A Typhoon Track Monitoring Method Integrating GNSS Atmospheric Precipitation and Fengyun 4B Brightness Temperature Disturbance Factors

Abstract

This invention discloses a typhoon path monitoring method that integrates GNSS atmospheric precipitation data and Fengyun 4B brightness temperature perturbation factors. First, observation data from satellite-borne GNSS reflection signals are acquired and preprocessed. Then, the latitude and longitude coordinates of the typhoon center are obtained through satellite-borne GNSS inversion based on the preprocessed observation data. Next, brightness temperature data from geostationary satellites is acquired and geometrically corrected, and the typhoon's precise maritime positioning result is obtained through processing. Ground-based GNSS observation data and precise solution products are collected to determine the center location of the typhoon during its landfall phase. Finally, the latitude and longitude coordinates of the typhoon center, the precise maritime positioning result, and the center location of the typhoon during its landfall phase are adaptively and elastically fused based on the typhoon's life cycle stage and the real-time availability of each data source. The fused monitoring points are then subjected to spatiotemporal consistency checks and trajectory smoothing, outputting a movement path covering the entire life cycle of the typhoon from deep sea to nearshore to landfall.

Description

Technical Field

[0001] This invention relates to the field of satellite remote sensing and meteorological disaster monitoring technology, specifically a typhoon path monitoring method that integrates GNSS atmospheric precipitation data and Fengyun 4B brightness temperature disturbance factors. Background Technology

[0002] Typhoons, as highly destructive weather systems, bring strong winds, torrential rains, and storm surges throughout their formation, movement, and landfall, posing a serious and direct threat to the safety of life and property, as well as socio-economic development in coastal areas. In recent years, influenced by global climate change, typhoons have exhibited significantly increased intensity and abrupt changes in their paths, drastically increasing the difficulty and pressure of flood and disaster prevention forecasting. Especially in the early stages of typhoon formation in deep-sea areas, in open seas during typhoon movement, and in the complex land-sea interface before and after landfall, continuous, real-time, and high-precision monitoring of typhoon paths around the clock has become a core requirement for improving the timeliness and accuracy of meteorological disaster warnings, and a key technological issue that urgently needs to be addressed in the field of meteorological disaster prevention and mitigation.

[0003] Currently, real-time monitoring methods for typhoon paths are mainly divided into three categories. Each method relies on different observation principles to achieve location, but due to the inherent characteristics of the technology, there are significant deficiencies in terms of monitoring range, continuity, and accuracy. Furthermore, existing technologies have not yet achieved effective fusion and complementarity of multi-source observation data, and cannot meet the monitoring needs of the entire typhoon life cycle. Specific technical problems are as follows:

[0004] Traditional weather radar and optical / infrared remote sensing satellite monitoring have inherent limitations. Coastal weather radar is the mainstream method for monitoring near-shore typhoons. Although it has high detection accuracy, its detection range is strictly limited by the electromagnetic wave line of sight, covering only a local near-shore area. It has no monitoring capability for typhoons in the open sea and deep sea areas, resulting in large-scale blind spots in deep-sea monitoring. While optical and infrared remote sensing satellites have a wide observation angle, their detection signals are easily blocked by thick cloud systems and heavy rainfall on the outer periphery of the typhoon. In the early stages of typhoon formation or when the structure is loose, it is difficult to penetrate clouds and rain to obtain effective information about the core wind field inside the typhoon. At the same time, these methods are mostly based on the geometry of the typhoon cloud system for positioning, which is easily affected by asymmetrical cloud systems, causing the positioning results to deviate from the true dynamic center of the typhoon, making it difficult to guarantee monitoring accuracy.

[0005] The method of tracking the moisture center of a typhoon using meteorological parameters such as PWV retrieved from ground-based GNSS has limitations in scene adaptation. Although it can achieve high-precision positioning in near-shore areas, it is limited by the deployment conditions of GNSS observation stations. Existing stations are mainly concentrated on land and coastal islands and reefs, while stations in the vast deep sea and offshore areas are extremely sparse or even non-existent. This makes it impossible to effectively capture the path changes of typhoons in the early stages of formation and during their movement at sea. There are serious monitoring blind spots in the offshore area, making it difficult to achieve continuous tracking of typhoons from the deep sea to the near shore.

[0006] Technical Shortcomings of Spaceborne GNSS Reflectance Measurement (CYGNSS) Monitoring Methods: Spaceborne GNSS reflectance measurement technology, represented by CYGNSS satellites, uses L-band signals that can penetrate clouds and rain to directly detect sea surface roughness, making it an important means of monitoring deep-sea typhoons. However, this method has two major problems: First, due to the limitation of satellite revisit frequency, it cannot provide continuous high-frequency real-time positioning, resulting in spatiotemporal discontinuities in monitoring and easily creating monitoring gaps in typhoon paths. Second, existing CYGNSS typhoon center extraction algorithms heavily rely on prior paths as initial values. Without accurate prior guidance, clustering algorithms are prone to getting stuck in local optima or even failing to locate, leading to low accuracy and reliability of initial typhoon monitoring and failing to provide stable, high-confidence initial anchor points for subsequent monitoring.

[0007] The existing multi-source data monitoring system suffers from gaps in integration and lacks synergy. Currently, the industry lacks a mature multi-source data fusion monitoring solution, failing to effectively integrate the advantages of spaceborne GNSS reflection signals, geostationary satellite brightness temperature data, and ground-based GNSS water vapor inversion data. Initial positioning using geostationary satellite infrared cloud images relies heavily on manual experience, lacking standardized automatic positioning strategies and resulting in large positioning errors. During the maritime monitoring phase, there is a lack of positioning correction methods based on typhoon dynamics, failing to address the deviation between the geometric center and the dynamic center. During typhoon landfall, satellite remote sensing signals are easily obstructed by terrain and complex cloud and rain interference, leading to decreased positioning accuracy, while high-precision water vapor monitoring data from ground-based GNSS is not fully utilized for seamless relay monitoring. Overall, the existing monitoring system exhibits multiple monitoring gaps and accuracy shortcomings throughout the entire typhoon lifecycle (deep sea-nearshore-landfall). The continuity, robustness, and overall accuracy of path monitoring cannot meet the practical application needs of meteorological disaster prevention and mitigation.

[0008] In summary, developing a technical method that can integrate the advantages of multi-source observation data from spaceborne GNSS, geostationary satellites, and ground-based GNSS, overcome the monitoring blind spots, accuracy deficiencies, and continuity problems of single methods, and achieve high-precision, all-weather, continuous, and seamless monitoring of typhoons throughout their entire life cycle has become an urgent need in the field of satellite remote sensing and meteorological disaster monitoring. Summary of the Invention

[0009] This invention relates to the field of satellite remote sensing and meteorological disaster monitoring technology, and specifically discloses a typhoon path monitoring method that integrates GNSS atmospheric precipitable water and Fengyun 4B brightness temperature disturbance factors. It aims to solve the technical problems of existing typhoon path monitoring methods, such as deep-sea monitoring blind spots, poor positioning continuity, deviation of geometric center from dynamic center, and insufficient monitoring accuracy in the landfall segment, so as to achieve high-precision, all-weather, continuous and seamless monitoring of the entire life cycle of typhoons from deep sea to nearshore to landfall.

[0010] To achieve the above objectives, the technical solution specifically adopted by the present invention is as follows:

[0011] A method for monitoring typhoon tracks that integrates GNSS atmospheric precipitation data and Fengyun 4B brightness temperature perturbation factors includes the following steps:

[0012] Step 1: Acquire observation data of the spaceborne GNSS reflection signal and preprocess it. Then, perform spaceborne GNSS inversion based on the preprocessed observation data to obtain the latitude and longitude coordinates of the typhoon center. The observation data includes normalized bistatic radar cross section, integral delay waveform leading edge slope, mirror point latitude and longitude, and timestamp information.

[0013] Step 2: Obtain brightness temperature data from geostationary satellites and perform geometric correction. Use a dual first guess strategy to determine the first guess center of the typhoon. Delineate the region of interest based on the first guess center. Calculate the brightness temperature gradient field and solve for divergence and curl within the region. Construct a brightness temperature perturbation factor characterizing the dynamic structure of the typhoon. Correct the first guess center based on the brightness temperature perturbation factor to obtain the precise maritime positioning result of the typhoon.

[0014] Step 3: Collect ground-based GNSS observation data and precise calculation products, use precise single-point positioning technology to calculate the total zenith delay, separate the zenith wet delay and invert it into atmospheric precipitable water; based on the spatiotemporal distribution characteristics of atmospheric precipitable water, determine the center location of the typhoon's landfall phase.

[0015] Step 4: Based on the typhoon's life cycle stage and the real-time availability of each data source, adaptively and elastically fuse the latitude and longitude coordinates of the typhoon center, the precise maritime positioning results of the typhoon, and the center position of the typhoon during the landfall stage obtained in Steps 1-3 respectively; perform spatiotemporal consistency verification and trajectory smoothing on the fused monitoring points, and output the movement path covering the entire life cycle of the typhoon from deep sea to nearshore to landfall.

[0016] Preferably, in step 1, the preprocessing method includes constructing a data quality control mask, removing outlier sampling points containing non-numerical values, and performing normalization processing.

[0017] Preferably, in step 1, the spaceborne GNSS inversion method is as follows: the preprocessed observation data is fused and the scattering index is calculated, and then the fused scattering index sequence is smoothed and filtered; then the region of interest is determined based on the smoothing and filtering results, and the data in the region of interest is normalized twice, and the latitude and longitude coordinates of the typhoon center are obtained by using the weighted centroid algorithm to invert the data.

[0018] Preferably, in the spaceborne GNSS inversion method, the normalization process is a maximum-minimum normalization method, which maps the observation data to a numerical range of 0 to 1; the smoothing filtering process uses an adaptive moving average filter, and the window size of the adaptive moving average filter is 7.

[0019] Preferably, in step 1, the method for determining the region of interest based on the smoothing filtering result is as follows: search for the global minimum value of the smoothed filtering sequence, and use the global minimum value as the anchor point to extract local data of a preset width as the region of interest, wherein the preset width is 17 sampling points; the weighted centroid algorithm is an exponentially decaying weighted centroid algorithm.

[0020] Preferably, the execution process of the exponentially strong attenuation weighted centroid algorithm is as follows: the fusion scattering index in the region of interest is locally normalized twice to obtain the local normalized value of each sampling point; based on the local normalized value, a cubic attenuation function with inverse weights is constructed to calculate the weight of each sampling point; and then the weighted centroid formula is used in combination with the latitude and longitude coordinates of each sampling point and the corresponding weight to calculate the latitude and longitude coordinates of the typhoon center.

[0021] Preferably, in step 2, the geostationary satellite is FY-4B, and the brightness temperature data is the equivalent blackbody temperature (TBB) data of channel 12 of the Advanced Geostationary Radiation Imager (AGRI) of FY-4B. The dual first guess strategy is as follows: if a valid inversion result from step 1 exists at the current moment, it is used as the first guess center; if not, density clustering is performed on the low-temperature cloud pixels, and the geometric center of the largest cluster is extracted as the first guess center.

[0022] As a preferred method, when performing density clustering on low-temperature cloud pixels, the DBSCAN density clustering algorithm is used. First, a search area with a preset geographical range is defined and invalid pixels are removed. Then, the brightness temperature threshold is adaptively determined using the Fisher criterion. After filtering out the set of low-temperature pixels, the clustering operation is performed.

[0023] Preferably, in step 2, the brightness temperature gradient field is calculated using the central difference method to obtain the brightness temperature gradient components in both longitude and latitude directions; the brightness temperature perturbation factor is the Euclidean norm of the divergence and curl of the brightness temperature gradient field, that is, the square root of the sum of the squares of the divergence and the curl.

[0024] As a preferred method, the method of correcting the initial guess center based on the brightness temperature perturbation factor is as follows: search for the geometric centroid of the annular extreme region of the brightness temperature perturbation factor, and use the geometric centroid to perform dynamic position correction on the initial guess center. The window range of the annular extreme region is 20km.

[0025] Preferably, in step 3, the precise calculation products include the precise ephemeris and precise clock error provided by IGS; the method for obtaining the zenith wet delay is as follows: the zenith static delay is calculated based on the station parameters using the Saastamoinen model, and the zenith wet delay is obtained by the difference between the total zenith delay and the zenith static delay.

[0026] Preferably, the process of converting zenith wet delay into atmospheric precipitable water is as follows: calculate the conversion coefficient, which includes liquid water density, water vapor gas constant, atmospheric refractive index constant and atmospheric weighted average temperature, and multiply the zenith wet delay by the conversion coefficient to obtain atmospheric precipitable water.

[0027] Preferably, in step 3, the method for determining the center location of the typhoon landfall stage based on the spatiotemporal distribution characteristics of atmospheric precipitable water is as follows: execute the maximum value positioning strategy, identify the ground-based GNSS station where the instantaneous maximum value of atmospheric precipitable water in the entire network is located, and take the geographical latitude and longitude of the station as the center location of the typhoon landfall stage.

[0028] Preferably, in step 4, the adaptive elastic fusion rule is as follows: when there are valid spaceborne GNSS inversion results in the deep sea, the spaceborne GNSS inversion results and the geostationary satellite precise positioning results are fused; in the deep sea period when spaceborne GNSS data is missing, the geostationary satellite precise positioning results are used; in the near-shore and landing phases, the ground-based GNSS water vapor inversion results are introduced to weight and correct other valid monitoring results.

[0029] Preferably, in step 4, the spatiotemporal consistency check is performed by: calculating the moving speed and direction of the typhoon center monitoring point at adjacent times, eliminating abnormal jump points that exceed the typhoon dynamic limits, and repairing them using effective point interpolation; the trajectory smoothing process uses the Kalman filter algorithm.

[0030] This invention has the following characteristics and beneficial effects:

[0031] By employing the method described in this invention, seamless high-precision monitoring of the entire life cycle of typhoons from "deep sea to nearshore to landfall" is effectively achieved through the flexible fusion of GNSS-PWV, Fengyun-4B satellite perturbation factors, and CYGNSS sea surface scattering characteristics. Compared with existing single monitoring methods, this invention utilizes an adaptive smoothing and nonlinear weighted CYGNSS inversion mechanism to effectively filter out random scintillation noise under high sea states. With prior path guidance, it can provide high-precision initial anchor points far superior to traditional cloud images. By leveraging the dual first-guessing strategy and brightness-temperature perturbation factor algorithm of Fengyun-4B satellite, it solves the initial positioning problem when CYGNSS data is missing through independent clustering, and corrects the defect of traditional cloud image geometric positioning being susceptible to asymmetric cloud interference through dynamic characteristics, significantly compensating for the monitoring blind spots caused by the sparseness of marine GNSS stations. During the typhoon landfall phase, a relay monitoring method based on GNSS ZWD inversion of high-precision PWV is adopted. Utilizing the physical characteristics of high water vapor convergence in the typhoon core area, the typhoon location is locked by tracking the high-value center of the spatiotemporal distribution of PWV across the entire network. This effectively overcomes the limitations of single satellite methods that are susceptible to terrain and complex cloud and rain obstruction during the landfall phase, achieving seamless and accurate tracking of the typhoon's process from ocean to land. In summary, this invention effectively avoids the risk of failure of a single method in a specific scenario by complementing the advantages of multiple data sources and flexibly integrating them, and greatly improves the continuity and robustness of typhoon path monitoring. Attached Figure Description

[0032] Figure 1 This is a flowchart of typhoon center inversion based on GNSS reflection signals in an embodiment of the present invention.

[0033] Figure 2 This is a flowchart of typhoon precision positioning based on the FY-4B brightness temperature perturbation factor in an embodiment of the present invention.

[0034] Figure 3 This is a flowchart of typhoon center monitoring based on GNSS-PWV inversion in an embodiment of the present invention.

[0035] Figure 4 This is a flowchart of the multi-source data elastic fusion and full lifecycle path generation in an embodiment of the present invention.

[0036] Figure 5 This invention provides a typhoon track monitoring method that integrates GNSS atmospheric precipitable water and Fengyun 4B brightness temperature disturbance factors—taking Typhoon "Mangkut" as an example. Detailed Implementation

[0037] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0038] A method for monitoring typhoon tracks that integrates GNSS atmospheric precipitable water data and Fengyun 4B brightness temperature perturbation factors, such as... Figure 1 As shown, it includes the following steps:

[0039] Step 1: Preprocessing of spaceborne GNSS reflection signals and inversion of typhoon center:

[0040] Specifically, L1-level observation data from spaceborne GNSS is acquired. This data includes the Normalized Bistatic Radar Cross Section (NBRCS), the slope of the leading edge of the integral delay waveform (LES), the latitude and longitude of the mirror point, and timestamp information. After preprocessing the observation data, the latitude and longitude coordinates of the typhoon center are obtained through fusion scattering index calculation, smoothing filtering, region of interest delineation, and weighted centroid algorithm inversion. The specific execution is as follows:

[0041] 1.1 Data Preprocessing: Construct a data quality control mask, traverse the latitude and longitude sequences of NBRCS, LES and mirror points, identify and remove outlier sampling points containing non-numerical (NaN) values ​​to ensure the numerical stability of subsequent calculations;

[0042] 1.2 Normalization and Calculation of Fusion Scattering Index: The NBRCS and LES, which have inconsistent dimensions, are mapped to the [0,1] interval using the maximum-minimum normalization method, and then the fusion scattering index is calculated. The calculation formula is as follows:

[0043] (1)

[0044] In the formula, i is the sampling point index. , These are the minimum and maximum statistical values ​​of NBRCS within this trajectory segment, respectively. , These are the minimum and maximum values ​​of LES within this trajectory segment, respectively.

[0045] 1.3 Smoothing Filtering: An adaptive moving average filter with a window size of 7 is used to smooth the fused scattering exponential sequence. Time-series smoothing is performed to filter out high-frequency random noise and retain low-value signal features. The smoothed sequence The calculation formula is:

[0046] (2)

[0047] In the formula, W is the size of the sliding window, W=7 in this embodiment, and p is the sampling point index of the smoothed sequence;

[0048] 1.4 Region of Interest Delineation: Searching for Smoothed Sequences global minimum value Using the global minimum value as the anchor point, 8 sampling points are extracted before and after the global minimum value to form local data with a window width of 17, and a region of interest (ROI) containing the core information of the typhoon is constructed.

[0049] 1.5 Weighted Centroid Algorithm for Typhoon Center Inversion: The fused scattering index within the region of interest is subjected to secondary local normalization to obtain the local normalized value for each sampling point. A cubic decay function with inverse weights is constructed based on the locally normalized values, and the weights of each sampling point are calculated. The formula is:

[0050] (3)

[0051] The latitude and longitude of the typhoon center are then calculated using the weighted centroid formula, as follows:

[0052] (4)

[0053] (5)

[0054] In the formula, The latitude of the typhoon center obtained through inversion; For the longitude of the typhoon center obtained by inversion, Each has its own weight.

[0055] Step 2: FY-4B brightness temperature data processing and precise typhoon location at sea:

[0056] Specifically, such as Figure 2 As shown, brightness temperature data from geostationary satellites is acquired and geometrically corrected. A dual-first-guessing strategy is used to determine the initial typhoon center. The initial guessing center is then dynamically corrected using a brightness temperature perturbation factor to obtain the precise maritime positioning result of the typhoon. In this embodiment, the geostationary satellite is FY-4B, and the brightness temperature data is the equivalent blackbody temperature (TBB) data from channel 12 of the Advanced Geostationary Orbital Radiation Imager (AGRI) on FY-4B. The specific execution is as follows:

[0057] 2.1 Geometric Correction of Brightness Temperature Data: Based on the CGCS2000 reference ellipsoid parameters and the satellite geocentric distance (42164km), rigorous geometric correction was performed on each TBB pixel using the scan angle (d,f). The intermediate variable Sn and geographic latitude and longitude (lat,lon) were calculated using the following formula:

[0058] (6)

[0059] In the formula, The satellite's geocentric distance is 42,164 km. a and b represent the semi-major axis of the Earth's reference ellipsoid at 6,378.137 km and 6,356.7523 km, respectively. The nominal longitude of the satellite's nadir point is indicated.

[0060] 2.2 Dual-Guess Strategy for Determining the First Guess Center: Determine if a valid onboard GNSS inversion result obtained in step 1 exists at the current time. If it exists, directly use it as the first guess center; otherwise, perform the following steps to determine the first guess center:

[0061] 2.2.1 Define a rectangular search area of ​​100°E~170°E and 5°N~45°N, and remove invalid TBB pixels with abnormal quality labels and fill values ​​within the area;

[0062] 2.2.2 The brightness temperature threshold is adaptively determined using the Fisher criterion to filter out the set of low-temperature cloud pixels within the search area;

[0063] 2.2.3 The DBSCAN density clustering algorithm is used to cluster the low-temperature cloud pixel set, remove the interference of scattered and fragmented cloud systems, extract the largest cluster with the most pixels in the clustering results, calculate the geometric center of the largest cluster and use it as the first guess center;

[0064] 2.3 Construction of Brightness-Temperature Perturbation Factor: A circular region of interest with a radius of 300 km was delineated based on the initial typhoon center. Within this region, the brightness-temperature gradient field, divergence, and curl were calculated to construct a brightness-temperature perturbation factor characterizing the dynamic structure of the typhoon eyewall.

[0065] 2.3.1 Calculate the brightness temperature gradient components in the longitude and latitude directions of TBB data at pixel location (d,f) using the central difference method. , The formula is:

[0066] (7)

[0067] (8)

[0068] In the formula, and These represent the pixel spatial intervals in the longitude and latitude directions, respectively. This gradient vector field It characterizes the maximum rate of change of brightness temperature in a two-dimensional plane and its direction.

[0069] 2.3.2 Based on brightness temperature gradient components , The divergence divG(d,f) and curl curlG(d,f) of the brightness temperature gradient field are calculated using the following formulas:

[0070] (9)

[0071] (10)

[0072] 2.3.3 Construct the brightness temperature perturbation factor P(d,f), which is the Euclidean norm of divergence and curl, and the formula is:

[0073] (11)

[0074] 2.4 Dynamic Correction of the Initial Prediction Center: Search for the maximum region of the brightness temperature disturbance factor P(d,f), delineate an annular extreme region with a window range of 20km outside the maximum region, calculate the geometric centroid of the annular extreme region, use the geometric centroid to correct the dynamic position of the initial prediction center, and output the corrected maritime precise positioning result of Typhoon FY-4B.

[0075] Step 3: Ground-based GNSS observation data processing and typhoon landfall monitoring:

[0076] Specifically, such as Figure 3 As shown, ground-based GNSS observation data and precise solution products were collected. Precise single-point positioning technology was used to calculate the total zenith delay. After separating the wet zenith delay, it was inverted into precipitable water volume (PWV). Based on the spatiotemporal distribution characteristics of PWV, the center location of the typhoon's landfall phase was determined. The specific execution is as follows:

[0077] 3.1 Data Acquisition: Collect raw ground-based GNSS observation data (including pseudorange and carrier phase) from coastal and island stations, and obtain precise calculation products such as precise ephemeris and precise clock error provided by IGS;

[0078] 3.2 Zenith Delay Calculation: Precise Point Positioning (PPP) technology was used to process the combined ionospheric observations. While estimating the receiver position coordinates and clock error, a high-precision total zenith delay (ZTD) was calculated. Using the Saastamoinen model, the latitude (in radians), altitude, and sea-level pressure parameters of the station were substituted to calculate the zenith static delay (ZHD). The formula is as follows:

[0079] (12)

[0080] In the formula, Pg is the pressure at sea level where the receiver is located, Hn is the latitude of the receiver, and H is the altitude of the receiver.

[0081] The zenith wet delay (ZWD) is obtained by separating the total zenith delay from the zenith static delay, as shown in the formula:

[0082] (13)

[0083] 3.3 ZWD Inversion to PWV: Based on the physical conversion relationship between ZWD and PWV, ZWD is inverted into PWV with a clear physical meaning. First, the conversion coefficient Π is calculated using the following formula:

[0084] (14)

[0085] In the formula, For conversion factors, The density of liquid water is given by a value of . , Let be the water vapor gas constant, with a value of . , , Here is the atmospheric refractive index constant, and its empirical value is usually 1. , Tm represents the weighted average temperature of the atmosphere, which can be estimated using a global empirical model.

[0086] Finally, according to ZWD, The conversion relationship between PWV and PWV can be used to obtain PWV. The specific calculation formula is as follows:

[0087] (15)

[0088] Save the PWV and corresponding time values ​​calculated above for direct use in later steps.

[0089] 3.4 Typhoon landfall location determination: By traversing the PWV inversion values ​​of all stations in the coastal monitoring network, a spatiotemporal distribution field of water vapor in the typhoon landfall area is constructed; by executing the maximum value positioning strategy, the ground-based GNSS station where the instantaneous maximum PWV value of the entire network is located is automatically identified, and the geographical latitude and longitude of the station is directly determined as the center location of the typhoon landfall stage.

[0090] Step 4: Elastic fusion of multi-source data and generation of the typhoon's full lifecycle path:

[0091] Specifically, such as Figure 4 As shown, based on the different life cycle stages of the typhoon (deep sea, nearshore, and land), and the real-time availability of the satellite-borne GNSS inversion results (step 1), the FY-4B precise positioning results (step 2), and the ground-based GNSS-PWV inversion results (step 3), the three types of monitoring results are adaptively and elastically fused. After spatiotemporal consistency verification and trajectory smoothing, the movement path covering the entire life cycle of the typhoon is output. The specific execution is as follows:

[0092] 4.1 Adaptive and Flexible Fusion: Multi-source data is fused according to the principle of "penetration priority, cloud image relay, and ground-based dominance":

[0093] 4.1.1 When the typhoon is in the deep sea stage and there are effective satellite GNSS inversion results, the satellite GNSS inversion results and FY-4B fine positioning results are integrated, with the satellite GNSS results as the benchmark and the FY-4B results as the correction.

[0094] 4.1.2 When a typhoon is in the deep-sea stage but spaceborne GNSS data is missing, the FY-4B precise positioning results are directly used to fill the spatiotemporal monitoring blind spots;

[0095] 4.1.3 When the typhoon is near the coast or during the landfall phase, the ground-based GNSS-PWV inversion results are introduced to weight and correct the effective monitoring results of the spaceborne GNSS / FY-4B, with the ground-based GNSS-PWV results as the core.

[0096] 4.2 Spatiotemporal consistency check: Physical consistency check is performed on the fused discrete typhoon center monitoring point sequence. The moving speed and direction of the typhoon center monitoring points at adjacent times are calculated. If an abnormal jump point is detected that exceeds the dynamic limit of the typhoon, it is judged as a gross error and removed. Linear interpolation is used to repair it using the effective monitoring points at the previous and next times.

[0097] 4.3 Trajectory Smoothing: The Kalman filter algorithm is used to smooth the interpolated and repaired monitoring point sequence to eliminate random positioning errors and generate a spatiotemporally continuous and smooth typhoon path;

[0098] 4.4 Output Results: The output covers the high-precision movement path of the typhoon throughout its entire life cycle, from deep sea to nearshore to landfall, including the latitude and longitude coordinates of the typhoon center, movement speed, and direction at each time point.

[0099] In this embodiment, taking Typhoon "Mangkut" as an example, the typhoon path monitoring error results based on Typhoon Mangkut are shown in Table 1.

[0100]

[0101] Table 1 shows the typhoon track monitoring error results based on Typhoon Mangkhut.

[0102] As can be seen from Table 1, the typhoon track monitoring method of the present invention, which integrates GNSS atmospheric precipitable water and Fengyun 4B brightness temperature perturbation factors, has achieved good positioning results through actual monitoring of Typhoon Mangkhut. Table 1 includes two parts: a table of Typhoon Mangkhut track monitoring error data and a comparison diagram of the spatial distribution of the typhoon's true center and inverted center. This visually presents the positioning accuracy and actual monitoring effect of the method of the present invention. Figure 5 The following core information was obtained from Table 1:

[0103] I. Quantitative error data reflects that the present invention has high positioning accuracy.

[0104] The error data table uses the monitoring time dimension from 6:00 AM on September 4th to 6:00 AM on September 7th, 2024, to simultaneously record the wind speed of Typhoon Mangkhut at different monitoring times and the positioning error (km) of the method of this invention. It covers 14 key monitoring time nodes, with wind speeds ranging from 45 m / s to 62 m / s, encompassing the high-intensity movement phase of the typhoon, and quantitatively demonstrating the positioning performance of this invention.

[0105] The positioning error is generally low: within the monitoring range of 45m / s to 62m / s during Typhoon Mangkhut, the path monitoring error of this invention is mostly controlled within 30km, with the error concentrated in the range of 10km to 25km at most times. For example, the error was only 16.16km at 18:00 on September 4 when the wind speed was 58m / s and only 9.99km at 15:00 on September 5 when the wind speed was 62m / s. Even during the strong wind phase of the typhoon with wind speeds of 62m / s, it can still maintain a low positioning deviation, indicating that this invention still has stable positioning accuracy in high sea state and strong typhoon scenarios.

[0106] Multiple high-precision positioning nodes exist: During the monitoring process, there were multiple moments with positioning errors of less than 10km, such as 6.27km at 9:00 AM on September 5th when the wind speed was 58m / s and 6.98km at 0:00 AM on September 7th when the wind speed was 55m / s. The positioning accuracy of these nodes has reached the high-precision requirements for typhoon monitoring, proving that the present invention can effectively improve the accuracy of typhoon center positioning through multi-source data fusion and dynamic correction.

[0107] The positioning error showed no significant positive correlation with typhoon wind speed: as the typhoon wind speed increased from 45 m / s to 62 m / s, the positioning error did not show a significant upward trend with the increase in wind speed. For example, when the wind speed increased from 58 m / s to 62 m / s, the error was 17.94 km at 0:00 on September 5 and only 9.99 km at 15:00 on September 5. This indicates that the core components of this invention, such as the spaceborne GNSS signal processing and FY-4B brightness temperature disturbance factor correction, can effectively overcome interference from sea surface signal noise and cloud asymmetry caused by high wind speeds, ensuring monitoring stability.

[0108] Based on the above quantitative error data, the practicality, stability and high accuracy of the typhoon path monitoring method of this invention, which integrates GNSS atmospheric precipitable water and Fengyun 4B brightness temperature disturbance factors, have been fully verified. Compared with traditional single monitoring methods, this invention effectively solves technical problems such as monitoring blind spots in the deep sea, poor positioning continuity and deviation of geometric center from dynamic center through the complementary advantages of multi-source data. In actual typhoon monitoring, it can achieve seamless and high-precision path monitoring throughout the entire life cycle from deep sea to nearshore to landfall.

[0109] Meanwhile, this embodiment uses Typhoon Mangkhut as a real monitoring case, providing real data support for the engineering application of the method of the present invention, proving that the present invention can maintain stable monitoring accuracy at different intensities and different stages of movement of typhoons, and can be effectively applied to typhoon disaster monitoring and early warning work in coastal areas of my country, providing reliable technical support for flood control and disaster prevention.

[0110] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for monitoring typhoon paths that integrates GNSS atmospheric precipitation data and the Fengyun 4B brightness temperature perturbation factor, characterized in that, Includes the following steps: Step 1: Acquire observation data of the spaceborne GNSS reflection signal and preprocess it. Then, perform spaceborne GNSS inversion based on the preprocessed observation data to obtain the latitude and longitude coordinates of the typhoon center. The observation data includes normalized bistatic radar cross section, integral delay waveform leading edge slope, mirror point latitude and longitude, and timestamp information. Step 2: Obtain brightness temperature data from geostationary satellites and perform geometric correction. Use a dual first guess strategy to determine the first guess center of the typhoon. Delineate the region of interest based on the first guess center. Calculate the brightness temperature gradient field and solve for divergence and curl within the region. Construct a brightness temperature perturbation factor characterizing the dynamic structure of the typhoon. Correct the first guess center based on the brightness temperature perturbation factor to obtain the precise maritime positioning result of the typhoon. Step 3: Collect ground-based GNSS observation data and precise calculation products, use precise single-point positioning technology to calculate the total zenith delay, separate the zenith wet delay and invert it into atmospheric precipitable water; based on the spatiotemporal distribution characteristics of atmospheric precipitable water, determine the center location of the typhoon's landfall phase. Step 4: Based on the typhoon's life cycle stage and the real-time availability of each data source, adaptively and elastically fuse the latitude and longitude coordinates of the typhoon center, the precise maritime positioning results of the typhoon, and the center position of the typhoon during the landfall stage obtained in Steps 1-3 respectively; perform spatiotemporal consistency verification and trajectory smoothing on the fused monitoring points, and output the movement path covering the entire life cycle of the typhoon from deep sea to nearshore to landfall.

2. The typhoon path monitoring method according to claim 1, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factor, is characterized in that... In step 1, the preprocessing method includes constructing a data quality control mask, removing outlier sampling points containing non-numerical values, and performing normalization processing.

3. The typhoon path monitoring method according to claim 1, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factors, is characterized in that... In step 1, the spaceborne GNSS inversion method is as follows: the preprocessed observation data is fused and the scattering index is calculated, and then the fused scattering index sequence is smoothed and filtered. Based on the smoothing filter results, the region of interest is determined. After the data within the region of interest is normalized twice, the latitude and longitude coordinates of the typhoon center are obtained by weighted centroid algorithm.

4. The typhoon path monitoring method according to claim 3, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factor, is characterized in that... In the preprocessing method, the normalization process is a max-min normalization method, which maps the observed data to a numerical range of 0 to 1; the smoothing filtering process uses an adaptive moving average filter, and the window size of the adaptive moving average filter is 7.

5. The typhoon path monitoring method according to claim 3, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factor, is characterized in that... In step 1, the method for determining the region of interest based on the smoothing filtering result is as follows: search for the global minimum value of the smoothed filtering sequence, and use the global minimum value as the anchor point to extract local data of a preset width as the region of interest, where the preset width is 17 sampling points; the weighted centroid algorithm is the exponentially decaying weighted centroid algorithm.

6. The typhoon path monitoring method according to claim 5, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factor, is characterized in that... The execution process of the exponentially strong attenuation weighted centroid algorithm is as follows: the fusion scattering index in the region of interest is locally normalized twice to obtain the local normalized value of each sampling point. Based on the local normalized value, a cubic attenuation function with inverse weights is constructed to calculate the weight of each sampling point. Then, the weighted centroid formula is used in combination with the latitude and longitude coordinates of each sampling point and the corresponding weight to calculate the latitude and longitude coordinates of the typhoon center.

7. The typhoon track monitoring method according to claim 1, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factor, is characterized in that... In step 2, the geostationary satellite is FY-4B, and the brightness temperature data is the equivalent blackbody temperature (TBB) data of channel 12 of the Advanced Geostationary Radiation Imager (AGRI) of FY-4B. The dual first guess strategy is as follows: if there is a valid inversion result from step 1 at the current moment, it is used as the first guess center; if not, density clustering is performed on the low-temperature cloud pixels, and the geometric center of the largest cluster is extracted as the first guess center.

8. The typhoon path monitoring method according to claim 7, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factor, is characterized in that... When performing density clustering on low-temperature cloud pixels, the DBSCAN density clustering algorithm is used. First, a search area with a preset geographical range is defined and invalid pixels are removed. Then, the Fisher criterion is used to adaptively determine the brightness temperature threshold. After filtering out the set of low-temperature pixels, the clustering operation is performed.

9. The typhoon track monitoring method according to claim 1, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factor, is characterized in that... In step 2, the brightness temperature gradient field is calculated using the central difference method to obtain the brightness temperature gradient components in the longitude and latitude directions; the brightness temperature perturbation factor is the Euclidean norm of the divergence and curl of the brightness temperature gradient field, that is, the square root of the sum of the squares of the divergence and the curl.

10. The typhoon path monitoring method according to claim 9, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factor, is characterized in that... The method for correcting the initial guess center based on the brightness temperature perturbation factor is as follows: search for the geometric centroid of the annular extreme region of the brightness temperature perturbation factor, and use this geometric centroid to perform dynamic position correction on the initial guess center. The window range of the annular extreme region is 20km.

11. The typhoon track monitoring method according to claim 10, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factor, is characterized in that... In step 3, the precise calculation products include the precise ephemeris and precise clock error provided by IGS; the method for separating the zenith wet delay is as follows: the zenith static delay is calculated based on the station parameters using the Saastamoinen model, and the zenith wet delay is obtained by the difference between the total zenith delay and the zenith static delay.

12. The typhoon path monitoring method according to claim 11, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factor, is characterized in that... The process of converting zenith wet delay into atmospheric precipitable water is as follows: calculate the conversion coefficient, which includes liquid water density, water vapor gas constant, atmospheric refractive index constant, and atmospheric weighted average temperature, and multiply the zenith wet delay by the conversion coefficient to obtain atmospheric precipitable water.

13. The typhoon path monitoring method according to claim 1, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factor, is characterized in that... In step 3, the method for determining the center location of the typhoon landfall stage based on the spatiotemporal distribution characteristics of atmospheric precipitable water is as follows: execute the maximum value positioning strategy, identify the ground-based GNSS station where the instantaneous maximum value of atmospheric precipitable water in the entire network is located, and take the geographical latitude and longitude of the station as the center location of the typhoon landfall stage.

14. The typhoon track monitoring method according to claim 1, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factor, is characterized in that... In step 4, the rule for adaptive elastic fusion is: when in the deep sea and there are effective spaceborne GNSS inversion results, the spaceborne GNSS inversion results are fused with the geostationary satellite precise positioning results; In the deep-sea period where satellite-borne GNSS data is missing, geostationary satellite precise positioning results are used; in the near-shore and landing phases, ground-based GNSS water vapor inversion results are introduced to weight and correct other effective monitoring results.

15. The typhoon path monitoring method according to claim 1, which integrates GNSS atmospheric precipitation and Fengyun 4B brightness temperature disturbance factor, is characterized in that... In step 4, the spatiotemporal consistency check is performed by: calculating the moving speed and direction of the typhoon center monitoring point at adjacent times, eliminating abnormal jump points that exceed the typhoon dynamic limits, and repairing them using effective point interpolation; the trajectory smoothing process uses the Kalman filter algorithm.

Images