array element. The farthest detection distance of HFSWR to the sea surface is expressed as R1, and the farthest detection distance to the ionosphere is expressed as R2. Q is the farthest point of ionospheric detection, and the corresponding wave arrival direction is (φ, θ). The projection point on the XOY plane is Q'. φ represents the azimuth angle, ranging from (-π, π), that is, the angle between the line connecting the OQ' point and the positive half axis of the X axis; θ represents the pitch angle, ranging from (0, π / 2), that is, the angle between the OQ point and the positive half axis of the Z axis. The coordinate value of point Q in the XY coordinate system can be expressed as:

[0031]

[0032]

[0033] ,

[0034] When TIDs at z altitude propagate to within the range of the corresponding (x, y) points, the HFSWR of the L array can achieve effective detection of the ionosphere.

[0035] Based on the GNSS observation data, the position coordinates of the ionospheric puncture point and the corresponding sub-satellite point are calculated and obtained.

[0036] Specifically, the intersection of the GNSS receiver's observation line of sight with the single layer of the ionosphere is called the puncture point, and its projection point on the Earth's surface is called the sub-satellite point. p 、φ p They are:

[0037]

[0038] ,

[0039] Where λ and φ are the longitude and latitude of the receiver, α is the geocentric angle between the receiver and the satellite penetration point, and A is the azimuth of the satellite relative to the receiver.

[0040] In this way, the ionospheric distance information obtained by the two-dimensional L-array HFSWR during the oblique detection process is converted into horizontal distance, which can achieve accurate matching with the GNSS puncture point position information.

[0041] Step 2: Data preprocessing.

[0042] Based on the RD spectrum obtained by the two-dimensional L-array HRSWR, the disturbance time series is extracted.

[0043] Specifically, the 2D-SNR method is used to identify and extract first-order ocean echoes. That is, the signal-to-noise ratio is calculated by selecting the echo intensity of the first-order spectrum area of ​​the two-dimensional window of the upper and lower adjacent distance cells of the current distance cell and the noise area of ​​the two surrounding two-dimensional windows. The continuous distribution of the entire first-order spectrum in distance is used to highlight the difference between the first-order spectrum and the surrounding noise, so as to detect the first-order spectrum boundary. For the extracted ocean echoes, the flow velocity information is inverted based on the Bragg scattering principle. The specific formula is as follows:

[0044] ,

[0045] Where σ1 represents the echo power spectrum of the first-order scattering, ω represents the Doppler frequency, k0 represents the wave number, S represents the directional wave spectrum, m = ±1 represents the positive and negative directions of the first-order spectrum, g represents the gravitational acceleration, V cr Indicates flow rate;

[0046] After extracting and removing the ocean echo, the OSTU image segmentation method is used to extract the ionospheric echo. For the extracted ionospheric echo, its time-frequency distribution result is obtained to capture TIDs in the ionosphere.

[0047] The ionospheric disturbance time series is extracted based on the TEC data acquired by GNSS.

[0048] Specifically, based on the GNSS observation data, the carrier phase smoothed pseudorange method is used to obtain the total electron content data of the oblique path within the monitoring range, that is,

[0049]

[0050] ,

[0051] Where STEC is the slant path total electron content, f1 and f2 are the two carrier frequencies of GNSS; φ1 and φ2 are the phases of the two carriers; λ1 and λ2 are the corresponding carrier wavelengths; n is the number of observation epochs; is the carrier ambiguity term; B s is the hardware delay of the satellite; B r is the hardware delay of the receiver; P1 and P2 are the pseudoranges of the two bands; k is the observation epoch number.

[0052] The STEC is converted into vertical total electron content by the ionospheric thin layer model, i.e.

[0053] ,

[0054] where VTEC is the vertical total electron content, R is the radius of the Earth, H is the height of the ionospheric thin layer, and z is the zenith distance of the satellite penetration point.

[0055] In order to extract the disturbance information in the VTEC time series, it is necessary to eliminate the background trend and bias terms. The second-order numerical difference method is used to process the VTEC data, and its expression is:

[0056]

[0057] ,

[0058] Where τ is the difference step size, which is set to 300s to effectively obtain gravity wave information. v(k) is the VTEC at epoch k, Δv(k), Δ 2 v(k) is the VTEC disturbance time series extracted by first-order and second-order numerical differences, respectively.

[0059] After the VTEC time series of all stations are extracted, a two-dimensional ionospheric disturbance time-distance map can be generated based on the distance between the puncture point and the tsunami epicenter.

[0060] In order to detect whether a tsunami has entered the monitoring range, it is necessary to extract the changes in VTEC at a certain measuring station separately, use a bandpass filter to eliminate high-frequency noise, and highlight the wave characteristics such as gravity waves.

[0061] Step 3: Detection of ionospheric information anomalies.

[0062] The Mann-Kendall (MK) test method is used to detect abnormal information on the VTEC time series after bandpass filter processing. The critical line of the test is determined to be ±1.96 based on the normal distribution quantile corresponding to the significance level α=0.05. The specific process is as follows:

[0063]

[0064]

[0065]

[0066] ,

[0067] In this way, UF is calculated k and its reverse UB k Two curves: If the two curves intersect and the intersection is within ±1.96 of the critical line, the time corresponding to the intersection is the moment when the mutation begins. At this time, continue to judge whether the ionospheric information obtained based on the two-dimensional L-array HFSWR is abnormal; if there is no intersection or the intersection exceeds the critical line range or there are multiple intersections, it is considered that the monitoring area is not affected by the tsunami.

[0068] The slope detection method is used to detect abnormal information of the ionospheric information obtained by the two-dimensional L-array HFSWR. The specific process is as follows:

[0069] The Doppler value corresponding to the maximum echo amplitude at each moment is selected as the sampling point, and the cubic spline interpolation fitting curve is used;

[0070] The slope of the fitted curve is then monitored to determine the state of the ionosphere. If the slope continues to change dramatically, it indicates normal ionospheric activity. If the slope remains stable, a TID is considered to be present. Because TIDs typically exhibit a certain periodicity, only disturbances lasting more than 10 minutes are considered to be the result of a tsunami.

[0071] When the ionospheric disturbance lasts for more than 10 minutes, it is considered that the ionospheric anomaly caused by the tsunami has spread to the monitoring range of the two-dimensional L-array HFSWR, and the "primary warning" stage is entered. The primary warning results can provide a basis for relevant departments to take further emergency measures; if no abnormal disturbance meeting the above conditions is detected, it is considered that the monitoring area is not affected by the tsunami.

[0072] Step 4: Detect flow field information anomaly.

[0073] The Q-factor test method is used to detect abnormal information in the velocity data observed by the two-dimensional L-array HFSWR. The specific process is as follows:

[0074] Within the detection range of the 2D L-array HFSWR flow field, a zone is divided every two kilometers. For zone b at time t, calculate the velocity v b (t) Average speed a in the past 5 minutes b (t) and standard deviation s b (t);

[0075] For each zone, the velocity deviation index d relative to the background is calculated b (t), i.e.

[0076] ,

[0077] Then the speed deviation function D(t) is

[0078] ;

[0079] For each zone, calculate the velocity change index Δv in two adjacent time intervals from t−2δ to t b (t), where δ is the time resolution, set the value of δ to 5 minutes:

[0080]

[0081] Then the velocity increment function ΔV(t)

[0082] ;

[0083] Calculate the correlation function C(t), which describes the correlation between the velocities in three adjacent bands at time t, t−δ, and t−2δ. C(t) defaults to 1 unless the velocities increase or decrease with time from t−2δ to t−δ and from t−δ to t in all three bands, in which case C(t) is set to 100.

[0084] Thus, the calculation formula of the tsunami detection factor at time t is:

[0085] ;

[0086] If q(t) ≥ threshold Q0, Q0 ≥ 200, the system enters the “emergency warning” stage; if no abnormal value is detected, the monitoring area is considered to be unaffected by the tsunami and everything is normal.

[0087] Example

[0088] The effect of the present invention is further illustrated by the following data:

[0089] The GNSS data are derived from GEONET observation data during the 2011 Tohoku earthquake and tsunami downloaded from the Geospatial Information Authority of Japan. Figure 3 The MK test method of the present invention is used to detect the ionospheric information. Figure 3 As can be seen from the figure, the present invention can determine whether tsunami-induced anomalies appear in ionospheric information acquired via GNSS. The main system parameters of the two-dimensional L-array HFSWR are: radar operating frequency of 4.7 MHz, L-shaped array, total number of array elements of 7, uniform spacing of 11.4 m between array elements, and bandwidth of 40 kHz. Using the slope detection method described in the present invention, a pair of RD spectra from this measurement system were tested. Figure 4 is the detection result of ionospheric information. Figure 4 It can be seen from the figure that the present invention can determine whether TIDs exist in the ionosphere. Figure 5 The results of anomaly detection on HFSWR flow field data using the Q factor test method of the present invention are obtained by inverting the echo spectrum of the 2011 Japan tsunami HFSWR simulated based on the COMCOT numerical model. Figure 5 It can be seen from the figure that the present invention can effectively identify whether the flow velocity has undergone abnormal changes caused by a tsunami.

[0090] The above embodiments of the present invention are merely examples for the purpose of clearly illustrating the present invention and are not intended to limit the embodiments of the present invention. Those skilled in the art will appreciate that other variations or modifications may be made based on the above description. It is not necessary and impossible to enumerate all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the claims of the present invention.

Claims

1. A multi-level tsunami early warning method combining two-dimensional L-array HFSWR and GNSS anomaly information detection, characterized in that: include: (1) Spatial matching: converting the ionospheric distance information obtained by the two-dimensional L-array HFSWR during oblique detection into horizontal distance information to achieve accurate matching with the GNSS penetration point position information, including: establishing a spatial positioning model based on the detection range of the two-dimensional L-array HFSWR. The model uses the two-dimensional L-array HFSWR operating frequency as a parameter to define the detection range of the ocean flow field and the ionosphere; and calculating and obtaining the position coordinates of the ionospheric penetration point and the corresponding sub-satellite point based on the GNSS observation data. (2) Data preprocessing: extracting disturbance time series based on the RD spectrum obtained by the two-dimensional L-array HFSWR, including: extracting ocean echoes according to the RD spectrum and inverting to obtain velocity information; extracting ionospheric echoes according to the RD spectrum after extracting and removing ocean echoes, and obtaining their time-frequency distribution results; extracting ionospheric disturbance time series based on the total electron content data obtained by GNSS, including: obtaining oblique path total electron content data within the monitoring range based on GNSS dual-frequency observation data using the carrier phase smoothing pseudorange method; converting the oblique path total electron content into vertical total electron content through the ionospheric thin layer model; processing the vertical total electron content data using the second-order numerical difference method to obtain the GNSS ionospheric disturbance time series; (3) Ionospheric information anomaly detection: the Mann-Kendall test method is used to detect anomaly information on the GNSS ionospheric disturbance time series: if an anomaly value is detected, the ionospheric time-frequency distribution result obtained based on the two-dimensional L-array HFSWR is judged to be abnormal; otherwise, it is considered that the monitoring area is not affected by the tsunami; the slope detection method is used to detect anomaly information on the ionospheric time-frequency distribution result obtained by the two-dimensional L-array HFSWR: if an anomaly value is detected, the "primary warning" stage is entered; otherwise, it is considered that the monitoring area is not affected by the tsunami; (4) Flow field information anomaly detection: the flow velocity information obtained based on the two-dimensional L-array HFSWR is tested for anomaly using the Q factor test method: if an abnormal value is detected, the "emergency warning" stage is entered; otherwise, it is considered that the monitoring area is not affected by the tsunami and everything is normal.

2. The multi-level tsunami early warning method combining two-dimensional L-array HFSWR and GNSS anomaly information detection according to claim 1 is characterized in that: The Mann-Kendall test method is used to detect abnormal information of the GNSS ionospheric disturbance time series, including: (1) Extract the vertical total electron content variation of a single station and use a bandpass filter to eliminate high-frequency noise while highlighting the fluctuation characteristics of gravity waves; (2) The Mann-Kendall test method is used to detect abnormal information in the vertical total electron content time series after bandpass filter processing in (1). The critical line of this test is determined by the normal distribution quantile corresponding to the significance level α = 0.05 as ±1.

96. The specific process is as follows: The calculated statistics are , Where n is the sample size of the time series, x i 、x j is the vertical total electron content at time i and j, sign is the sign function; The expected value of the calculated statistic is ; The variance of the calculated statistic is ; Calculate the standardized statistic as ; Calculating UF k The reverse sequence is ; In this way, UF is calculated k and its reverse UB k Two curves: If the two curves intersect and the intersection is within ±1.96 of the critical line, the time corresponding to the intersection is the moment when the mutation begins. At this time, continue to judge whether the ionospheric information obtained based on the two-dimensional L-array HFSWR is abnormal; if there is no intersection or the intersection exceeds the critical line range or there are multiple intersections, it is considered that the monitoring area is not affected by the tsunami.

3. The multi-level tsunami early warning method combining two-dimensional L-array HFSWR and GNSS anomaly information detection according to claim 1 is characterized in that: Anomaly information detection is performed on the ionospheric time-frequency distribution result obtained by the two-dimensional L-array HFSWR using a slope detection method, including: (1) Based on the RD spectrum obtained by the two-dimensional L-array HFSWR, the first-order ocean echo is identified and extracted using the 2D-SNR method. That is, the signal-to-noise ratio is calculated by selecting the echo intensity of the first-order spectrum area of ​​the two-dimensional window of the upper and lower adjacent distance cells of the current distance cell and the noise area of ​​the two surrounding two-dimensional windows. The first-order spectrum boundary is detected by highlighting the difference between the first-order spectrum and the surrounding noise using the characteristic of the continuous distribution of the entire first-order spectrum in distance. After extracting and removing the ocean echo, the ionospheric echo is extracted using the Ostu image segmentation method. For the extracted ionospheric echo, its time-frequency distribution result is obtained to capture the ionospheric traveling disturbance in the ionosphere. (2) Select the Doppler value corresponding to the maximum echo amplitude at each moment in the ionospheric time-frequency distribution result as the sampling point, and use cubic spline interpolation to fit the curve; then detect the slope change of the fitting curve to judge the ionospheric state: if the slope continues to change dramatically, it indicates that the ionospheric change is normal; if the slope does not change significantly, it is judged that the ionospheric traveling disturbance phenomenon has occurred in the ionosphere. Since the ionospheric traveling disturbance is periodic, only disturbances lasting more than 10 minutes are considered to be the result of tsunami; When ionospheric disturbances persist for more than 10 minutes, the ionospheric anomaly caused by the tsunami is deemed to have propagated into the monitoring range of the two-dimensional L-array HFSWR, entering the "primary warning" stage. The primary warning results can provide a basis for relevant departments to take further emergency measures. If no abnormal disturbances meeting the above conditions are detected, the monitoring area is considered to be unaffected by the tsunami.

4. The multi-level tsunami early warning method combining two-dimensional L-array HFSWR and GNSS anomaly information detection according to claim 1 is characterized in that: The Q factor test method is used to detect abnormal information of the flow velocity information obtained based on the two-dimensional L-array HFSWR, including: (1) For the extracted ocean echo, the velocity information is inverted based on the Bragg scattering principle. The specific formula is as follows: ， Where σ1 represents the echo power spectrum of the first-order scattering, ω represents the Doppler frequency, k0 represents the wave number, S represents the directional wave spectrum, m = ±1 represents the positive and negative directions of the first-order spectrum, δ represents the Dirac function, g represents the gravitational acceleration, and V cr Indicates flow rate; (2) Within the detection range of the two-dimensional L-array HFSWR flow field, a zone is divided every two kilometers, v b (t) is the flow velocity in zone b at time t, calculate the flow velocity v b (t) Average speed a in the past 5 minutes b (t) and standard deviation s b (t); (3) For each zone, calculate the velocity deviation index d relative to the background b (t), i.e. ， Then the speed deviation function D(t) is ； (4) For each regional band, calculate the velocity change index Δv in two adjacent time intervals from t−2δ to t b (t), where δ is the time resolution, set the value of δ to 5 minutes: ， Then the velocity increment function ΔV(t) ； (5) Calculate the correlation function C(t), which is used to describe the correlation between the velocities in three adjacent bands between time t, t−δ, and t−2δ. The value of C(t) is 1 unless the velocities from t−2δ to t−δ and from t−δ to t increase or decrease with time in all three bands, in which case the value of C(t) is set to 100. (6) Thus, the calculation formula of the tsunami detection factor q(t) at time t is: ， If q(t) ≥ threshold Q0, Q0 ≥ 200, the system enters the "emergency warning" stage; if no abnormal value is detected, the monitoring area is considered to be unaffected by the tsunami and everything is normal.

Images