A radar coordination monitoring system for complex terrain

Abstract

The application relates to the technical field of radar image sensing and weather monitoring, and discloses a radar cooperative monitoring system for complex terrains, which comprises the following modules: an echo acquisition module, which is used for acquiring multi-waveband radar volume scanning data; a space-time mapping module, which is used for mapping polar coordinate data to a three-dimensional equidistant grid space; a deviation correction module, which is used for constructing a binary deviation function based on a beam blocking rate derived from an elevation model to correct a reflectivity factor; an index reconstruction module, which is used for reconstructing vertical cumulative liquid water content and echo top height in combination with a nonlinear scaling factor to output a unified early warning index; and a cooperative decision module, which is used for outputting a thunderstorm early warning signal. In the application, the beam blocking rate is embedded into an attenuation correction function as a physical weight, distortion caused by terrain blocking is reduced, and the amplification of errors in the calculation of an operational index is inhibited through a nonlinear scaling factor, so that the consistency of multi-waveband early warning indexes is improved.

Description

Technical Field

[0001] This invention belongs to the field of radar image sensing and weather monitoring technology, and particularly relates to a radar collaborative monitoring system for complex terrain. Background Technology

[0002] Weather radar monitoring networks typically employ a multi-band radar collaborative approach to monitor severe convective weather processes. S-band radars, with their lower electromagnetic wave attenuation and longer detection range, are often used as the backbone of the radar network. C-band radars are primarily used for regional networking. X-band radars, with their high temporal and spatial resolution, are frequently used to supplement low-altitude detection blind spots and acquire more detailed echo images. In mountainous environments, current technologies typically interpolate and map the three-dimensional volume scan echo data acquired by different band radars onto a unified three-dimensional grid space to facilitate collaborative processing of monitoring data from different band radars. However, in situations where severe convective weather develops rapidly and mountainous terrain significantly obstructs data, existing data fusion methods are prone to spatial location deviations and index calculation errors. The rapid movement of echo cells in severe convective weather, coupled with time differences in volume scan sampling by different band radars, can lead to inaccurate alignment of echo positions in three-dimensional space. Commonly used nearest-neighbor interpolation methods do not consider the influence of horizontal wind fields on the movement and morphological changes of echo cells, easily causing ghosting or false structures in overlapping areas of echo images from different radars.

[0003] Existing data interpolation and rasterization methods do not fully consider physical attenuation mechanisms when processing radar base data. For example, Chinese invention patent CN108169751B discloses a three-dimensional rasterization method for weather radar base data, a computer-readable storage medium, and an electronic device. It establishes a raster cube in a rectangular coordinate system and determines the echo intensity of the raster unit based on the geometric relative position of the polar coordinate scanning point in the adjacent azimuth and elevation layers. However, this technical solution relies only on interpolation and value selection based on pure geometric spatial relationships and does not consider the physical attenuation caused by continuous obstruction of the radar beam by mountains under complex terrain conditions. It also does not establish a quantitative mapping relationship between elevation obstruction and attenuation compensation. In addition, the attenuation error of the reflectivity factor will be continuously amplified in the calculation of the seventh-fourth power of the vertical cumulative liquid water content and the vertical integration process, causing nonlinear deviations in the warning indicators calculated by radars of different bands, making it difficult to accurately apply a unified warning threshold. The existing low-level reflectivity factor correction process also lacks the constraint of the upper-level meteorological adiabatic rising envelope, which may cause the upper-level echo top height to rise or fall in a way that is not in line with reality.

[0004] Therefore, how to improve the spatiotemporal alignment accuracy of radar echo data in different bands, compensate for the continuous beam obstruction and attenuation caused by mountainous terrain, and suppress the accumulation and amplification of errors in the calculation of early warning indicators have become technical problems that need to be solved. Summary of the Invention

[0005] To address the problems in the background art, the technical solution of the present invention is as follows: A radar cooperative monitoring system for complex terrain, comprising:

[0006] The echo acquisition module is used to simultaneously acquire three-dimensional volume scan echo data from S-band, C-band, and X-band radars in the monitoring area and output a three-dimensional polar coordinate dataset.

[0007] The spatiotemporal mapping module, connected to the echo acquisition module, is used to interpolate and map the data from the X-band and C-band radars in the three-dimensional polar coordinate dataset to a three-dimensional equidistant grid space, based on the coordinate grid of the S-band radar.

[0008] The deviation correction module, connected to the spatiotemporal mapping module, is used to construct a binary deviation function that varies with radial distance and the benchmark based on the reflectivity factor of the mapped S-band radar, extract digital elevation model data to calculate beam obstruction rate, embed the beam obstruction rate as a physical weight into the binary deviation function, and generate a terrain-weighted reflectivity factor.

[0009] The index reconstruction module, connected to the deviation correction module, is used to calculate the vertical cumulative liquid water content and echo top height of each grid point based on the reflectivity factor corrected by terrain weighting. It uses a nonlinear scaling factor that varies with radial distance and reference to superimpose gain and intercept compensation on the vertical cumulative liquid water content and outputs a unified early warning index.

[0010] The collaborative adjudication module, connected to the indicator reconstruction module, is used to classify thunderstorm intensity levels and output warning signals based on unified early warning indicators.

[0011] Preferably, the deviation correction module calculates the reflectance factor after terrain weighting correction. The formula for its calculation is: ,in, This represents the original reflectivity factor of the X-band radar at the grid points. It is a bivariate deviation function. For beam blocking rate, This is the terrain influence weighting coefficient, with a value ranging from 0.0 to 1.0.

[0012] Preferably, the construction process of the binary deviation function is as follows: taking the reflectivity factor of the S-band radar as the unattenuated reference benchmark, obtaining the radial distance from the grid point to the radar site, calculating the deviation of the reflectivity factor of the X-band radar relative to the unattenuated reference benchmark, and using historical thunderstorm sample sets to fit and determine the deviation compensation model composed of the power function product and the constant intercept.

[0013] Preferably, the process of the deviation correction module to extract the beam obstruction rate is as follows: based on the elevation of the radar site, the antenna elevation angle and the atmospheric refraction model, the elevation of the radar beam centerline and the beam half-power angle spread range in the three-dimensional grid space are determined; the elevation obstruction relationship of the digital elevation model data on the ray path from the radar site to the target grid point is extracted; the proportion of the beam cross-sectional area intercepted by the terrain obstruction surface is calculated to obtain the beam obstruction rate.

[0014] Preferably, the process of constructing the nonlinear scaling factor by the index reconstruction module is as follows: extract the radial distance from the grid point to the radar site, multiply the radial distance by the power term of the reference using the bias adjustment parameter, and add a constant to generate the nonlinear scaling factor; the index reconstruction module multiplies the nonlinear scaling factor by the vertical cumulative liquid water content and adds an intercept correction term to obtain the compensated vertical cumulative liquid water content.

[0015] Preferably, the spatiotemporal mapping module interpolates and maps the X-band and C-band radar data in the three-dimensional polar coordinate dataset to a three-dimensional equidistant grid space with a spatial resolution of 1km×1km×0.5km, and controls the time difference of each band radar volume scan in the echo acquisition module to be within 3 minutes.

[0016] Preferably, the spatiotemporal mapping module is also used to obtain the horizontal wind field Doppler velocity vector of the monitoring area, calculate the spatial displacement of the echo centroid along the horizontal wind field Doppler velocity vector within the sampling time difference based on the time difference of the multi-band radar volume scan sampling, and compensate the coordinates for the reverse displacement in the three-dimensional equidistant grid space.

[0017] Preferably, the collaborative adjudication module includes a cross-layer verification submodule, which is used to apply spatial gradient matching calculation to the vertical cumulative liquid water content and echo top height. When the low-layer reflectivity factor changes numerically after topographic weighting correction, the meteorological thermodynamic adiabatic rising envelope is retrieved as the physical boundary to synchronously correct the high-layer echo top height.

[0018] Preferably, the collaborative adjudication module classifies thunderstorm intensity levels as follows: the output unified warning index is input into a preset two-dimensional threshold matrix with two parameters, and the intensity level of the thunderstorm is determined by cross-matching of the two parameters; when the numerical combination of the vertical cumulative liquid water content and the echo top height in the unified warning index exceeds the critical alarm threshold, a warning signal of the corresponding level is generated.

[0019] Preferably, the deviation correction module is also used to establish a binary deviation compensation model of the C-band radar's reflectivity factor relative to the benchmark, based on the reflectivity factor of the mapped S-band radar, and calculate the beam obstruction rate of the C-band radar at each grid point by combining the extracted digital elevation model data, and apply terrain-weighted correction to the original reflectivity factor of the C-band radar.

[0020] Compared with existing technologies, the radar cooperative monitoring system for complex terrain of the present invention has the following advantages:

[0021] 1. In radar collaborative monitoring of complex terrain, by using the volume scan sampling time difference of different radar bands and combining it with the horizontal wind field Doppler velocity vector to correct the echo position, the spatial misalignment caused by the movement of strong convective echo cells within the sampling interval can be reduced. This avoids ghosting and false structures in the overlapping areas of echo images from different radars, thereby improving the spatial alignment accuracy of multi-band radar data and providing a more accurate spatiotemporal reference for subsequent reflectivity factor correction and early warning index calculation.

[0022] 2. Calculate the continuous beam obstruction rate based on digital elevation model data, antenna elevation angle, and atmospheric refraction, and use the beam obstruction rate for reflectivity factor correction. This can compensate for the actual degree of radar beam obstruction by mountains. Compared with using binary masks or local interpolation, this processing can reduce echo intensity distortion caused by terrain obstruction and attenuation of low-elevation beams, and improve the accuracy of echo data in low-altitude obstruction areas in mountainous regions.

[0023] 3. By using a nonlinear scaling factor that varies with radial distance and reference reflectivity factor to compensate for the gain and intercept of vertical cumulative liquid water content, the cumulative amplification of reflectivity factor attenuation error during power operation and vertical integration can be suppressed, reducing the nonlinear deviation between radar calculation results of different bands. This can improve the consistency between vertical cumulative liquid water content and echo top height, enabling the unified thunderstorm warning threshold to be used more accurately for multi-band radar collaborative monitoring. Attached Figure Description

[0024] Figure 1 This is a flowchart of the radar echo data processing and early warning indicator reconstruction of the present invention;

[0025] Figure 2 This is a diagram of the field calibration and multi-node hardware deployment architecture of this invention. Detailed Implementation

[0026] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0027] A radar cooperative monitoring system for complex terrain includes:

[0028] The echo acquisition module is used to simultaneously acquire three-dimensional volume scan echo data from S-band, C-band, and X-band radars in the monitoring area and output a three-dimensional polar coordinate dataset.

[0029] The spatiotemporal mapping module, connected to the echo acquisition module, is used to interpolate and map the data from the X-band and C-band radars in the three-dimensional polar coordinate dataset to a three-dimensional equidistant grid space, based on the coordinate grid of the S-band radar.

[0030] The deviation correction module, connected to the spatiotemporal mapping module, is used to construct a binary deviation function that varies with radial distance and the benchmark based on the reflectivity factor of the mapped S-band radar, extract digital elevation model data to calculate beam obstruction rate, embed the beam obstruction rate as a physical weight into the binary deviation function, and generate a terrain-weighted reflectivity factor.

[0031] The index reconstruction module, connected to the deviation correction module, is used to calculate the vertical cumulative liquid water content and echo top height of each grid point based on the reflectivity factor corrected by terrain weighting. It uses a nonlinear scaling factor that varies with radial distance and reference to superimpose gain and intercept compensation on the vertical cumulative liquid water content and outputs a unified early warning index.

[0032] The collaborative adjudication module, connected to the indicator reconstruction module, is used to classify thunderstorm intensity levels and output warning signals based on unified early warning indicators.

[0033] Preferably, the deviation correction module calculates the reflectance factor after terrain weighting correction. The formula for its calculation is: ,in, This represents the original reflectivity factor of the X-band radar at the grid points. It is a bivariate deviation function. For beam blocking rate, This is the terrain influence weighting coefficient, with a value ranging from 0.0 to 1.0.

[0034] Preferably, the construction process of the binary deviation function is as follows: taking the reflectivity factor of the S-band radar as the unattenuated reference benchmark, obtaining the radial distance from the grid point to the radar site, calculating the deviation of the reflectivity factor of the X-band radar relative to the unattenuated reference benchmark, and using historical thunderstorm sample sets to fit and determine the deviation compensation model composed of the power function product and the constant intercept.

[0035] Preferably, the process of the deviation correction module to extract the beam obstruction rate is as follows: based on the elevation of the radar site, the antenna elevation angle and the atmospheric refraction model, the elevation of the radar beam centerline and the beam half-power angle spread range in the three-dimensional grid space are determined; the elevation obstruction relationship of the digital elevation model data on the ray path from the radar site to the target grid point is extracted; the proportion of the beam cross-sectional area intercepted by the terrain obstruction surface is calculated to obtain the beam obstruction rate.

[0036] Preferably, the process of constructing the nonlinear scaling factor by the index reconstruction module is as follows: extract the radial distance from the grid point to the radar site, multiply the radial distance by the power term of the reference using the bias adjustment parameter, and add a constant to generate the nonlinear scaling factor; the index reconstruction module multiplies the nonlinear scaling factor by the vertical cumulative liquid water content and adds an intercept correction term to obtain the compensated vertical cumulative liquid water content.

[0037] Preferably, the spatiotemporal mapping module interpolates and maps the X-band and C-band radar data in the three-dimensional polar coordinate dataset to a three-dimensional equidistant grid space with a spatial resolution of 1km×1km×0.5km, and controls the time difference of each band radar volume scan in the echo acquisition module to be within 3 minutes.

[0038] Preferably, the spatiotemporal mapping module is also used to obtain the horizontal wind field Doppler velocity vector of the monitoring area, calculate the spatial displacement of the echo centroid along the horizontal wind field Doppler velocity vector within the sampling time difference based on the time difference of the multi-band radar volume scan sampling, and compensate the coordinates for the reverse displacement in the three-dimensional equidistant grid space.

[0039] Preferably, the collaborative adjudication module includes a cross-layer verification submodule, which is used to apply spatial gradient matching calculation to the vertical cumulative liquid water content and echo top height. When the low-layer reflectivity factor changes numerically after topographic weighting correction, the meteorological thermodynamic adiabatic rising envelope is retrieved as the physical boundary to synchronously correct the high-layer echo top height.

[0040] Preferably, the collaborative adjudication module classifies thunderstorm intensity levels as follows: the output unified warning index is input into a preset two-dimensional threshold matrix with two parameters, and the intensity level of the thunderstorm is determined by cross-matching of the two parameters; when the numerical combination of the vertical cumulative liquid water content and the echo top height in the unified warning index exceeds the critical alarm threshold, a warning signal of the corresponding level is generated.

[0041] Preferably, the deviation correction module is also used to establish a binary deviation compensation model of the C-band radar's reflectivity factor relative to the benchmark, based on the reflectivity factor of the mapped S-band radar, and calculate the beam obstruction rate of the C-band radar at each grid point by combining the extracted digital elevation model data, and apply terrain-weighted correction to the original reflectivity factor of the C-band radar.

[0042] Example 1: The current echo acquisition module simultaneously acquires three-dimensional volume scan echo data from S-band, C-band, and X-band radars within the mountainous severe convective weather monitoring area. It outputs a three-dimensional polar coordinate dataset containing longitude, latitude, altitude, and reflectivity factors. The S-band radar serves as a non-attenuation reference, the C-band radar is used for regional coverage, and the X-band radar is used to fill low-altitude blind spots. The spatiotemporal mapping module uses the S-band radar's coordinate grid as a reference, interpolating and mapping the X-band and C-band radar data to a three-dimensional equidistant grid space with a spatial resolution of 1km × 1km × 0.5km. The volume scan sampling time difference between each radar band is controlled within 3 minutes. The spatiotemporal mapping module acquires the horizontal wind field Doppler velocity vector of the monitoring area, calculates the spatial displacement of the echo centroid along the horizontal wind field Doppler velocity vector based on the sampling time difference, and then compensates for the reverse displacement of the X-band and C-band radar grid coordinates within the three-dimensional equidistant grid space to correct alignment deviations and pseudo-structure ghosting of the echo cells in the spatially overlapping coverage area.

[0043] The deviation correction module maps the reflectivity factor of the S-band radar. Calculate the radial distance using the reference point. X-band radar raw reflectivity factor The deviation relative to this benchmark, and the construction of radial distance. and benchmark Variational bivariate deviation function: ,in, The radial distance from the grid point to the radar site. This is the reference reflectivity factor for the S-band radar at this grid point. , and To obtain the model coefficients determined by fitting a historical thunderstorm sample set, the bias correction module retrieves digital elevation model (DEM) data. Based on the radar site elevation, antenna elevation angle, and atmospheric refraction model, it determines the beam centerline elevation and beam half-power angular spread range of the radar beam within the three-dimensional grid space. The elevation obstruction relationship is extracted from the DEM data along the ray path from the radar site to the target grid point. The proportion of the beam cross-section intercepted by the terrain obstruction surface is calculated, yielding a beam obstruction rate ranging from 0.0 to 1.0. , beam blocking rate The terrain-weighted reflectivity factor is calculated as follows, using a binary bias function embedded as a physical weight: ,in, The deviation value is the output of the binary deviation function. This is the terrain influence weighting coefficient, with a value ranging from 0.0 to 1.0. For the C-band radar, the deviation correction module also uses the reflectivity factor of the mapped S-band radar as the benchmark to establish a binary deviation compensation model of the C-band radar reflectivity factor relative to the benchmark; it calculates the beam obstruction rate of the C-band radar at each grid point by combining the digital elevation model data, and applies terrain-weighted correction to the original reflectivity factor of the C-band radar.

[0044] The index reconstruction module is based on the topographically weighted reflectance factor. The linear reflectivity factor is calculated using the following formula: At the grid height layer thickness is At that time, the cumulative liquid water content of a single layer is calculated according to the following formula: ,in, It is a linear reflectivity factor. For the height layer thickness, To calculate the vertical cumulative liquid water content based on the topographically weighted reflectivity factor, without gain and intercept compensation, the index reconstruction module extracts the radial distance. Using the bias adjustment parameter to adjust the radial distance Compared with the benchmark Multiplying the powers of the terms and adding the constant 1, we obtain the nonlinear scaling factor: ,in, and As the bias adjustment parameter, the nonlinear scaling factor is combined with Multiply and add the intercept correction term. The compensated vertical cumulative liquid water content is obtained as follows: ,in, For the intercept correction term, The vertical cumulative liquid water content after compensation.

[0045] In mountainous thunderstorm monitoring, data from key grid points were selected for numerical calculations, with the radial distance from each grid point to the radar station site set as an example. km, measured reflectivity factor of S-band radar dBZ, the original reflectivity factor of the X-band radar to be corrected dBZ, beam obstruction rate calculated by digital elevation model Terrain influence weighting coefficient Height layer thickness km; the deviation model coefficients are respectively , and The parameters of the nonlinear scaling factor are respectively , and kg / m², substituting the above parameters into the bivariate deviation function, we get: The bias value output by the binary bias function is 2.9132 dBZ, and the terrain-weighted gain term is: Thus, the total deviation correction is obtained: The total bias correction is 3.3502 dBZ, and the topographically weighted reflectance factor is: Right now The value is 38.3502 dBZ. Converting it to a linear reflectance factor, we get: Linear reflectivity factor It is 6839.43mm 6 / m³, the vertical cumulative liquid water content of this height layer is: Right now for kg / m², nonlinear scaling factor is: Gain and intercept compensation were performed based on the topographically weighted vertical cumulative liquid water content, resulting in: Compensated vertical cumulative liquid water content for kg / m², from the uncorrected reflectivity factor dBZ, the calculated original vertical cumulative liquid water content for kg / m²; During the parameter calibration process of the nonlinear scaling factor, this original value is substituted into the scaling relationship, and the calculation result is: This result is compared with the reflectivity factor calculated after terrain weighting correction. The values ​​are close, and are used to verify the calibration results of the nonlinear scaling factor and intercept correction term.

[0046] The collaborative adjudication module includes a cross-layer verification submodule. This submodule calculates the spatial difference between the vertical cumulative liquid water content and the echo top height between adjacent grid points, compares the gradient directions of the two and the grids where the gradient changes occur, and completes the spatial gradient matching calculation. When the lower-level reflectivity factor changes due to terrain weighting correction, the cross-layer verification submodule retrieves the meteorological thermodynamic adiabatic ascending envelope calculated from the radiosonde data of the monitoring area. It obtains the maximum buoyancy height that the lower-level thermodynamic ascending airflow can support from this envelope and uses it as the physical boundary of the upper-level echo top height. When the current echo top height exceeds this physical boundary, the echo top height is corrected to within this boundary range; otherwise, the original value is maintained. The collaborative adjudication module combines the compensated vertical cumulative liquid water content and the synchronously corrected echo top height into a unified early warning index and inputs it into a preset two-parameter two-dimensional threshold matrix. The thunderstorm intensity level is determined through two-parameter cross-matching. When the numerical combination of the vertical cumulative liquid water content and the echo top height exceeds the corresponding critical alarm threshold, a thunderstorm warning signal of the corresponding level is output.

[0047] Example 2: This example uses a historical weather radar volume scan dataset containing mountain elevation blocking and severe convective precipitation processes for multi-band radar collaborative monitoring tests. Gaussian white noise with a signal-to-noise ratio of 20dB and 50Hz power frequency interference harmonics are superimposed on the radar echo signals to form a radar image sensing interference background. The test platform uses a computing node configured with a digital elevation model processing unit and a multi-band signal alignment processor. The resolution of the three-dimensional equidistant grid space is set to 1km × 1km × 0.5km, and the volume scan sampling time difference between each radar band is set to 3 minutes. Least square fitting is performed based on a historical thunderstorm case sample set to obtain the model coefficients of the binary bias function. , and The parameter of the nonlinear scaling factor is , Intercept correction term kg / m², based on the geometric attenuation relationship of the beam cross-section intercepted by the terrain obstruction surface and the constraint of the echo reconstruction intensity in the low-altitude blind zone, the terrain influence weighting coefficient is applied. The value range is set to 0.0 to 1.0.

[0048] Test sample group 1-A corresponds to the low occlusion boundary condition, radial distance The reference reflectivity factor for an S-band radar with a range of 30km. The raw reflectivity factor of X-band radar is 30 dBZ. 28dBZ, beam obstruction rate The terrain influence weighting coefficient is 0.05. The reflectance factor is 0.00 after the deviation correction module processes the data according to the above parameters and then undergoes terrain-weighted correction. The value is 29.1302 dBZ; after gain and intercept compensation by the index reconstruction module, the compensated vertical cumulative liquid water content is... for The weight per square meter (kg / m²) has a root mean square error of 4.2% and an echo top height deviation of 0.12 km.

[0049] Test sample group 1-B uses standard parameters, radial distance For 50km, reference reflectivity factor 40 dBZ, original reflectivity factor 35dBZ, beamout rate The terrain influence weighting coefficient is 0.30. The reflectivity factor is 0.50, which is obtained by embedding the beam obstruction rate as a physical weight into the binary bias function and then adjusting it for terrain weighting. The value is 38.3502 dBZ; after compensation by nonlinear scaling factor and intercept correction term, the vertical cumulative liquid water content is... for kg / m², with a root mean square error of 4.8% and an echo top height deviation of 0.15km.

[0050] Test sample group 1-C corresponds to the high shading limit condition, radial distance 80km, reference reflectivity factor 50 dBZ, original reflectivity factor 38dBZ, beam obstruction rate The terrain influence weighting coefficient is 0.65. The reflectance factor is 1.00 after the deviation correction module completes terrain-weighted correction. The value is 46.8521 dBZ; after the index reconstruction module completes gain and intercept compensation, the vertical cumulative liquid water content is... for The root mean square error of the vertical cumulative liquid water content is 5.3% and the echo top height deviation is 0.18km. As can be seen from the above three working conditions, under the condition that the beam obstruction rate increases with the radial distance, the root mean square error of the vertical cumulative liquid water content after terrain weighting correction and nonlinear scaling compensation remains within a small range.

[0051] Control group 1, with terrain-weighted correction removed, was used to compare the results when beamout rate was not used as a physical weight. Control group 1-A used radial distance. For 50km, reference reflectivity factor 40 dBZ, original reflectivity factor For 35dBZ and beam obstruction rate The input data is 0.30, and the terrain influence weighting coefficient is applied. Set to 0.00, after processing with the binary bias function, the reflectivity factor is... The vertical cumulative liquid water content is 37.9132 dBZ. for kg / m², with a root mean square error of 14.6% and an echo top height deviation of 0.42 km. Control sample group 1-B used radial distance... For 80km, reference reflectivity factor 50 dBZ, original reflectivity factor 38dBZ and beam obstruction rate Input data with a terrain influence weighting coefficient of 0.65 Similarly, setting it to 0.00, the resulting reflectance factor is... The vertical cumulative liquid water content is 42.1150 dBZ. for The root mean square error of the sample is 22.8%, and the echo top height deviation is 0.65km. As the beam obstruction rate increases, the processing results without introducing the physical weight of the beam obstruction rate show a large reflectivity attenuation deviation, which is transmitted to the vertical cumulative liquid water content and echo top height.

[0052] Control group 2 retains the topographic weighting correction but does not use the nonlinear scaling factor and intercept correction term to compensate for the vertical cumulative liquid water content. Control group 2-A uses radial distance. For 50km, reference reflectivity factor 40 dBZ, original reflectivity factor 35dBZ, beamout The weighting factor is 0.30 and the terrain influence factor. The reflectance factor is calculated based on input data of 0.50 and adjusted for terrain weighting. The value is 38.3502 dBZ. The vertical cumulative liquid water content obtained directly from integrating this reflectivity factor is... kg / m², with a root mean square error of 18.5% and an echo top height deviation of 0.51 km. Control sample group 2-B used radial distance... For 80km, reference reflectivity factor 50 dBZ, original reflectivity factor 38dBZ, beamout rate The weighting factor is 0.65 and the terrain influence factor. The reflectance factor is calculated based on input data of 1.00 and adjusted for terrain weighting. The value is 46.8521 dBZ, and the vertical cumulative liquid water content obtained by direct integration is... The root mean square error of the reflectivity factor is 28.2%, and the echo top height deviation is 0.78 km. If the reflectivity factor is not nonlinearly scaled and intercept compensated after terrain weighting correction, the calculation error of the vertical cumulative liquid water content will still increase with the increase of radial distance and reference reflectivity factor.

[0053] Control group 3 uses the topographic influence weighting coefficient The value was set to 1.30 to examine overcorrection when it exceeded the range of 0.0 to 1.0. The control sample group 3-A used radial distance... For 50km, reference reflectivity factor 40 dBZ, original reflectivity factor For 35dBZ and beam obstruction rate The reflectance factor obtained after processing input data of 0.30 The vertical cumulative liquid water content is 38.7845 dBZ. for kg / m², with a root mean square error of 12.3% and an echo top height deviation of 0.38 km. The control sample group 3-B used radial distance... For 80km, reference reflectivity factor 50 dBZ, original reflectivity factor 38dBZ and beam obstruction rate The reflectance factor obtained after processing input data of 0.65 The vertical cumulative liquid water content is 48.2610 dBZ. for The root mean square error of the reflectivity factor is 19.4%, and the echo top height deviation is 0.55km. After the topographic influence weighting coefficient exceeds the set range, the reflectivity factor becomes too high, and the deviation between the vertical cumulative liquid water content and the echo top height increases accordingly.

[0054] Under conditions of mountainous elevation obstruction and strong electromagnetic interference, terrain weighting is applied using beam obstruction rate, and compensation for vertical cumulative liquid water content is performed using nonlinear scaling factor and intercept correction term. As a result, the root mean square error of vertical cumulative liquid water content output by the system is reduced by more than 40%, the echo top height deviation is reduced by more than 30%, and the unified early warning indicators output by S-band radar, C-band radar and X-band radar remain consistent.

[0055] Example 3: This example compares the test results in a mountainous beam-blocking area. Without applying terrain-weighted correction, the beam-blocking rate... Increased to 0.65, the topographically weighted reflectance factor of control group 1-B With a resolution of only 42.1150 dBZ, the root mean square error of vertical cumulative liquid water content reached 22.8%, and the echo top height deviation reached 0.65 km. This group only used a binary deviation function for compensation and did not use the beam obstruction rate as a physical weight in the calculation, thus failing to offset the beam cutoff caused by the mountainous terrain.

[0056] Beam blocking rate After being embedded as physical weights in the binary bias function, the topographically weighted reflectance factors of experimental groups 1-B and 1-C are... The values ​​are 38.3502 dBZ and 46.8521 dBZ, respectively, with root mean square errors of vertical cumulative liquid water content of 4.8% and 5.3%, respectively. Terrain-weighted correction is used to compensate for the attenuation echo in the blocked area according to the degree of beam blockage.

[0057] Control group 2 retained the topographic weighting correction, but did not utilize the nonlinear scaling factor and intercept correction term to compensate for the vertical cumulative liquid water content, which increased with radial distance. and reference reflectivity factor As the vertical cumulative liquid water content obtained by direct integration increases, it becomes downward biased, with the root mean square error rising to 28.2%. The index reconstruction module uses a nonlinear scaling factor for gain compensation and, after superimposing an intercept correction term, reduces the accumulation of reflectivity factor error in the power operation and vertical integration process.

[0058] Control group 3 uses the topographic influence weighting coefficient Setting it to 1.30 exceeds the range of 0.0 to 1.0. In this case, the topographically weighted reflectivity factor is too high, the root mean square error of vertical cumulative liquid water content rises to 12.3%, and the echo top height deviation increases to 0.38 km. Therefore, the topographic influence weighting coefficient should be adjusted. By keeping it between 0.0 and 1.0, the overcorrection of the reflectivity factor can be limited while compensating for terrain shading attenuation.

[0059] Example 4: This example combines Figures 1 to 2 This paper describes a radar-coordinated monitoring system for complex terrain, such as... Figure 1 As shown, the echo acquisition module synchronously acquires three-dimensional volume scan echo data from S-band, C-band, and X-band radars in the monitoring area and outputs a three-dimensional polar coordinate dataset. The spatiotemporal mapping module, connected to the echo acquisition module, uses the coordinate grid of the S-band radar as a reference to interpolate and map the data from the X-band and C-band radars in the three-dimensional polar coordinate dataset to a three-dimensional equidistant grid space and outputs the three-dimensional equidistant grid space data. The deviation correction module, connected to the spatiotemporal mapping module, uses the reflectedivity factor of the mapped S-band radar as a reference to construct a binary deviation function that varies with radial distance and the reference. It extracts digital elevation model data to calculate the beam obstruction rate, embeds the beam obstruction rate as a physical weight into the binary deviation function, and generates a terrain-weighted reflectedivity factor. The reconstruction module connects to the deviation correction module. Based on the reflectivity factor corrected by terrain weighting, it calculates the vertical cumulative liquid water content and echo top height of each grid point. It uses a nonlinear scaling factor that varies with radial distance and reference to superimpose gain and intercept compensation on the vertical cumulative liquid water content to output a unified early warning index. The index reconstruction module applies spatial gradient matching calculation to the cross-layer verification submodule. When the low-level reflectivity factor changes due to terrain weighting correction, the cross-layer verification submodule retrieves the meteorological thermodynamic adiabatic rising envelope as the physical boundary to synchronously correct the upper-level echo top height. The collaborative adjudication module connects the index reconstruction module and the cross-layer verification submodule. Based on the unified early warning index and the synchronous correction results of the cross-layer verification submodule, it classifies the thunderstorm intensity level and outputs an early warning signal.

[0060] like Figure 2As shown, the radar site includes an optoelectronic focusing device and a true north calibrator, used to perform zero-point pointer operations on the antenna elevation angles of S-band, C-band, and X-band radars in the weather radar monitoring network. The weather radar monitoring network transmits the acquired three-dimensional volume scan echo data to the computing node. The computing node is equipped with a multi-band signal alignment processor and a digital elevation model processing unit, used to process the three-dimensional volume scan echo data and generate fitting parameters. The computing node then performs a write operation to store the generated fitting parameters in the system parameter memory.

[0061] Example 5: Before the deployment and operation of the multi-band radar collaborative monitoring system, retrieve no less than 500 sets of unattenuated heavy precipitation body sweep sample data accumulated over the years in the monitoring area, using the S-band radar reflectivity factor. As a reference standard without attenuation, radial distance is extracted from the sample data. X-band radar raw reflectivity factor at ranges of 10km to 120km Data points ranging from 20 dBZ to 65 dBZ were used; the deviation of each data point relative to the unattenuated reference was calculated, a sample matrix for model fitting was constructed, and then the binary deviation function was nonlinearly fitted using least squares estimation to determine the model coefficients. , and constant intercept .

[0062] A linear reflectance factor is obtained by converting the topographically weighted reflectance factor, and the corresponding vertical cumulative liquid water content is calculated. This vertical cumulative liquid water content is then compared with the original vertical cumulative liquid water content calculated based on the uncorrected reflectance factor. By performing ratio comparisons, the parameters of the nonlinear scaling factor are optimized, and the bias adjustment parameters are calibrated. , and intercept correction term After calibration, the fitting parameters of the binary deviation function and the nonlinear scaling factor are written into the system parameter memory for use by the deviation correction module and the index reconstruction module during real-time monitoring.

[0063] During the on-site installation and initialization of the spatiotemporal mapping module at the radar site, the measured elevation of the site was retrieved. Using a high-precision digital elevation model mesh, the antenna elevation angle was calibrated using an electromagnetic focusing instrument and a true north calibrator. The zero-position pointer controls the mechanical pointing tolerance within 0.05°, and then a standard atmospheric refractive index model is used to adjust the antenna elevation angle. The beam is scanned layer by layer from 0.5° to 19.5° to calculate the elevation envelope of the beam centerline corresponding to each elevation angle; the intersection of the beam propagation path and the elevation blocking surface of the digital elevation model is obtained by tracing the 360 ​​azimuth ray paths radiating outward from the radar site.

[0064] For each grid point in the three-dimensional equidistant grid space, calculate the blocking area of ​​the beam cross-section intercepted by the terrain obstruction surface within the antenna half-power angular diffusion range, and use the ratio of this blocking area to the total beam cross-section as the beam obstruction rate. This generates a field basic verification library for beam obstruction rate. After the system receives real-time three-dimensional volume scan echo data, the deviation correction module directly calls the corresponding beam obstruction rate in the verification library based on the grid point coordinates, embeds it as a physical weight into the binary deviation function, and applies terrain-weighted correction to the original reflectivity factor of the X-band radar.

[0065] Example 6: In response to the 2.50dBZ hardware power attenuation of the X-band radar transmitter and the ground clutter interference caused by the super-refraction of ground waves in the valley terrain, the collaborative adjudication module retrieves the real-time parameters of the spatiotemporal mapping module and the deviation correction module to perform online offset calibration and clutter removal on the multi-band radar image sensing data link.

[0066] During a clear-sky window with no precipitation, using the S-band radar reference reflectivity factor As a benchmark, the systematic negative deviation of the X-band radar at the zero-obstruction far-field grid points is calculated, and based on this negative deviation, the constant intercept in the binary deviation function is determined. For calibration, the binary deviation function is: In this embodiment, the constant intercept is... The calibration was changed from 1.20dBZ to 3.70dBZ to compensate for the echo intensity deviation caused by the 2.50dBZ hardware power attenuation of the X-band radar transmitter.

[0067] For advection clutter generated by elevation obstruction in valley terrain and surface superrefractive areas, the collaborative adjudication module uses the zero variance of the Doppler velocity field and the vertical reflectivity gradient threshold of 8 dBZ / km as joint criteria to screen abnormal grids in the three-dimensional equidistant grid space. After identifying clutter-contaminated grids, the reflectivity factor and horizontal wind field Doppler velocity vector of the surrounding 26 clutter-free grid points are extracted. Based on the interpolation weight according to the distance between the grid points and the clutter-contaminated grids, the reflectivity factor and horizontal wind field Doppler velocity vector of the clutter-contaminated grids are reconstructed using an inverse distance weighted interpolation algorithm.

[0068] The index reconstruction module obtains the compensated vertical cumulative liquid water content based on the reflectivity factor after offset calibration and clutter removal. and echo top height The two are combined into a unified early warning indicator and transmitted to the collaborative decision-making module. The collaborative decision-making module calls a preset two-parameter two-dimensional threshold matrix to classify the thunderstorm intensity level: when kg / m² and When km, it is judged to be of weak convection level; when kg / m² kg / m², or km When km, it is judged to be of moderate to strong convection level; when kg / m² and At a speed of km, it was determined to be a severe thunderstorm and hail level.

[0069] At the target grid points in the deep valley area, the beam obstruction rate of the X-band radar The reflectance factor correction value is 0.80, obtained after terrain-weighted correction. The concentration was 6.50 dBZ, and the vertical cumulative liquid water content in the lower single layer increased to [missing value]. kg / m², the cross-layer verification submodule calculates the spatial difference of the low-layer reflectivity factor and the spatial difference of the high-layer echo top height between the target grid point and the adjacent grid point, and compares the gradient direction and gradient change position of the two. When the spatial gradient change direction of the low-layer reflectivity factor after terrain weighting correction is consistent with the gradient change direction of the high-layer echo top height, the echo top height linkage correction is performed.

[0070] The cross-layer verification submodule retrieves the meteorological thermodynamic adiabatic ascent envelope calculated from the radiosonde data of the monitoring area, obtains the upper limit of thermodynamic buoyancy height of 10.5km, and uses it as the physical boundary of the upper echo top height. Based on the spatial gradient matching results, the upper uncorrected echo top height of 8.2km is corrected to 9.8km. Since this correction value does not exceed the upper limit of thermodynamic buoyancy height of 10.5km, 9.8km is retained as the corrected echo top height. The collaborative adjudication module inputs the compensated vertical cumulative liquid water content and the corrected echo top height into a two-parameter two-dimensional threshold matrix to determine whether the target grid point reaches the level of severe thunderstorm and hail, and outputs a severe thunderstorm imminent warning signal containing three-dimensional spatial location and intensity level.

[0071] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit of this application and the scope of protection of this invention, and all of these forms are within the protection scope of this application.

Claims

1. A radar-coordinated monitoring system for complex terrain, characterized in that, include: The echo acquisition module is used to simultaneously acquire three-dimensional volume scan echo data from S-band, C-band, and X-band radars in the monitoring area and output a three-dimensional polar coordinate dataset. The spatiotemporal mapping module, connected to the echo acquisition module, is used to interpolate and map the data from the X-band and C-band radars in the three-dimensional polar coordinate dataset to a three-dimensional equidistant grid space, based on the coordinate grid of the S-band radar. The deviation correction module, connected to the spatiotemporal mapping module, is used to construct a binary deviation function that varies with radial distance and the benchmark based on the reflectivity factor of the mapped S-band radar, extract digital elevation model data to calculate beam obstruction rate, embed the beam obstruction rate as a physical weight into the binary deviation function, and generate a terrain-weighted reflectivity factor. The index reconstruction module, connected to the deviation correction module, is used to calculate the vertical cumulative liquid water content and echo top height of each grid point based on the reflectivity factor corrected by terrain weighting. It uses a nonlinear scaling factor that varies with radial distance and reference to superimpose gain and intercept compensation on the vertical cumulative liquid water content and outputs a unified early warning index. The collaborative adjudication module, connected to the indicator reconstruction module, is used to classify thunderstorm intensity levels and output warning signals based on unified early warning indicators.

2. The radar cooperative monitoring system for complex terrain according to claim 1, characterized in that, The deviation correction module calculates the terrain-weighted reflectance factor. The formula for its calculation is: ,in, This represents the original reflectivity factor of the X-band radar at the grid points. It is a bivariate deviation function. For beam blocking rate, This is the terrain influence weighting coefficient, with a value ranging from 0.0 to 1.

0.

3. The radar cooperative monitoring system for complex terrain according to claim 2, characterized in that, The construction process of the binary deviation function is as follows: taking the reflectivity factor of the S-band radar as the unattenuated reference benchmark, the radial distance from the grid point to the radar site is obtained, the deviation of the reflectivity factor of the X-band radar relative to the unattenuated reference benchmark is calculated, and the deviation compensation model consisting of the product of power functions and the constant intercept is determined by fitting the historical thunderstorm sample set.

4. A radar cooperative monitoring system for complex terrain according to claim 1, characterized in that, The process of extracting beam obstruction rate by the deviation correction module is as follows: Based on the elevation of the radar site, the antenna elevation angle and the atmospheric refraction model, the elevation of the radar beam centerline and the beam half-power angle spread range in the three-dimensional grid space are determined. The elevation obstruction relationship of the digital elevation model data on the ray path from the radar site to the target grid point is extracted. The proportion of the beam cross-sectional area intercepted by the terrain obstruction surface is calculated to obtain the beam obstruction rate.

5. A radar cooperative monitoring system for complex terrain according to claim 1, characterized in that, The process of constructing the nonlinear scaling factor in the index reconstruction module is as follows: extract the radial distance from the grid point to the radar site, multiply the radial distance by the power term of the reference using the bias adjustment parameter, and superimpose a constant to generate the nonlinear scaling factor. The index reconstruction module multiplies the vertical cumulative liquid water content by a nonlinear scaling factor and adds an intercept correction term to obtain the compensated vertical cumulative liquid water content.

6. A radar cooperative monitoring system for complex terrain according to claim 1, characterized in that, The spatiotemporal mapping module interpolates and maps the X-band and C-band radar data in the three-dimensional polar coordinate dataset to a three-dimensional equidistant grid space with a spatial resolution of 1km×1km×0.5km, and controls the time difference of each band radar volume scan in the echo acquisition module to be within 3 minutes.

7. A radar cooperative monitoring system for complex terrain according to claim 1, characterized in that, The spatiotemporal mapping module is also used to obtain the horizontal wind field Doppler velocity vector of the monitoring area. Based on the time difference of multi-band radar volume scan sampling, it calculates the spatial displacement of the echo centroid along the horizontal wind field Doppler velocity vector within the sampling time difference, and compensates for the reverse displacement of the coordinates in the three-dimensional equidistant grid space.

8. A radar cooperative monitoring system for complex terrain according to claim 1, characterized in that, The collaborative adjudication module includes a cross-layer verification submodule, which is used to apply spatial gradient matching calculation to the vertical cumulative liquid water content and echo top height. When the low-layer reflectivity factor changes due to topographic weighting correction, the meteorological thermodynamic adiabatic rising envelope is retrieved as the physical boundary to synchronously correct the high-layer echo top height.

9. A radar cooperative monitoring system for complex terrain according to claim 1, characterized in that, The collaborative adjudication module classifies thunderstorm intensity levels as follows: the output unified warning index is input into a preset two-dimensional threshold matrix with two parameters, and the intensity level of thunderstorms is determined by cross-matching of the two parameters; when the numerical combination of vertical cumulative liquid water content and echo top height in the unified warning index exceeds the critical alarm threshold, a warning signal of the corresponding level is generated.

10. A radar cooperative monitoring system for complex terrain according to claim 1, characterized in that, The deviation correction module is also used to establish a binary deviation compensation model of the C-band radar's reflectivity factor relative to the benchmark, based on the reflectivity factor of the mapped S-band radar. It calculates the beam obstruction rate of the C-band radar at each grid point by combining the extracted digital elevation model data, and applies terrain-weighted correction to the original reflectivity factor of the C-band radar.

Images