Three-dimensional identification method and device for convection-triggered gravity wave

Abstract

This invention provides a method and apparatus for three-dimensional identification of convection-triggered gravity waves. The method includes: fusing multi-source meteorological observation data to generate three-dimensional grid data; identifying the movement path of the convective system based on the three-dimensional grid data; sampling the vertical profile of the three-dimensional grid data along the movement path within a preset statistical radius to obtain a time-series vertical profile set containing dynamic and microphysical parameters; performing physically constrained spectral filtering and wavenumber spectrum analysis on the meso-β-scale vertical wind shear time-series vertical profile in the time-series vertical profile set to extract gravity wave signals; calculating the spectral slope matching degree, the vertical coherence of gravity waves between adjacent height layers, and the frequency range matching degree of the gravity wave signals; and obtaining the comprehensive confidence level of the gravity wave signals based on the spectral slope matching degree, vertical coherence, and frequency range matching degree. This invention achieves objective, quantitative, and accurate three-dimensional identification of gravity waves.

Description

Technical Field

[0001] This invention relates to the field of meteorological monitoring technology, and in particular to a method and device for three-dimensional identification of convection-triggered gravity waves. Background Technology

[0002] Convection-triggered gravity waves are a key dynamic process connecting small- and medium-scale convective systems with mesoscale circulation. Accurate identification of their activity is of great significance for short-term forecasting and early warning of severe convective weather (such as thunderstorms, strong winds, and short-duration heavy precipitation).

[0003] However, existing gravity wave identification technologies mostly rely on subjective experience or spectral analysis of a single physical quantity, which presents the following prominent bottlenecks:

[0004] (1) Subjective identification and lack of objective quantitative criteria: It is difficult to achieve stable identification that is automated and business-oriented.

[0005] (2) Single analysis dimension: Most of them are limited to two-dimensional planes or single vertical cross sections, which cannot fully reveal the structure, propagation and evolution of gravity waves in three-dimensional space.

[0006] (3) The credibility of the results is difficult to assess: There is a lack of a quantitative system to distinguish the real gravity wave signal from environmental noise, turbulence and other disturbances. Summary of the Invention

[0007] To address the problems existing in the prior art, this invention provides a method and apparatus for three-dimensional identification of convection-triggered gravity waves, enabling objective, quantitative, and accurate three-dimensional identification of gravity waves.

[0008] This invention provides a method for three-dimensional identification of convection-triggered gravity waves, comprising:

[0009] Multi-source meteorological observation data are fused to generate three-dimensional grid data, and the movement path of the convective system is identified based on the three-dimensional grid data.

[0010] Vertical profile sampling is performed on the three-dimensional grid data within a preset statistical radius along the movement path to obtain a time-series vertical profile set containing dynamic and microphysical parameters;

[0011] Physically constrained spectral filtering and wavenumber spectrum analysis are performed on the meso-β-scale vertical wind shear time-series vertical profiles in the aforementioned time-series vertical profile set to extract gravity wave signals;

[0012] The spectral slope matching degree, vertical coherence of gravity waves between adjacent height layers, and frequency range matching degree of the gravity wave signal are calculated. Based on the spectral slope matching degree, vertical coherence, and frequency range matching degree, the overall confidence level of the gravity wave signal is obtained.

[0013] According to the three-dimensional identification method for convection-triggered gravity waves provided by the present invention, vertical profile sampling is performed on the three-dimensional grid data within a preset statistical radius along the movement path to obtain a time-series vertical profile set containing dynamic and microphysical parameters, including:

[0014] Differential propagation phase shift rate in the three-dimensional grid data The field searches for vertical cumulative values ​​within a preset statistical radius centered on the position at each moment in the movement path. The maximum horizontal position is taken as the centroid of the convective microphysics.

[0015] For the vertical wind shear VS field in the three-dimensional grid data, at each moment in the movement path The horizontal position with the largest vertical cumulative absolute VS within a preset statistical radius centered on the location is used as the optimal point for dynamic perturbation.

[0016] Vertical profiles of polarization are extracted at the convective microphysical centroid at each moment, and vertical profiles of dynamic parameters are extracted at the dynamic perturbation optimum at each moment.

[0017] Based on the vertical profiles of polarization and dynamic parameters, a time-series vertical profile set containing dynamic and microphysical parameters is generated.

[0018] According to the convective-triggered gravity wave three-dimensional identification method provided by the present invention, physical-constrained spectral filtering and wavenumber spectrum analysis are performed on the meso-β-scale vertical wind shear time-series vertical profiles in the time-series vertical profile set to extract gravity wave signals, including:

[0019] The sliding window method is used to divide the mid-β scale vertical wind shear time series vertical profile into multiple overlapping sequences, and the sequence at a specific height level within each sliding window is multiplied point by point with the window function to obtain the windowed sequence;

[0020] The complex spectrum and power spectral density of the gravity wave signal are obtained by performing a discrete Fourier transform on the windowed sequence.

[0021] The physical constraints that the frequency of the gravity wave signal must satisfy are determined based on key atmospheric parameters;

[0022] A bandpass filter is determined based on the physical constraints, and the filtered gravity wave spectrum is extracted from the complex spectrum using the bandpass filter.

[0023] The filtered gravity wave spectrum is subjected to inverse Fourier transform to reconstruct the time-domain gravity wave disturbance signal;

[0024] Calculate the filtered power spectral density based on the filtered gravity wave spectrum.

[0025] According to the convection-triggered three-dimensional recognition method for gravity waves provided by the present invention, the frequency of the gravity wave signal must satisfy the following physical constraint:

[0026]

[0027]

[0028] in, The frequency of the gravity wave signal is... and These are the minimum and maximum frequency thresholds of the gravity wave signal, respectively. It is the Coriolis frequency. The buoyancy frequency is near the top of the troposphere.

[0029] According to the three-dimensional identification method for convection-triggered gravity waves provided by the present invention, the boundary of the bandpass filter adopts a Gaussian transition to smooth the truncation effect, as shown in the formula:

[0030]

[0031] in, Let be the response function of the bandpass filter. It is the Nyquist frequency.

[0032] According to the convection-triggered gravity wave three-dimensional identification method provided by the present invention, the calculation of the spectral slope matching degree of the gravity wave signal includes:

[0033] The frequency range of the gravity wave signal Within the system, the filtered power spectral density is fitted with linear least squares in a double logarithmic coordinate system to obtain the slope of the fitted spectrum.

[0034] Calculate the degree of matching between the slope of the fitted spectrum and the classical value of the linear theory of gravity waves.

[0035] According to the convection-triggered gravity wave three-dimensional identification method provided by the present invention, the vertical coherence of gravity waves between adjacent height layers is calculated, including:

[0036] The time series of adjacent vertical layers in the time-domain gravity wave disturbance signal are standardized to obtain a standardized time series.

[0037] Zero-lag correlation coefficient, sign consistency rate, and waveform similarity are calculated for standardized time series of adjacent vertical layers;

[0038] The zero-hysteresis correlation coefficient, sign consistency rate, and waveform similarity are weighted and fused to obtain the vertical coherence of gravity waves.

[0039] According to the convection-triggered gravity wave three-dimensional identification method provided by the present invention, calculating the frequency range matching degree of the gravity wave signal includes:

[0040] Calculate the gravity wave signal in the frequency range The proportion of energy within the frequency band to the total energy of the entire frequency band is used as the frequency range matching degree.

[0041] The three-dimensional identification method for convection-triggered gravity waves provided by the present invention, after calculating the spectral slope matching degree of the gravity wave signal, the vertical coherence of gravity waves between adjacent height layers, and the frequency range matching degree of the gravity wave signal, further includes:

[0042] The spectral slope matching degree of the gravity wave signal, the vertical coherence of gravity waves between adjacent height layers, and the frequency range matching degree are fused with the microphysical field and dynamic field in the time-series vertical profile set for three-dimensional visualization.

[0043] This invention also provides a convection-triggered gravity wave three-dimensional recognition device, comprising:

[0044] The identification module is used to fuse multi-source meteorological observation data to generate three-dimensional grid data, and to identify the movement path of the convective system based on the three-dimensional grid data;

[0045] The sampling module is used to perform vertical profile sampling on the three-dimensional grid data within a preset statistical radius along the movement path to obtain a time-series vertical profile set containing dynamic and microphysical parameters.

[0046] The analysis module is used to perform physically constrained spectral filtering and wavenumber spectrum analysis on the mid-β-scale vertical wind shear time-series vertical profiles in the time-series vertical profile set to extract gravity wave signals.

[0047] The matching module is used to calculate the spectral slope matching degree, the vertical coherence of gravity waves between adjacent height layers, and the frequency range matching degree of the gravity wave signal. Based on the spectral slope matching degree, vertical coherence, and frequency range matching degree, the overall confidence level of the gravity wave signal is obtained.

[0048] The present invention provides a method and apparatus for three-dimensional identification of convection-triggered gravity waves. It proposes an objective identification process based on a time-series feature dataset of vertical profiles of classified strong convection polarization quantities (TSQVP), using the meso-β-scale vertical wind shear time-series vertical profiles (TSQVP) in the dataset. VSUsing physical constraints as the core input, the system objectively extracts gravity wave signals through spectral filtering and wavenumber spectrum analysis. It constructs a multi-index comprehensive confidence model, innovatively integrating spectral slope matching, vertical consistency, and frequency range matching to achieve quantitative scoring and grading of the existence of gravity waves and the reliability of their three-dimensional structure. Furthermore, it enables dynamic collaborative visualization of multiple physics fields, deeply integrating the identified three-dimensional features of gravity waves with the microphysics and dynamic fields in TSQVP to generate a dynamic three-dimensional visualization product that supports interactive operation.

[0049] In summary, this invention is the first to realize a complete objective technical process of "data-identification-assessment-visualization" for convection-triggered gravity waves, providing a new tool that can be applied in business to improve the accuracy and predictability of short-term warnings for severe convection. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0051] Figure 1 This is a flowchart illustrating the three-dimensional identification method for convection-triggered gravity waves provided by the present invention;

[0052] Figure 2 This is a visualization analysis diagram of the confidence level of gravity waves in the three-dimensional identification method for convection-triggered gravity waves with a statistical radius of 20km provided by the present invention.

[0053] Figure 3 This is a visualization analysis diagram of the confidence level of gravity waves in the three-dimensional identification method for convection-triggered gravity waves with a statistical radius of 40km provided by the present invention.

[0054] Figure 4 This is a three-dimensional volume drawing view of the gravity wave confidence level of the three-dimensional recognition method for convection-triggered gravity waves provided by the present invention;

[0055] Figure 5 This is a schematic diagram of the structure of the convection-triggered gravity wave three-dimensional recognition device provided by the present invention. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0057] The following is combined with Figure 1 A method for three-dimensional identification of convection-triggered gravity waves according to the present invention includes:

[0058] Step 101: Fuse multi-source meteorological observation data to generate three-dimensional grid data, and identify the movement path of the convective system based on the three-dimensional grid data;

[0059] Step 102: Sampling the three-dimensional grid data along the movement path within a preset statistical radius to obtain a time-series vertical profile set containing dynamic and microphysical parameters.

[0060] Step 103: Perform physical constraint spectral filtering and wavenumber spectrum analysis on the mid-β-scale vertical wind shear time-series vertical profile in the time-series vertical profile set to extract gravity wave signals;

[0061] Step 104: Calculate the spectral slope matching degree, vertical coherence of gravity waves between adjacent height layers, and frequency range matching degree of the gravity wave signal. Based on the spectral slope matching degree, vertical coherence, and frequency range matching degree, obtain the overall confidence level of the gravity wave signal.

[0062] The difficulty in this embodiment lies in:

[0063] 1) Difficult signal extraction: Gravity wave signals are usually weak and mixed in with background wind fields, turbulence and other synoptic-scale fluctuations. It is necessary to design sophisticated filtering algorithms to separate them from the time series data of TSQVP.

[0064] 2) Difficulty in three-dimensional verification: There is a lack of independent and direct observation methods to verify the three-dimensional structure of gravity waves. It is necessary to construct a comprehensive verification system with multiple indicators and clear physical meaning.

[0065] 3) Dynamic visualization is difficult: It is necessary to effectively integrate the fourth dimension of time evolution with the three-dimensional spatial structure and achieve synchronous and interactive visualization with the multiphysics field in TSQVP.

[0066] Based on the TSQVP dataset, this embodiment proposes a complete technical solution from structured data supply to three-dimensional wave signal identification and evaluation, and then to multi-field collaborative visualization, aiming to overcome the three major difficulties in gravity wave identification: weak signal extraction, three-dimensional structure verification, and dynamic visualization fusion.

[0067] First, a dedicated, standardized precursor dataset is constructed, namely the Time Series Feature Dataset of Vertical Profiles of Strong Convection Polarization (TSQVP). This dataset provides high-quality, structured input for this embodiment, and its core construction logic is as follows:

[0068] Multi-source data fusion: Integrating multi-source observation data such as regional dual-polarization weather radar network, automatic weather stations, and lightning locators, and through strict quality control and three-dimensional gridding stitching, a standardized grid field with high spatiotemporal resolution (1km horizontally, 0.5km vertically, and 6 minutes in time) is generated.

[0069] Precursor signal focusing: Starting from the location of the actual occurrence of severe convection, the improved cross-correlation tracking method is used to trace the movement path of the convective system 0-4 hours before the disaster, thereby locking in the key precursor time period and spatial area.

[0070] Multi-scale feature structuring: Along the tracking path, within different preset statistical radii, vertical profile sampling and averaging of 3D grid data are performed to generate a set of data including the horizontal reflectivity factor (Z). H ), differential reflectance (Z) DR ), differential propagation phase shift rate (K) DP This dataset is a collection of time-series vertical profiles of dynamic and microphysical parameters such as vertical wind shear (VS) and jet stream (JET) (variable definitions are shown in Table 1). This dataset systematically characterizes the spatiotemporal evolution of dynamics and microphysical structures before the occurrence of strong convection, laying a data foundation for automatically extracting precursor signals from massive amounts of data.

[0071] The construction of the TSQVP dataset mainly includes the following steps:

[0072] Input data source preparation: Acquire high spatiotemporal resolution dual-polarization weather radar network base data, wind profiler radar data, automatic weather station precipitation and wind field data, and lightning location data for the target area. Radar data must undergo rigorous quality control, including non-meteorological echo removal, system bias correction, phase defolding, and smoothing.

[0073] 3D Gridding and Convection Identification: Multiple radar base data are fused using a range-exponential weighted mosaic algorithm to generate a unified 3D Cartesian grid data updated every 6 minutes within the region. The grid has a horizontal resolution of 1 km and a vertical range of 1-21 km with a vertical resolution of 0.5 km. Based on this grid data, convection regions are identified using echo intensity and gradient characteristics on radar contour surfaces. Using the 6-minute contour surface wind field 3D grid data within the network, the 6-minute contour surface vertical wind shear 3D grid data is calculated by subtracting the lowest-level wind speed from the upper-level wind speed; a low-pass filtering method is used to calculate the 6-minute contour surface wind speed jet 3D grid data.

[0074] Constructing a Temporal Feature Dataset of Vertical Profiles of Strong Convection Polarization Quantities (TSQVP): This is the core foundational dataset for this embodiment. The construction process is as follows:

[0075] (1) Real-time label generation: Based on automatic station and lightning data, according to meteorological industry standards, generate a binary real-time grid field with a spatial resolution of 5 km for short-duration heavy precipitation (≥20 mm / h) and thunderstorm wind (gusts ≥17 m / s and accompanied by lightning).

[0076] (2) Tracing of convective systems: Starting from the actual location of the disaster, the improved cross-correlation tracking method (CLTREC) is used in conjunction with the correction of the significant feature location of dual polarization radar to trace the movement path of the convective system 0-4 hours before the disaster.

[0077] (3) Multi-scale vertical profile sampling: Along the tracking path, for each analysis time, the optimal vertical profile sampling of the three-dimensional grid data is performed within a set of preset statistical radii R ∈ {5, 10, 20, 30, 40, 50, 60} km.

[0078] (4) Calculation of feature parameters: The TSQVP dataset consists of the physical and dynamic vertical profile time series feature variables shown in Table 1. It is a standardized time series vertical profile dataset containing multi-scale information, providing structured input for subsequent analysis.

[0079] surface Temporal features of physical and dynamic vertical profiles in the TSQVP dataset

[0080]

[0081] This embodiment proposes an objective identification process based on TSQVP, using the meso-β-scale vertical wind shear time-series vertical profile (TSQVP) in the dataset. VS Using physical constraints as the core input, the system objectively extracts gravity wave signals through spectral filtering and wavenumber spectrum analysis. It constructs a multi-index comprehensive confidence model, innovatively integrating spectral slope matching, vertical consistency, and frequency range matching to achieve quantitative scoring and grading of the existence of gravity waves and the reliability of their three-dimensional structure. Furthermore, it enables dynamic collaborative visualization of multiple physics fields, deeply integrating the identified three-dimensional features of gravity waves with the microphysics and dynamic fields in TSQVP to generate a dynamic three-dimensional visualization product that supports interactive operation.

[0082] Based on the above embodiments, this embodiment performs vertical profile sampling on the three-dimensional grid data within a preset statistical radius along the movement path to obtain a time-series vertical profile set containing dynamic and microphysical parameters, including:

[0083] Differential propagation phase shift rate in the three-dimensional grid data The field searches for vertical cumulative values ​​within a preset statistical radius centered on the position at each moment in the movement path. The maximum horizontal position is taken as the centroid of the convective microphysics.

[0084] For the vertical wind shear VS field in the three-dimensional grid data, at each moment in the movement path The horizontal position with the largest vertical cumulative absolute VS within a preset statistical radius centered on the location is used as the optimal point for dynamic perturbation.

[0085] Vertical profiles of polarization are extracted at the convective microphysical centroid at each moment, and vertical profiles of dynamic parameters are extracted at the dynamic perturbation optimum at each moment.

[0086] Based on the vertical profiles of polarization and dynamic parameters, a time-series vertical profile set containing dynamic and microphysical parameters is generated.

[0087] This embodiment employs dual-field independent optimal sampling centered on time-series nodes for the 3D grid data. The analysis time is at the j-th time-series position within a 4-hour time series. Instead of sampling using a traditional single fixed point or nested method, it uses the reference position of the convective system's moving path determined by the tracking algorithm at that moment. Centered on the same predefined statistical radius Within, independent optimal searches are performed in two different three-dimensional physical quantity fields.

[0088] In three-dimensional differential propagation phase shift rate ( In the field, search for vertical accumulation Maximum horizontal position As the centroid of convection microphysics, it reflects the core region of condensate concentration and latent heat release. Its time-varying sequence { This is used to analyze the overall evolution and microphysical processes of convective systems. In equations (1) and (2), max represents the maximum search. Indicates Statistical radius centered on Search within a range; l represents the level, traversing from the bottom level 1 to the top level. .

[0089] (1)

[0090] In a three-dimensional vertical wind shear (VS) field, independently search for the horizontal position that maximizes the vertically accumulated absolute vertical wind shear. The location of the optimal point of dynamic disturbance at the current moment reflects the dynamic instability core most relevant to the excitation and propagation of waves. Its time-varying sequence { This is used for subsequent gravity wave signal extraction and analysis.

[0091] (2)

[0092] Based on the results of the independent search described above, two types of high-quality signal sequences are extracted in parallel:

[0093] Microphysical evolution sequence extraction: Microphysical centroid at each time step At this point, extract the complete polarization quantity ( , , Vertical profiles form the microphysical evolution time-series dataset in the TSQVP dataset: TSQVP ZH TSQVP ZDR and TSQVP KDP .

[0094] Dynamic disturbance sequence extraction: the optimal point of dynamic disturbance at each time step At this point, complete vertical profiles of dynamic parameters such as vertical wind shear and jet stream are extracted to form the dynamic disturbance time series dataset in the TSQVP dataset: TSQVP VS and TSQVP JET TSQVP VS It serves as the primary input signal for gravity wave identification.

[0095] Based on the above embodiments, this embodiment performs physically constrained spectral filtering and wavenumber spectrum analysis on the meso-β-scale vertical wind shear time-series vertical profiles in the time-series vertical profile set to extract gravity wave signals, including:

[0096] The sliding window method is used to divide the mid-β scale vertical wind shear time series vertical profile into multiple overlapping sequences, and the sequence at a specific height level within each sliding window is multiplied point by point with the window function to obtain the windowed sequence;

[0097] The complex spectrum and power spectral density of the gravity wave signal are obtained by performing a discrete Fourier transform on the windowed sequence.

[0098] The physical constraints that the frequency of the gravity wave signal must satisfy are determined based on key atmospheric parameters, and the boundaries are determined by the key atmospheric parameters.

[0099] A bandpass filter is determined based on the physical constraints, and the filtered gravity wave spectrum is extracted from the complex spectrum using the bandpass filter.

[0100] The filtered gravity wave spectrum is subjected to inverse Fourier transform to reconstruct the time-domain gravity wave disturbance signal;

[0101] Calculate the filtered power spectral density based on the filtered gravity wave spectrum.

[0102] This embodiment is based on TSQVP. VSThe data was processed independently for each altitude layer, and the wave components whose temporal variation patterns conformed to gravity wave theory were initially extracted. For a selected spatial location, the power spectral density calculation process for the vertical wind shear time series at a specific altitude layer was as follows: First, a Hanning window was used to reduce spectral leakage; then, a Discrete Fourier Transform (DFT) was performed on the windowed sequence to obtain its complex spectrum and power spectral density. The specific altitude layers were each altitude layer above 0°C.

[0103] First, the sliding window method is used to divide the continuous TSQVP. VS The time series was divided into a series of short, overlapping analysis periods. For each sliding window, the time series at a specific height level l was analyzed. , the original sequence Multiplying the window function point by point yields the windowed sequence. This is to reduce leakage errors in subsequent spectrum analysis.

[0104] (3)

[0105] Among them, the Hanning window function The discrete form is:

[0106]

[0107] in, The discrete-time index within the window is an integer. The length of the time series (number of data points) within the current analysis window, an integer.

[0108] Then, the windowed sequence Perform a Discrete Fourier Transform (DFT) to obtain its complex spectrum. and the corresponding power spectral density .

[0109]

[0110] (4)

[0111] in, For frequency index, integer, value range 0, 1, ... -1; in formula (4) The imaginary unit satisfies ; Discrete frequency, unit: Hz; For frequency Complex spectrum value at , unit and Consistent; For frequency The power spectral density at that point, in units of (m / s). 2 / Hz or dimensionless; For time resolution (6 minutes).

[0112] Based on the above embodiments, in this embodiment, according to the linear gravity wave theory, the physical constraint that the frequency of the gravity wave signal must satisfy is:

[0113]

[0114] (5)

[0115] in, The frequency of the gravity wave signal is... and These are the minimum and maximum frequency thresholds for gravity waves, respectively, in Hz; It is the Coriolis frequency. This is the Earth's rotational angular velocity. The latitude of the center of this study area; This represents the typical buoyancy frequency near the tropopause in a certain region during spring.

[0116] Based on the above embodiments, the bandpass filter in this embodiment uses a Gaussian transition at the boundary to smooth out the truncation effect, as shown in the formula:

[0117] (6)

[0118] in, Let be the response function of the bandpass filter, which is dimensionless; It is the Nyquist frequency.

[0119] Finally, the spectrum of the gravity wave components extracted by frequency domain multiplication can be expressed as:

[0120] (7)

[0121] in, The filtered gravity wave spectrum components, in units of . Consistent; Discrete frequency The filter response value at that point.

[0122] right Perform inverse Fourier transform to reconstruct the time-domain gravity wave disturbance signal:

[0123] (8)

[0124] In the formula: Represents the discrete inverse Fourier transform operator; The reconstructed time-domain gravity wave disturbance signal is a real number sequence.

[0125] The specific formula for calculating the inverse Fourier transform is as follows:

[0126] (9)

[0127] in, is the time-domain index, and N is the sequence length.

[0128] To assess whether the signal conforms to the theoretical spectral characteristics of gravity waves, the power spectral density of the filtered signal was analyzed. The calculation formula is:

[0129] (10)

[0130] Based on the above embodiments, this embodiment calculates the spectral slope matching degree of the gravity wave signal, including:

[0131] The frequency range of the gravity wave signal Within, the filtered power spectral density Perform linear least squares fitting in a log-log coordinate system to obtain the slope of the fitted spectrum. :

[0132] log (11)

[0133] in, To fit the spectral amplitude constant, This indicates logarithmic operations with base 10.

[0134] Calculate the degree of matching between the slope of the fitted spectrum and the classical value of the linear theory of gravity waves.

[0135] Spectral slope matching degree ( The slope of the observed spectrum is fitted by performing a spatial Fourier transform on the filtered signal data. And calculate its degree of matching with the classical value (-3) of the linear theory of gravity waves:

[0136] (12)

[0137] when A score of 1.0 is awarded when the value is -3.0; a deviation exceeding this value results in a score of 0.0. You get zero points when you score 1.5.

[0138] Based on the above embodiments, this embodiment calculates the vertical coherence of gravity waves between adjacent height layers, including:

[0139] The time series of adjacent vertical layers in the time-domain gravity wave disturbance signal are standardized to obtain a standardized time series.

[0140] Zero-lag correlation coefficient, sign consistency rate, and waveform similarity are calculated for standardized time series of adjacent vertical layers;

[0141] The zero-hysteresis correlation coefficient, sign consistency rate, and waveform similarity are weighted and fused to obtain the vertical coherence of gravity waves.

[0142] Vertical consistency ( The calculation of the vertical coherence of gravity waves between signals at adjacent altitude layers includes the following data processing and calculation steps:

[0143] Input the reconstructed time-domain gravity wave perturbation signal sequence of adjacent vertical layers and First, standardization is performed to obtain a standardized time series with zero mean and unit variance. , ;

[0144] (13)

[0145] in, This represents the i-th observation of the original time series; This represents the i-th value of the standardized time series. and These are the means of the corresponding original time series. and To correspond to the standard deviation of the original time series, M represents the number of time series data points.

[0146] Based on this, three similarity indices are then calculated simultaneously:

[0147] Zero-lag correlation coefficient: (14)

[0148] Sign consistency rate: (15)

[0149] Among them, the function Let be the indicator function, which is defined as:

[0150]

[0151] For symbolic functions, their definition is:

[0152]

[0153] Waveform similarity (16)

[0154] Finally, the results are fused together using fixed weights to obtain a comprehensive coherence determination for vertical consistency:

[0155] (17)

[0156] Based on the above embodiments, this embodiment calculates the frequency range matching degree of the gravity wave signal, including:

[0157] Calculate the gravity wave signal in the frequency range The proportion of energy within the specified frequency band to the total energy across the entire frequency band is used as the frequency range matching degree. ):

[0158] (18)

[0159] The weighted fusion model was used to calculate the overall confidence level of gravity waves. :

[0160] (19)

[0161] in, , and Let be the weighting coefficient, satisfying =1, preferred , , .

[0162] Based on the above embodiments, this embodiment, after calculating the spectral slope matching degree of the gravity wave signal, the vertical coherence of the gravity wave between adjacent height layers, and the frequency range matching degree of the gravity wave signal, further includes:

[0163] The spectral slope matching degree of the gravity wave signal, the vertical coherence of gravity waves between adjacent height layers, and the frequency range matching degree are fused with the microphysical field and dynamic field in the time-series vertical profile set for three-dimensional visualization.

[0164] Existing technologies lack sufficient visualization capabilities: there is a lack of effective means to coordinate, dynamically, and three-dimensionally display the identified gravity wave structure with key environmental fields (such as vertical wind shear and microphysical fields), which is not conducive to forecasters' understanding of its physical mechanisms and early warning value.

[0165] This embodiment visualizes the three-dimensional gravity wave comprehensive confidence score data identified above. The visualization method combines volume rendering and isosurface rendering techniques: the comprehensive confidence score of gravity waves is visualized... Displayed in 3D space using semi-transparent colored volume rendering, with color mapping representing confidence level and transparency representing signal strength. It dynamically shows the spatial distribution and evolution of gravity wave signals within 0-4 hours before a strong convection event. Interactive image rotation, scaling, and cross-sectioning operations are supported.

[0166] A typical warm-sector strong convective event on April 2, 2024, was selected for case application and analysis. This event, which lasted from 08:00 to 20:00 UTC on April 2, 2024, exhibited clear warm-sector convective characteristics and encompassed various disaster types, including thunderstorms and strong winds (widespread thunderstorms with wind speeds of 8-10 on the Beaufort scale from noon to night, with maximum wind speeds of 14 on the Beaufort scale) and localized short-duration heavy rainfall (maximum hourly rainfall of 31 mm / h). This provided a relatively ideal case condition for verifying the effectiveness and practicality of the gravity wave identification method.

[0167] Figures 2 to 4 The results demonstrate the effectiveness and operational potential of the proposed method: it successfully extracts gravity wave activity signals that are spatiotemporally closely coupled with the occurrence and development of severe convection from observational statistics (meso-β-scale vertical wind shear time series vertical profiles), with high-confidence signals appearing more than 3 hours in advance; providing new physical insights for classification and graded early warning. Therefore, the method of this patent provides a reliable technical tool for the routine and automated monitoring of gravity waves, a key precursor signal, in the spring warm-sector severe convection monitoring and early warning operations in Zhejiang.

[0168] The convection-triggered gravity wave three-dimensional recognition device provided by the present invention will be described below. The convection-triggered gravity wave three-dimensional recognition device described below can be referred to in correspondence with the convection-triggered gravity wave three-dimensional recognition method described above.

[0169] like Figure 5 As shown, the device includes an identification module 501, a sampling module 502, an analysis module 503, and a matching module 504, wherein:

[0170] The identification module 501 is used to fuse multi-source meteorological observation data to generate three-dimensional grid data, and to identify the movement path of the convective system based on the three-dimensional grid data;

[0171] The sampling module 502 is used to perform vertical profile sampling on the three-dimensional grid data within a preset statistical radius along the movement path to obtain a time-series vertical profile set containing dynamic and microphysical parameters.

[0172] Analysis module 503 is used to perform physically constrained spectral filtering and wavenumber spectrum analysis on the meso-β-scale vertical wind shear time-series vertical profile in the time-series vertical profile set to extract gravity wave signals;

[0173] The matching module 504 is used to calculate the spectral slope matching degree of the gravity wave signal, the vertical coherence of the gravity wave between adjacent height layers, and the frequency range matching degree of the gravity wave signal. Based on the spectral slope matching degree, vertical coherence, and frequency range matching degree, the comprehensive confidence level of the gravity wave signal is obtained.

[0174] This embodiment proposes an objective identification process based on TSQVP, using the meso-β-scale vertical wind shear time-series vertical profile (TSQVP) in the dataset. VS Using physical constraints as the core input, the system objectively extracts gravity wave signals through spectral filtering and wavenumber spectrum analysis. It constructs a multi-index comprehensive confidence model, innovatively integrating spectral slope matching, vertical consistency, and frequency range matching to achieve quantitative scoring and grading of the existence of gravity waves and the reliability of their three-dimensional structure. Furthermore, it enables dynamic collaborative visualization of multiple physics fields, deeply integrating the identified three-dimensional features of gravity waves with the microphysics and dynamic fields in TSQVP to generate a dynamic three-dimensional visualization product that supports interactive operation.

[0175] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for three-dimensional identification of convection-triggered gravity waves, characterized in that, include: Multi-source meteorological observation data are fused to generate three-dimensional grid data, and the movement path of the convective system is identified based on the three-dimensional grid data. Vertical profile sampling is performed on the three-dimensional grid data within a preset statistical radius along the movement path to obtain a time-series vertical profile set containing dynamic and microphysical parameters; Physically constrained spectral filtering and wavenumber spectrum analysis are performed on the meso-β-scale vertical wind shear time-series vertical profiles in the aforementioned time-series vertical profile set to extract gravity wave signals; The spectral slope matching degree, vertical coherence of gravity waves between adjacent height layers, and frequency range matching degree of the gravity wave signal are calculated. Based on the spectral slope matching degree, vertical coherence, and frequency range matching degree, the overall confidence level of the gravity wave signal is obtained.

2. The three-dimensional identification method for convection-triggered gravity waves according to claim 1, characterized in that, Vertical profile sampling is performed on the three-dimensional grid data within a preset statistical radius along the movement path to obtain a time-series vertical profile set containing dynamic and microphysical parameters, including: Differential propagation phase shift rate in the three-dimensional grid data The field searches for vertical cumulative values ​​within a preset statistical radius centered on the position at each moment in the movement path. The maximum horizontal position is taken as the centroid of the convective microphysics. For the vertical wind shear VS field in the three-dimensional grid data, at each moment in the movement path The horizontal position with the largest vertical cumulative absolute VS within a preset statistical radius centered on the location is used as the optimal point for dynamic perturbation. Vertical profiles of polarization are extracted at the convective microphysical centroid at each moment, and vertical profiles of dynamic parameters are extracted at the dynamic perturbation optimum at each moment. Based on the vertical profiles of polarization and dynamic parameters, a time-series vertical profile set containing dynamic and microphysical parameters is generated.

3. The three-dimensional identification method for convection-triggered gravity waves according to claim 1, characterized in that, Physically constrained spectral filtering and wavenumber spectrum analysis are performed on the meso-β-scale vertical wind shear time-series vertical profiles in the aforementioned time-series vertical profile set to extract gravity wave signals, including: The sliding window method is used to divide the mid-β scale vertical wind shear time series vertical profile into multiple overlapping sequences, and the sequence at a specific height level within each sliding window is multiplied point by point with the window function to obtain the windowed sequence; The complex spectrum and power spectral density of the gravity wave signal are obtained by performing a discrete Fourier transform on the windowed sequence. The physical constraints that the frequency of the gravity wave signal must satisfy are determined based on key atmospheric parameters; A bandpass filter is determined based on the physical constraints, and the filtered gravity wave spectrum is extracted from the complex spectrum using the bandpass filter. The filtered gravity wave spectrum is subjected to inverse Fourier transform to reconstruct the time-domain gravity wave disturbance signal; Calculate the filtered power spectral density based on the filtered gravity wave spectrum.

4. The three-dimensional identification method for convection-triggered gravity waves according to claim 3, characterized in that, The physical constraint that the frequency of the gravity wave signal must satisfy is: ； ； in, The frequency of the gravity wave signal is... and These are the minimum and maximum frequency thresholds of the gravity wave signal, respectively. It is the Coriolis frequency. The buoyancy frequency is near the top of the troposphere.

5. The three-dimensional identification method for convection-triggered gravity waves according to claim 4, characterized in that, The bandpass filter employs a Gaussian transition at its boundary to smooth out the truncation effect, as shown in the formula: ； in, Let be the response function of the bandpass filter. It is the Nyquist frequency.

6. The three-dimensional identification method for convection-triggered gravity waves according to claim 4, characterized in that, Calculating the spectral slope matching degree of the gravity wave signal includes: The frequency range of the gravity wave signal Within the system, the filtered power spectral density is fitted with linear least squares in a double logarithmic coordinate system to obtain the slope of the fitted spectrum. Calculate the degree of matching between the slope of the fitted spectrum and the classical value of the linear theory of gravity waves.

7. The three-dimensional identification method for convection-triggered gravity waves according to claim 3, characterized in that, Calculating the vertical coherence of gravity waves between adjacent height layers includes: The time series of adjacent vertical layers in the time-domain gravity wave disturbance signal are standardized to obtain a standardized time series. Calculate the zero-lag correlation coefficient, sign consistency rate, and waveform similarity for standardized time series from adjacent vertical layers; The zero-hysteresis correlation coefficient, sign consistency rate, and waveform similarity are weighted and fused to obtain the vertical coherence of gravity waves.

8. The three-dimensional identification method for convection-triggered gravity waves according to claim 4, characterized in that, Calculating the frequency range matching degree of the gravity wave signal includes: Calculate the gravity wave signal in the frequency range The proportion of energy within the frequency band to the total energy of the entire frequency band is used as the frequency range matching degree.

9. The three-dimensional identification method for convection-triggered gravity waves according to any one of claims 1-8, characterized in that, After calculating the spectral slope matching degree of the gravity wave signal, the vertical coherence of the gravity waves between adjacent height layers, and the frequency range matching degree of the gravity wave signal, the method further includes: The spectral slope matching degree of the gravity wave signal, the vertical coherence of gravity waves between adjacent height layers, and the frequency range matching degree are fused with the microphysical field and dynamic field in the time-series vertical profile set for three-dimensional visualization.

10. A convection-triggered gravity wave three-dimensional recognition device, characterized in that, include: The identification module is used to fuse multi-source meteorological observation data to generate three-dimensional grid data, and to identify the movement path of the convective system based on the three-dimensional grid data; The sampling module is used to perform vertical profile sampling on the three-dimensional grid data within a preset statistical radius along the movement path to obtain a time-series vertical profile set containing dynamic and microphysical parameters. The analysis module is used to perform physically constrained spectral filtering and wavenumber spectrum analysis on the mid-β-scale vertical wind shear time-series vertical profiles in the time-series vertical profile set to extract gravity wave signals. The matching module is used to calculate the spectral slope matching degree, the vertical coherence of gravity waves between adjacent height layers, and the frequency range matching degree of the gravity wave signal. Based on the spectral slope matching degree, vertical coherence, and frequency range matching degree, the overall confidence level of the gravity wave signal is obtained.

Images