Atmospheric boundary layer turbulence statistic inversion method and system based on 6-beam ground-based light quantum wind lidar DBS scanning mode, medium and program product

Abstract

The invention discloses an atmospheric boundary layer turbulence statistic inversion method and system based on a 6-beam ground-based light quantum wind measurement laser radar DBS scanning mode, a medium and a program product, and relates to the technical field of atmospheric remote sensing and boundary layer detection. A multi-direction radial wind speed time sequence is collected, and the radial speed variance of each beam is calculated on the basis of data quality control; and further establishing a relationship between the velocity variance and the covariance of the velocity in the three directions, constructing an observation equation set and inverting the turbulence statistics, and combining coordinate rotation to obtain the turbulence statistics in the downwind direction, the transverse wind direction and the vertical direction. And finally, calculating turbulent flow kinetic energy, and outputting a turbulent flow statistic vertical profile of each height layer. According to the method, continuous monitoring and high-precision description of the atmospheric boundary layer turbulence statistics can be realized, and the requirements of a service system on stability, precision and deployability of a boundary layer turbulence inversion product are met.

Description

Technical Field

[0001] The present invention belongs to the field of atmospheric remote sensing and boundary layer detection technology, and relates to laser radar wind field detection and turbulence parameter inversion technology. Specifically, it is a method, system, medium and program product for inverting atmospheric boundary layer turbulence statistics based on a 6-beam ground-based photon wind measurement laser radar DBS scanning mode. Background Art

[0002] The atmospheric boundary layer is a key area for the exchange of matter and energy between the surface and the free atmosphere. Its turbulent structure and evolution have an important impact on weather forecasting, air quality assessment, wind energy resource development, and aviation safety. Therefore, accurate and continuous acquisition of the vertical distribution characteristics of turbulent statistics (such as velocity variance, momentum flux, turbulent kinetic energy, etc.) in the atmospheric boundary layer is of great theoretical and practical significance for understanding boundary layer dynamics, developing boundary layer parameterization schemes, and establishing high-precision numerical models.

[0003] Traditional means of observing the atmospheric boundary layer mainly include soundings, boundary layer towers, radar wind profilers, aircraft platforms and satellite remote sensing. Soundings and tower observations can obtain vertical distribution information of basic meteorological elements such as wind speed, temperature and humidity, but they are limited by low temporal resolution and limited spatial coverage, and it is difficult to meet the needs of continuous inversion of turbulence statistics. Although aircraft observations have high accuracy, they are only suitable for short-term, local area studies due to high cost and low frequency. Satellite remote sensing plays a role in large-scale meteorological monitoring, but its spatial resolution and ability to penetrate the low altitude of the boundary layer are still limited.

[0004] In recent years, with the development of lidar technology, especially the widespread application of photon Doppler wind lidar, it has become possible to obtain three-dimensional wind fields with high temporal and spatial resolution. This technology uses the Doppler frequency shift to invert wind speed information by emitting a laser beam and receiving the scattering signal of the laser by atmospheric aerosols. In actual observations, Doppler lidar can use different scanning modes to achieve wind field inversion and turbulence parameter estimation. Common modes include velocity azimuth display (VAD), elevation scanning (Range Height Indicator, RHI) and Doppler beam swinging scanning (DBS). VAD and RHI methods can be used to invert the average wind field and vertical velocity structure, but their accuracy in inverting turbulent statistics is limited. The DBS method obtains radial velocities in multiple directions by setting multiple laser beams with fixed azimuths and elevations, thereby inverting parameters such as velocity variance, momentum flux and turbulent kinetic energy.

[0005] However, the traditional DBS method mostly uses a 4-beam configuration, and its inversion results are easily affected by the angle distribution between beams, wind field non-uniformity and atmospheric disturbances, resulting in insufficient calculation accuracy of turbulence statistics. In the prior art, Chinese patent CN109814131B proposes a laser radar turbulence parameter inversion method based on discontinuous conical scanning, which reduces the speed measurement error of radial wind speed by pausing the scan, and uses the Kolmogorov model combined with the velocity structure function to eliminate the noise term, but its scanning mode is still mainly based on a single fixed zenith angle, and it is difficult to obtain the three-dimensional turbulence structure of the entire boundary layer. Its discontinuous scanning strategy causes the laser beam to have a window area in time, which limits the ability to fully capture high-frequency turbulent disturbances and is not conducive to reconstructing the continuous vertical profile of turbulence statistics in the entire boundary layer. In addition, some studies have attempted to increase the number of beams to 5 or 6 to enhance the inversion capability, but there is still a lack of standardized guidance on scanning strategies and parameter settings (such as beam azimuth, elevation angle, time interval, and spatial resolution). Especially in the deployment of operational observations, there is no mature solution for how to balance scanning efficiency and inversion accuracy. In addition, existing studies have focused on products such as average wind speed and boundary layer height. High-precision inversion products for turbulence statistics have not been systematically established, and data quality control and result reliability assessment are still insufficient.

[0006] In summary, the existing atmospheric boundary layer turbulence observation methods have obvious deficiencies in terms of coverage height, spatiotemporal resolution, inversion accuracy and product systematization. Therefore, how to build a turbulence statistics inversion method and system with high accuracy, strong stability and suitable for business deployment based on a ground-based photon wind lidar system with high spatiotemporal resolution is a technical problem that needs to be solved urgently. Summary of the invention

[0007] 1. Purpose of the invention In view of the above-mentioned defects and shortcomings of the prior art, the present invention aims to provide a method, system, medium and program product for inverting turbulence statistics in the atmospheric boundary layer based on the DBS scanning mode of a 6-beam ground-based photon wind measurement lidar. The method adopts 6-beam laser synchronous measurement and DBS scanning mode to obtain the spatiotemporal distribution of the atmospheric three-dimensional wind field in the boundary layer with high spatiotemporal resolution, and constructs a radial velocity variance matrix containing multi-beam observation directions to invert the turbulence statistics of the atmospheric boundary layer such as velocity variance, momentum flux, turbulent kinetic energy, etc., and introduces coordinate rotation to unify the results, thereby outputting the turbulence vertical profile with high spatiotemporal resolution, realizing continuous monitoring and high-precision characterization of the turbulence structure in the atmospheric boundary layer, and meeting the business system's requirements for the stability, accuracy and deployability of the boundary layer turbulence inversion products.

[0008] (II) Technical solution In order to achieve the purpose of the invention and solve the technical problems, the present invention adopts the following technical solutions: The first invention object of the present invention is to provide an atmospheric boundary layer turbulence statistics inversion method based on a 6-beam ground-based photon wind laser radar DBS scanning mode, which is used to obtain vertical profiles of turbulence statistics such as velocity variance, momentum flux and turbulent kinetic energy in the atmospheric boundary layer. The inversion method comprises at least the following steps when implemented: S100. 6-beam DBS scanning mode configuration: The ground-based photon wind lidar is controlled to observe in Doppler beam swing scanning mode (DBS), with 6 detection beams set, including 1 vertical zenith beam and 5 beams at set elevation angles. f And along the cone, the azimuth angles are equally spaced i Distributed tilt beams; S200. Radial wind speed data collection and processing: Synchronously obtain the radial wind speed time series data in each detection beam direction and record the corresponding azimuth i With elevation f Parameters, and quality control of the acquired radial wind speed data, screen out data points with a signal-to-noise ratio lower than a preset threshold, and remove abnormal data points that deviate significantly from the average value or have mutations; S300. Calculate the radial velocity variance of each beam: The radial wind speed time series data collected in each beam direction v ri ( t ) to perform statistical processing and calculate the radial velocity variance corresponding to each beam , and based on this, the radial velocity variance observation vector is constructed , is the radial velocity variance of the 1st to 6th beams, i =1,2,…,6; S400. Establish the relationship between radial velocity variance and turbulence statistics: According to the elevation angle corresponding to each beam f i and azimuth i i , respectively establish the radial velocity variance of each beam and six turbulence statistics The relationship between: in, The fluctuating wind speed components in the horizontal east-west direction, horizontal north-south direction and vertical direction are are the velocity variances in three directions, are the covariances of the velocity components in three directions, i.e., the momentum flux terms; S500. Construct radial velocity variance observation equations and parameter matrix: Based on the relationship between the radial velocity variance in the six beam directions and the turbulence statistics, a matrix-based linear equation system for radial velocity variance observation is constructed. , where S is the radial velocity variance observation vector, and M is the elevation angle of each beam. f i and azimuth i i The calculated 6×6 coefficient matrix, Σ is the turbulence statistics vector to be solved and ; S600. Turbulence statistics inversion and coordinate rotation: Based on the radial velocity variance observation equations , the inverse matrix method or the least squares fitting method is used to invert and solve the turbulence statistics vector Σ, and the coordinates of the inversion results are rotated based on the average wind direction to ensure that the velocity disturbance variance and covariance are expressed in a coordinate system aligned with the average wind direction; S700. Turbulent Kinetic Energy TKE calculation and result output: Calculate the turbulent kinetic energy based on the inverted velocity variance Finally, the vertical profiles of turbulence statistics including velocity variance, momentum flux and turbulent kinetic energy at each altitude layer are output to characterize the vertical structure of turbulence in the atmospheric boundary layer.

[0009] The second invention object of the present invention is to provide an atmospheric boundary layer turbulence statistics inversion system, which is used to implement the above-mentioned atmospheric boundary layer turbulence statistics inversion method based on the 6-beam ground-based photon wind laser radar DBS scanning mode of the present invention, and the system includes at least the following modules: The laser radar observation module controls the ground-based photon wind laser radar to launch six detection beams in DBS scanning mode, including one vertical zenith beam and five inclined beams evenly distributed along the cone surface; The data acquisition and processing module is used to synchronously obtain the radial wind speed time series data in each detection beam direction, record the corresponding azimuth and elevation parameters, and perform quality control on the radial wind speed data; Statistics calculation and matrix construction module, used to calculate the radial velocity variance of each beam at each altitude layer, and establish a linear equation system between it and turbulence statistics based on the beam geometry angle; The inversion solution and coordinate rotation module is used to solve the turbulence statistics vector using the inverse matrix method or the least squares fitting method, and perform coordinate transformation based on the average wind direction; The result calculation output module is used to calculate the turbulent kinetic energy based on the velocity variance obtained by inversion, and output the vertical profile of turbulence statistics at each altitude layer.

[0010] The third inventive object of the present invention is to provide a computer program product, including computer instructions, for executing the above-mentioned atmospheric boundary layer turbulence statistics inversion method based on the 6-beam ground-based photon wind measurement lidar DBS scanning mode of the present invention.

[0011] The fourth inventive object of the present invention is to provide a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned atmospheric boundary layer turbulence statistics inversion method based on the 6-beam ground-based photon wind measurement lidar DBS scanning mode of the present invention is implemented.

[0012] (III) Technical Effect Compared with the prior art, the atmospheric boundary layer turbulence statistics inversion method, system, medium and program product based on the 6-beam ground-based photon wind laser radar DBS scanning mode of the present invention have the following beneficial and significant technical effects: (1) By introducing a 6-beam DBS scanning mode (1 zenith beam and 5 inclined beams), the present invention can synchronously acquire radial wind speed data at multiple altitude layers in the atmospheric boundary layer, thereby realizing high spatiotemporal resolution observation of turbulence statistics. At the same time, by accurately constructing the mathematical relationship between radial velocity variance and turbulence statistics and optimizing the solution of the matrix equation, the present invention avoids the systematic error caused by the underdetermined equations of the traditional 4-5 beams, and significantly improves the inversion accuracy of turbulence statistics such as velocity variance and momentum flux.

[0013] (2) The present invention constructs an analytical relationship model between radial velocity variance and six turbulence statistics, and uniformly solves them in matrix form, avoiding the uncertainty of empirical estimation item by item, and realizing the systematic inversion of velocity variance and momentum flux. In addition, the present invention adopts a strategy combining quality control and coordinate rotation to establish a stable mapping process from the original observation to the turbulence component in the main wind direction coordinate system, which improves the physical interpretability of the inversion results and is particularly suitable for boundary layer environments with drastic wind direction changes.

[0014] (3) The present invention is based on ground-based quantum wind laser radar technology, which has the advantages of small size, light weight, and easy deployment. It can be widely used in atmospheric boundary layer turbulence observation in different scenarios such as urban environments, mountainous terrains, and ocean areas. At the same time, the present invention can realize continuous vertical profile observation of atmospheric boundary layer turbulence statistics, provide key parameters for atmospheric diffusion models, numerical weather forecasts, and climate models, and has significant scientific value and application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a flow chart of the atmospheric boundary layer turbulence statistics inversion method based on the 6-beam ground-based photon wind laser radar DBS scanning mode of the present invention; Figure 2 It is a schematic diagram of the atmospheric boundary layer turbulence statistics inversion system of the present invention; Figure 3 The time-height profiles of various turbulence statistics obtained by inverting the six-beam detection mode in Hefei area from 11-16 00:00 to 11-21 23:30 Beijing time in 2023, where: (a) u Directional velocity variance , (b) is v Directional velocity variance , (c) is the vertical velocity variance , (d) is the turbulent kinetic energy , pink and cyan are the boundary layer heights obtained from wind lidar and ERA5 reanalysis, respectively.

[0016] Description of reference numerals: LiDAR observation module 100, data acquisition and processing module 200, statistics calculation and matrix construction module 300, inversion solution and coordinate rotation module 400, result calculation and output module 500. DETAILED DESCRIPTION

[0017] The present invention aims to provide a method, system, medium and program product for inverting turbulence statistics of the atmospheric boundary layer based on the DBS scanning mode of a 6-beam ground-based photon wind laser radar, which is used to obtain the vertical profile of turbulence statistics such as velocity variance, momentum flux and turbulent kinetic energy in the atmospheric boundary layer. In order to make the purpose, technical scheme and advantages of the implementation of the present invention clearer, the technical scheme in the embodiment of the present invention will be described in more detail below in conjunction with the drawings in the embodiment of the present invention. The described embodiments are part of the embodiments of the present invention, not all of the embodiments, and the described embodiments are exemplary and are intended to be used to explain the present invention, but cannot be understood as limiting the present invention.

[0018] Example 1: Inversion method As a specific example, the embodiment of the present invention provides an atmospheric boundary layer turbulence statistics inversion method based on a 6-beam ground-based photon wind laser radar DBS scanning mode, such as Figure 1 As shown, the implementation mainly includes the following steps: S100. 6-beam DBS scanning mode configuration: The ground-based photon wind lidar is controlled to observe in Doppler beam swing scanning mode (DBS), with 6 detection beams set, including 1 vertical zenith beam and 5 beams at set elevation angles. f And along the cone, the azimuth angles are equally spaced i Distributed tilted beams.

[0019] In the embodiment of the present invention, five tilted beams are evenly distributed along the azimuth angle with the zenith beam as the center on the cone surface formed by the vertical beam, and the azimuth angle of each tilted beam is i i The laser beams are arranged at intervals of 72° starting from the true north direction to ensure the symmetry of the laser beam on the horizontal plane. That is, the azimuth angles are set to i 1 =0°, i 2 =72°, i 3 =144°, i 4 =216°, i 5 =288°, the elevation angles of the tilted beams are the same and are set in the range of 45°~75° to ensure that the detection altitude covers the main range of the atmospheric boundary layer, and the six beams are emitted in sequence at fixed time intervals to achieve omnidirectional coverage of typical-scale turbulence disturbance structures.

[0020] S200. Radial wind speed data collection and processing: Synchronously obtain the radial wind speed time series data in each detection beam direction and record the corresponding azimuth i With elevation f Parameters are set, and quality control is performed on the acquired radial wind speed data to screen out data points with a signal-to-noise ratio lower than a preset threshold, and to remove abnormal data points that deviate significantly from the average value or have mutations.

[0021] Preferably, the sampling period of the radial wind speed time series data traversing 6 beams is 20-30 seconds, and the spatial resolution is set to 15-30 meters to ensure the ability to capture the mesoscale and small-scale turbulent disturbance processes in the atmospheric boundary layer.

[0022] In addition, in step S200, radial wind speed data quality control preferably includes the following sub-steps: S201. Preprocessing the original radial wind speed data to remove invalid data or erroneous data caused by low detection echo intensity, disordered timestamp, oversaturated echo or abnormal system return code; S202. calculating the signal-to-noise ratio (CNR) of each detection point based on the characteristic parameters of the laser pulse echo signal, and filtering out data points whose signal-to-noise ratio is lower than a preset threshold (preferably between -18 dB and -20 dB) and whose signal strength is insufficient to support reliable velocity estimation; S203. Apply the sliding window statistical analysis method to the retained data to calculate the local mean and standard deviation, and use the Z-score method or the empirical upper and lower limit method to identify sharp jump points, extreme deviation points or data glitches, and remove data points that do not meet the characteristics of stable disturbances; S204. Perform linear interpolation or time reconstruction on the data that are partially missing due to screening to ensure that all 6 beams have a consistent sampling time reference within the same observation period.

[0023] S300. Calculate the radial velocity variance of each beam: The radial wind speed time series data collected in each beam direction v ri ( t ) to perform statistical processing and calculate the radial velocity variance corresponding to each beam , and based on this, the radial velocity variance observation vector is constructed is the radial velocity variance of the 1st to 6th beams, i =1,2,…,6.

[0024] In the embodiment of the present invention, the calculation of the radial velocity variance of each beam includes the following sub-steps: S301. For each beam direction i The altitude layer corresponding to each distance k , for radial wind speed time series data Calculate its mean within the time window T ,in i =1,2,…,6, k =1,2,…, N , N It is the effective number of height layers that can be covered in the vertical direction; S302. Based on the calculated mean Solve the radial velocity sample variance within the time window T ,in, n is the effective sample number, t j is the sampling time point; S303. For each beam direction and each altitude layer, a multi-altitude and multi-directional radial velocity variance matrix is ​​formed as the basic input data for the subsequent turbulence statistics inversion. If necessary, linear interpolation is performed for altitude layers with data gaps to ensure the structural continuity and calculation stability of the vertical profile.

[0025] S400. Establish the relationship between radial velocity variance and turbulence statistics: According to the elevation angle corresponding to each beam f i and azimuth i i , respectively establish the radial velocity variance of each beam and six turbulence statistics The relationship between: in, The fluctuating wind speed components in the horizontal east-west direction, horizontal north-south direction and vertical direction are are the velocity variances in three directions, are the covariances of the velocity components in three directions, namely, the momentum flux terms.

[0026] It should be pointed out that when establishing the relationship between radial velocity variance and turbulence statistics, based on the assumption of uniformity and isotropy of atmospheric turbulence (i.e., assuming that in the atmospheric boundary layer, the statistical characteristics of turbulence are uniformly distributed in space and the same in all directions), the radial velocity variance is expressed as six turbulence statistics (including velocity variance in three directions and the momentum flux in three directions ) is a linear combination of the beams; the coefficients of the linear combination are determined by the elevation angles of the beams. f i and azimuth i i It reflects the contribution of wind speed components in different directions to the radial velocity variance.

[0027] S500. Construct radial velocity variance observation equations and parameter matrix: Based on the relationship between the radial velocity variance in the six beam directions and the turbulence statistics, a matrix-based linear equation system for radial velocity variance observation is constructed. , where S is the radial velocity variance observation vector, and M is the elevation angle of each beam. f i and azimuth i i The calculated 6×6 coefficient matrix, Σ is the turbulence statistics vector to be solved and .

[0028] It should be pointed out that when determining the coefficient matrix M of the observation equation group, the elevation angle of each beam is f i and azimuth i i , calculate the projection coefficients of each beam in the three coordinate axis directions, and arrange these projection coefficients in a certain order to form a coefficient matrix M, whose dimension is 6×6, each row corresponds to a beam, and each column corresponds to a turbulence statistic, such as i The 6 matrix elements corresponding to the rows are The coefficient matrix reflects the sensitivity of each beam to different turbulence statistics, and its value directly affects the accuracy of turbulence statistics inversion. In actual calculations, it is necessary to ensure that the elevation angle of each beam f iand azimuth i i The measurement accuracy is improved to avoid the influence of the calculation error of the coefficient matrix M on the inversion result.

[0029] S600. Turbulence statistics inversion and coordinate rotation: Based on the radial velocity variance observation equations , the inverse matrix method or the least squares fitting method is used to invert and solve the turbulence statistics vector Σ, and the coordinates of the inversion results are rotated based on the mean wind direction to ensure that the velocity disturbance variance and covariance are expressed in a coordinate system aligned with the mean wind direction.

[0030] Preferably, the coordinate rotation may adopt a double coordinate rotation method, comprising the following sub-steps: S621. Calculate the horizontal average wind speed direction for each altitude layer based on the average wind speed component obtained by inversion ,in and are the east-west and north-south components of the average wind speed, respectively. Based on this, the first rotation matrix from the radar local coordinate system to the average wind direction coordinate system is constructed to eliminate the projection effect of the velocity distribution in the non-principal axis direction. S622. Based on the covariance term of the vertical velocity component and the horizontal wind speed component (such as ) The second rotation matrix is ​​constructed to completely align the vertical direction with the ground normal direction, project the turbulence disturbance tensor onto the principal axis system in the sense of atmospheric dynamics, achieve physically consistent expression of the turbulent flux terms, and eliminate the cross-covariance components of the unphysical solution.

[0031] In addition, when the inverse matrix method is used to invert and solve the turbulence statistics vector Σ, the following sub-steps are included: S611. Determine whether the coefficient matrix M of the observation equation group is invertible. If M is not invertible, use the singular value decomposition (SVD) method for processing; S612. Calculate the inverse matrix M of the coefficient matrix M -1 , then the radial velocity variance observation vector S and the inverse matrix M -1 Multiplying them together gives the solution to the turbulence statistics vector Σ.

[0032] It should be pointed out that when the least squares fitting method is used to invert and solve the turbulence statistic vector Σ, it is necessary to first construct an objective function, which represents the residual sum of squares between the radial velocity variance observation value and the model calculated value; then an optimization algorithm (such as the gradient descent method or the Levenberg-Marquardt algorithm) is used to minimize the objective function to obtain the optimal solution of the turbulence statistic vector Σ.

[0033] Furthermore, in step S600, an ill-conditioned matrix analysis and regularization processing method is used to solve the ill-conditioned problem that may exist in the coefficient matrix M, including: Calculate the condition number of the coefficient matrix M to determine the degree of ill-conditioning of the observation equation system; When the condition number of the coefficient matrix M exceeds the preset threshold, the regularization method is used to solve the modified system of equations ,in l is the regularization parameter, I is the identity matrix; Determine the optimal regularization parameter by L-curve method or generalized cross-validation method l ; The turbulence statistics Σ are solved based on the modified set of equations, and the uncertainty analysis of the inversion results is performed using the Monte Carlo simulation method to output the confidence interval of each turbulence statistic.

[0034] S700. Turbulent Kinetic Energy TKE calculation and result output: Calculate the turbulent kinetic energy based on the inverted velocity variance Finally, the vertical profiles of turbulence statistics including velocity variance, momentum flux and turbulent kinetic energy at each altitude layer are output to characterize the vertical structure of turbulent disturbances in the atmospheric boundary layer.

[0035] S800. Verify the physical validity of turbulence parameters, including: Verify the physical rationality of the inversion results based on the constraints of atmospheric turbulence theory, including checking the non-negativity of velocity variance ; Verify that the covariance satisfies the Cauchy-Schwarz inequality , , ; Calculate the turbulence anisotropy index And verify whether its variation range conforms to the turbulence characteristics of the atmospheric boundary layer; check whether the sign and vertical variation trend of the momentum flux under different atmospheric stability conditions conform to the physical laws, and mark or correct the results that do not meet the constraints.

[0036] S900. Error evaluation of the inverted turbulence statistics, including: S901. Multi-source comparison error assessment: Compare and verify the turbulence statistics obtained by inversion at each altitude with other existing observation data (such as temperature and wind field profiles provided by the sounding system, turbulent kinetic energy and momentum flux measured by the boundary layer tower mast ultrasonic anemometer, and high-resolution wind speed disturbance obtained by the airborne platform, etc.), calculate the deviation, root mean square error and / or correlation coefficient of the inversion results, and evaluate the physical consistency and observation credibility of the inversion results in the actual atmospheric background; S902. Error source identification: Identify the main sources of inversion errors, including the accuracy limitation of the lidar system, fluctuations in the signal-to-noise ratio of the echo signal, insufficient observation time window, statistical deviations caused by the non-steady state of boundary layer turbulence, and / or beam angle projection errors. In the identification process, the potential error dominant factors are marked for different altitude layers and different beam directions, providing a basis for subsequent propagation modeling and weighted processing; S903. Propagation Error Modeling and Quantitative Estimation: Matrix Structure Based on Inversion Algorithm Quantitative analysis of the inversion error of turbulence statistics and calculation of the covariance matrix of turbulence statistics , obtain the estimated error variance and covariance of each inversion component and the cross-correlation uncertainty between turbulence statistics, C S is the uncertainty covariance matrix of radial velocity variance observation; S904. Evaluation output and layered diagnosis: Output the relative error percentage and confidence interval boundaries of each turbulence statistical component in each altitude layer, identify the noise-dominated area and high-confidence section, and visualize the inversion results and error envelope for subsequent quality control and data release reference.

[0037] In summary, Example 1 elaborates in detail the complete technical process of the atmospheric boundary layer turbulence statistics inversion method based on the 6-beam ground-based photon wind measurement lidar DBS scanning mode of the present invention, from observation configuration, data processing to parameter inversion and error evaluation, and finally obtains the vertical profile of turbulence statistics with high temporal and spatial resolution, reflecting the comprehensive advantages of the present invention in terms of accuracy, stability and practicality.

[0038] Example 2: Inversion system Based on the atmospheric boundary layer turbulence statistics inversion method based on the 6-beam ground-based photon wind laser radar DBS scanning mode proposed in Example 1, this Example 2 further provides an inversion system structure that can be actually deployed and operated. The system combines the six-beam DBS scanning observation capability of the ground-based photon wind laser radar with a multi-module collaborative computing architecture, and can efficiently complete the extraction and output of turbulence statistics. Specifically, Figure 2 As shown. The system decomposes the complex inversion process into several independent modules with clear functions through modular design, which is convenient for system development, maintenance and upgrade. At the same time, standardized interfaces are used for data exchange between modules to ensure the flexibility and scalability of the system. The system mainly includes the following modules: The laser radar observation module 100 is used to control the ground-based photon wind laser radar to emit 6 detection beams in DBS scanning mode, including 1 vertical zenith beam and 5 tilted beams evenly distributed along the cone. In terms of hardware implementation, the module includes a laser emission submodule, a beam control submodule and a timing synchronization submodule. The laser emission submodule adopts the photon detection technology in the 1.5μm band, and generates laser pulses that meet the needs of atmospheric detection through a fiber amplifier; the beam control submodule realizes the rapid switching of six beams through a high-precision galvanometer system, in which the elevation angle of the tilted beam can be dynamically set in the range of 45°-75°, and the azimuth angle is evenly distributed at 72° intervals to form a complete ring cone structure; the timing synchronization submodule ensures that the emission and reception of each beam strictly follow the preset time sequence, and at the same time keeps the time reference synchronized with the subsequent data processing module. This module uses a preset scanning program and lidar control system to transmit six beams in sequence at fixed time intervals, and can adjust parameters such as scanning period and beam elevation according to different meteorological conditions or application requirements to meet the observation needs of multi-scale turbulent disturbances in the atmospheric boundary layer, providing stable, continuous, and well-covered original wind field information for subsequent data processing and parameter inversion.

[0039] The data acquisition and processing module 200 is used to synchronously acquire the radial wind speed time series data in each detection beam direction, record the corresponding azimuth and elevation parameters, and perform quality control on the radial wind speed data. In terms of hardware implementation, the module includes a multi-channel data acquisition card, a signal processing submodule, and a quality control submodule. The data acquisition card synchronously records the echo signals in the six beam directions at a sampling rate of not less than 1MHz. The signal processing submodule extracts the radial wind speed value of each range gate (15-30m resolution) through fast Fourier transform FFT, and calculates quality indicators such as carrier-to-noise ratio CNR. The quality control submodule first performs system preprocessing on the raw data, including abnormal code recognition, time alignment check, and range gate channel screening; then calculates the signal-to-noise ratio CNR based on the laser echo signal, and removes unreliable sample points below the preset threshold (–18 dB to –20 dB); further, the sliding window method and Z-score method are used to identify local wind speed mutation points and deviation abnormal values ​​to ensure that only data points that meet stable statistical characteristics are retained. In addition, to solve the data discontinuity problem caused by quality control, the module also supports linear interpolation or timing reconstruction of the time series to ensure that the sampling time base of all beams is consistent.

[0040] The statistics calculation and matrix construction module 300 is used to calculate the radial velocity variance of each beam at each altitude layer, and establish a linear equation system between it and the turbulence statistics based on the beam geometry angle. This module corresponds to the calculation process of steps S300-S500 in Example 1, and is composed of three submodules working together: The variance calculation submodule performs statistical analysis on the wind speed time series (30-60s window) of each beam and each altitude layer to calculate the radial velocity variance; the matrix construction submodule calculates the radial velocity variance based on the preset beam geometry parameters ( f i , i i ), automatically generates a 6×6 coefficient matrix M according to the trigonometric function relationship given in Example 1; an equation group assembly submodule, combines the observation vector S and the coefficient matrix M into a complete linear equation group S=M·Σ. To improve computing efficiency, the module adopts a parallel computing architecture to process data from multiple altitude layers simultaneously, and uses memory mapping technology to achieve fast access to large amounts of data. The module also has a cache mechanism inside. When some beam data is missing, it can be intelligently filled based on historical data to ensure the continuity of the solution of the equation group. In addition, the module also has the function of interpolating and completing the missing altitude layers to ensure the continuity of the vertical profile, and supports batch construction of multi-time equation group structures to achieve automatic preprocessing of large amounts of data under the sliding window.

[0041] The inversion solution and coordinate rotation module 400 is used to solve the turbulence statistics vector using the inverse matrix method or the least squares fitting method, and perform coordinate transformation based on the average wind direction. This module is the core computing unit of the system, and the solver module provides two algorithm options: for well-conditioned equations, the LU decomposition method is directly used for inversion; for pathological cases (condition number > 100), the Tikhonov regularization algorithm is enabled, and the optimal regularization parameter is automatically determined by the L-curve method. The coordinate rotation submodule uses the quaternion method to realize the three-dimensional coordinate system transformation: first, the first rotation angle is calculated based on the average horizontal wind to eliminate the influence of wind direction; then the second rotation angle is calculated based on the vertical flux to align the coordinate system with the main axis of the average wind field. This module also includes a result verification submodule to check whether the turbulence statistics obtained by inversion meet the requirements. Physical constraints such as the above can automatically trigger the recalculation process for abnormal results.

[0042] The result calculation and output module 500 calculates the turbulent kinetic energy based on the velocity variance obtained by inversion, and outputs the vertical profile of the turbulent statistics at each altitude layer. The turbulent kinetic energy calculation submodule is responsible for calculating the turbulent kinetic energy value at each altitude layer; the boundary layer height detection submodule automatically identifies the position of the boundary layer top through the vertical velocity variance threshold method (w′²<0.3 m² / s²); the data output submodule generates standard data products including velocity variance, momentum flux, turbulent kinetic energy and boundary layer height, and supports multiple scientific data formats such as NetCDF and HDF5. In addition, it also includes a quality control submodule to test the spatial consistency and temporal continuity of the inversion results, mark suspicious data points, and generate metadata information containing error estimates.

[0043] Through the coordinated operation of the above modules, the system can completely and efficiently implement all key links of the inversion method described in Example 1, and has good adaptability, scalability and engineering deployment capabilities.

[0044] Example 3: Example verification In order to verify the practicality and reliability of the atmospheric boundary layer turbulence statistics inversion method based on the 6-beam ground-based photon wind laser radar DBS scanning mode proposed in the present invention, a radar actual observation test was carried out for six consecutive days from 00:00 on November 16, 2023 to 23:30 on November 21 (Beijing time) in Hefei, Anhui Province as the experimental site. The observation equipment uses a ground-based photon wind laser radar with 6-beam DBS scanning capability. The beam configuration is consistent with that described in Example 1, the data sampling period is 1 second, the sliding time window is 60 seconds, the vertical resolution is set to 30 meters, and the coverage height is up to 2.7 km. After data collection, the radial wind speed quality control, velocity variance calculation, turbulence statistics matrix inversion, coordinate rotation and turbulence kinetic energy solution are completed in accordance with the technical process of Example 1.

[0045] Figure 3 The time-height profile distribution of typical turbulence statistics obtained by the inversion method of the present invention during the above observation period is shown. Figures (a)-(d) represent the east-west velocity variance, respectively. , north-south velocity variance , vertical velocity variance and turbulent kinetic energy The time-height distribution characteristics of the surface of the Earth are shown in Figure 2. The pink and cyan curves represent the boundary layer heights determined based on wind lidar and ERA5 reanalysis data, respectively. Z - z i As can be seen from the figure, the four turbulence statistics show good spatiotemporal structural characteristics in different daily variation cycles, and show a highly coupled correlation with the boundary layer height change. Especially in the daytime convection development stage, as well as The three-component velocity variances are significantly enhanced in the range of 500-1500m altitude, which can clearly depict the strong turbulence activities in the typical convective boundary layer. When the stable layer appears at night, the three-component velocity variances weaken rapidly and show the characteristics of concentrated distribution close to the ground, reflecting the sensitivity of this method to capturing the turbulence suppression process in the boundary layer. In addition, Z - z i With ERA5- z iThe daily variation trends of ERA5- z i There is an obvious hysteresis phenomenon. Z - z i The local turbulence enhancement process can still be identified, indicating that the inversion mechanism of the present invention has a higher time-effective response capability and vertical structure resolution capability. In terms of boundary layer height estimation, the dual-criteria method adopted by the present invention, namely, the vertical velocity variance The first time it is less than 0.3 m² / s² or the signal-to-noise ratio CNR is lower than 0.4, which is used as the boundary layer top criterion. Figure 3 Judging from the continuity and boundary clarity of the method, it has strong stability and discrimination ability in practical applications, and can be used for automatic extraction of boundary layer height and real-time publishing of turbulence products.

[0046] In summary, the measured analysis results of Example 3 fully demonstrate that the inversion method proposed in the present invention can not only accurately reflect the three-dimensional structural evolution process of turbulent disturbances, but also has good boundary layer structure recognition ability and reanalysis data comparison consistency, and has strong engineering adaptability and scientific research promotion value.

[0047] Through the above embodiments, the purpose of the present invention is fully and effectively achieved. Those skilled in the art can understand that the present invention includes but is not limited to the contents described in the drawings and the above specific embodiments. Although the present invention has been described with respect to the most practical and preferred embodiments currently considered, it should be understood that the present invention is not limited to the disclosed embodiments, and any modification that does not deviate from the functional and structural principles of the present invention will be included in the scope of the claims.

Claims

1. A method for inverting atmospheric boundary layer turbulence statistics based on a 6-beam ground-based photon wind laser radar DBS scanning mode, characterized in that: include: S100. Setting the ground-based photon wind laser radar to DBS scanning mode, and configuring one vertically upward zenith beam and five tilted beams distributed at set elevation angles and equally spaced in azimuth; S200. Synchronously obtain radial wind speed time series data in each beam direction and record the corresponding azimuth θ With elevation φ parameters and perform quality control on wind speed data; S300. Statistic the radial wind speed data of each beam and calculate the radial speed variance , i =1,2,…,6, and construct the radial velocity variance observation vector based on it ,in , , …, is the radial velocity variance of the 1st to 6th beams; S400. According to the elevation angle corresponding to each beam φ i and azimuth θ i , respectively establish the radial velocity variance and six turbulence statistics The relationship between: in, are the fluctuating wind speed components in the east-west, north-south and vertical directions, and are the velocity variance and covariance in three directions respectively; S500. Constructing radial velocity variance observation equations , where S is the radial velocity variance observation vector, and M is the elevation angle of each beam. φ i and azimuth θ i The calculated 6×6 coefficient matrix, Σ is the turbulence statistics vector to be solved and ; S600. Based on the radial velocity variance observation equations , the inverse matrix method or the least squares fitting method is used to invert and solve the turbulence statistics vector Σ, and the coordinates of the inversion results are rotated based on the average wind direction to ensure that the velocity disturbance variance and covariance are expressed in a coordinate system aligned with the average wind direction; S700. Calculate turbulent kinetic energy based on the inverted velocity variance , and finally outputs the vertical profiles of velocity variance, momentum flux and turbulent kinetic energy at each altitude layer, which are used to characterize the vertical structure of turbulence in the atmospheric boundary layer.

2. The method for inverting atmospheric boundary layer turbulence statistics based on the 6-beam ground-based photon wind laser radar DBS scanning mode according to claim 1 is characterized in that: In step S100, the five tilted beams are evenly distributed along the azimuth angle on the cone formed by the vertical beam with the zenith beam as the center. The azimuth angle of each tilted beam is θ i The six beams are arranged at intervals of 72° starting from the north direction to ensure the symmetry of the laser beam on the horizontal plane. The elevation angles of the inclined beams are the same and are set in the range of 45° to 75° to ensure that the detection altitude covers the main range of the atmospheric boundary layer. The six beams are emitted in sequence at fixed time intervals to achieve omnidirectional coverage of typical-scale turbulent disturbance structures.

3. The method for inverting atmospheric boundary layer turbulence statistics based on 6-beam ground-based photon wind laser radar DBS scanning mode according to claim 1 is characterized in that: In step S200, the sampling period of the radial wind speed of the six beams is 20 to 30 seconds, and the spatial resolution is set to 15 to 30 meters.

4. The method for inverting atmospheric boundary layer turbulence statistics based on 6-beam ground-based photon wind laser radar DBS scanning mode according to claim 1 or 3, characterized in that: In step S200, radial wind speed data quality control includes the following sub-steps: S201. Preprocessing the original radial wind speed data to remove invalid data or erroneous data caused by low detection echo intensity, disordered timestamp, oversaturated echo or abnormal system return code; S202. Calculating the signal-to-noise ratio CNR of each detection point based on the characteristic parameters of the laser pulse echo signal, and filtering out data points whose signal-to-noise ratio is lower than a preset threshold and whose signal strength is insufficient to support reliable velocity estimation; S203. Apply the sliding window statistical analysis method to the retained data to calculate the local mean and standard deviation, and use the Z-score method or the empirical upper and lower limit method to identify sharp jump points, extreme deviation points or data glitches, and remove data points that do not meet the characteristics of stable disturbances; S204. Perform linear interpolation or time reconstruction on the data that are partially missing due to screening to ensure that all 6 beams have a consistent sampling time reference within the same observation period.

5. The method for inverting atmospheric boundary layer turbulence statistics based on 6-beam ground-based photon wind laser radar DBS scanning mode according to claim 1 is characterized in that: In step S300, the calculation of the radial velocity variance of each beam includes the following sub-steps: S301. For each beam direction i The altitude layer corresponding to each distance k , for radial wind speed time series data Calculate its mean within the time window T ,in i =1,2,…,6, k =1,2,…, N , N It is the effective number of height layers that can be covered in the vertical direction; S302. Based on the calculated mean Solve the radial velocity sample variance within the time window T ,in, n is the effective sample number, t j is the sampling time point; S303. For each beam direction and each altitude layer, a multi-altitude and multi-directional radial velocity variance matrix is ​​formed as the basic input data for the subsequent turbulence statistics inversion.

6. The method for inverting atmospheric boundary layer turbulence statistics based on 6-beam ground-based photon wind laser radar DBS scanning mode according to claim 1 is characterized in that: In step S600, the coordinate rotation adopts a double coordinate rotation method, which includes the following sub-steps: S621. Calculate the horizontal average wind speed direction for each altitude layer based on the average wind speed component obtained by inversion ,in and are the east-west and north-south components of the average wind speed, respectively. Based on this, the first rotation matrix from the radar local coordinate system to the average wind direction coordinate system is constructed to eliminate the projection effect of the velocity distribution in the non-principal axis direction. S622. Based on the covariance terms of the vertical velocity component and the horizontal wind speed component, the second rotation matrix is ​​constructed to completely align the vertical direction with the ground normal direction, project the turbulent disturbance tensor onto the principal axis system in the sense of atmospheric dynamics, achieve physically consistent expression of the turbulent flux terms, and eliminate the cross-covariance components of the non-physical solution.

7. The method for inverting atmospheric boundary layer turbulence statistics based on 6-beam ground-based photon wind laser radar DBS scanning mode according to claim 1 is characterized in that: In step S600, when the inverse matrix method is used to invert and solve the turbulence statistics vector Σ, the following sub-steps are included: S611. Determine whether the coefficient matrix M of the observation equation group is reversible. If M is not reversible, use the singular value decomposition method for processing; S612. Calculate the inverse matrix M of the coefficient matrix M -1 , then the radial velocity variance observation vector S and the inverse matrix M -1 Multiplying them together gives the solution to the turbulence statistics vector Σ.

8. The method for inverting atmospheric boundary layer turbulence statistics based on 6-beam ground-based photon wind laser radar DBS scanning mode according to claim 1 is characterized in that: In step S600, an ill-conditioned matrix analysis and regularization processing method is used to solve the possible ill-conditioned problem of the coefficient matrix M: Calculate the condition number of the coefficient matrix M to determine the degree of ill-conditioning of the observation equation system; When the condition number of the coefficient matrix M exceeds the preset threshold, the regularization method is used to solve the modified system of equations ,in λ is the regularization parameter, I is the identity matrix; Determine the optimal regularization parameter by L-curve method or generalized cross-validation method λ ; The turbulence statistics Σ are solved based on the modified set of equations, and the uncertainty analysis of the inversion results is performed using the Monte Carlo simulation method to output the confidence interval of each turbulence statistic.

9. The method for inverting atmospheric boundary layer turbulence statistics based on 6-beam ground-based photon wind laser radar DBS scanning mode according to claim 1 is characterized in that: The step S800 is also included, in which the physical validity of the turbulence parameters is tested, including the following sub-steps: S801. Verify the physical rationality of the inversion results based on the constraints of atmospheric turbulence theory, including checking the non-negativity of velocity variance ; S802. Verify that the covariance satisfies the Cauchy-Schwarz inequality ; S803. Calculation of turbulence anisotropy index And verify whether its variation range is consistent with the turbulence characteristics of the atmospheric boundary layer; S804. Check whether the sign and vertical variation trend of momentum flux under different atmospheric stability conditions conform to physical laws, and mark or correct the results that do not meet the constraints.

10. The method for inverting atmospheric boundary layer turbulence statistics based on 6-beam ground-based photon wind laser radar DBS scanning mode according to claim 9 is characterized in that: The method further includes step S900 of performing error evaluation on the turbulence statistics obtained by inversion, wherein the error evaluation includes the following sub-steps: S901. Compare and verify the turbulence statistics obtained by inversion at each altitude with other existing observation data, calculate the deviation, root mean square error and / or correlation coefficient of the inversion results, and evaluate the physical consistency and observation credibility of the inversion results in the actual atmospheric background; S902. Identify the main sources of inversion errors, including laser radar system accuracy limitations, echo signal-to-noise ratio fluctuations, insufficient observation time windows, statistical biases caused by non-steady-state properties of boundary layer turbulence, and / or beam angle projection errors; S903. Matrix structure based on inversion algorithm , quantitatively analyze the inversion error of turbulence statistics and calculate the covariance matrix of turbulence statistics , obtain the estimated error variance and covariance of each inversion component and the cross-correlation uncertainty between turbulence statistics, where C S is the uncertainty covariance matrix of radial velocity variance observation; S904. Output the relative error percentage and confidence interval boundaries of each turbulence statistical component in each altitude layer, identify the noise-dominated area and high-confidence section, and visualize the inversion results and error envelope.

11. An atmospheric boundary layer turbulence statistics inversion system, used to implement the atmospheric boundary layer turbulence statistics inversion method based on 6-beam ground-based photon wind laser radar DBS scanning mode as described in any one of claims 1 to 10, characterized in that: include: The laser radar observation module controls the ground-based photon wind laser radar to launch six detection beams in DBS scanning mode, including one vertical zenith beam and five inclined beams evenly distributed along the cone surface; The data acquisition and processing module is used to synchronously obtain the radial wind speed time series data in each detection beam direction, record the corresponding azimuth and elevation parameters, and perform quality control on the radial wind speed data; Statistics calculation and matrix construction module, used to calculate the radial velocity variance of each beam at each altitude layer, and establish a linear equation system between it and turbulence statistics based on the beam geometry angle; The inversion solution and coordinate rotation module is used to solve the turbulence statistics vector using the inverse matrix method or the least squares fitting method, and perform coordinate transformation based on the average wind direction; The result calculation output module is used to calculate the turbulent kinetic energy based on the velocity variance obtained by inversion, and output the vertical profile of turbulence statistics at each altitude layer.

12. A computer program product comprising computer instructions, characterized in that: Used to execute the atmospheric boundary layer turbulence statistics inversion method based on the 6-beam ground-based photon wind measurement lidar DBS scanning mode as described in any one of claims 1 to 10.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for inverting atmospheric boundary layer turbulence statistics based on the DBS scanning mode of a 6-beam ground-based photon wind laser radar as described in any one of claims 1 to 10 is implemented.

Images