Short-time rainfall prediction method and device based on three-dimensional GNSS water vapor chromatography
By reconstructing the three-dimensional water vapor field of the target using 3D GNSS water vapor tomography and combining it with radar echo data for short-term rainfall prediction, the problem of insufficient spatiotemporal resolution in traditional monitoring methods is solved, enabling earlier and more accurate rainfall warnings, especially with improved adaptability in urban areas with high-rise buildings and areas with weak radar coverage at long distances.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-31
- Publication Date
- 2026-04-03
AI Technical Summary
Traditional radar or weather station monitoring lacks sufficient spatiotemporal resolution in urban short-term heavy rainfall warnings, making it difficult to accurately capture local water vapor bursts and rainfall triggering processes, resulting in low accuracy in short-term rainfall forecasts.
A method based on three-dimensional GNSS water vapor tomography is adopted. By acquiring multi-source data, a near-surface wet layer enhancement constraint with a preset three-dimensional layered and encrypted grid is constructed. The target three-dimensional water vapor field is reconstructed using an enhanced tomography algorithm. Short-term rainfall prediction is then performed by combining the water vapor burst comprehensive index and radar echo data.
Capturing water vapor accumulation and rising signals 10–30 minutes before rainfall occurs improves the accuracy and stability of urban short-term rainstorm forecasts, especially in areas with traditional blind spots, supporting rapid early warning and dispatching of urban stormwater risks.
Smart Images
Figure CN121784867A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of meteorological information and urban stormwater management, and in particular to a method and apparatus for short-term rainfall prediction based on three-dimensional GNSS water vapor tomography. Background Technology
[0002] Urban short-duration heavy rainfall develops rapidly, and traditional radar or weather station monitoring suffers from insufficient spatiotemporal resolution, making accurate early warning difficult. Water vapor changes are a crucial physical driver of rainfall formation, but current urban rainfall warning systems do not fully utilize high-resolution water vapor information. Data from a single radar or weather station is insufficient to capture local water vapor bursts and rainfall triggering processes. Limited water vapor observation accuracy makes it difficult to achieve millimeter-level precision in short-term forecasts, resulting in low accuracy in short-duration rainfall prediction. Summary of the Invention
[0003] In view of this, the present invention provides a method, device, storage medium and electronic device for short-term rainfall prediction based on three-dimensional GNSS water vapor tomography, the main purpose of which is to solve the problem of low accuracy of short-term rainfall prediction.
[0004] To address the aforementioned problems, this application provides a short-term rainfall prediction method based on three-dimensional GNSS water vapor tomography, comprising: Acquire multi-source data for short-term rainfall forecasting, including GNSS observation data, meteorological station surface observation data, and radar echo data; Near-ground wet layer enhancement constraint is constructed based on the GNSS observation data and the meteorological station surface observation data to build a preset three-dimensional layered and densified grid. Based on the enhanced constraint of the near-surface wet layer and the multi-source data, the enhanced tomography algorithm is used to reconstruct the water vapor field within the preset three-dimensional hierarchical densified grid to obtain the target three-dimensional water vapor field that meets the preset conditions. Based on the target three-dimensional water vapor field, a comprehensive water vapor burst index is obtained through calculation and processing. Short-term rainfall prediction results are obtained by performing short-term rainfall prediction based on the comprehensive water vapor burst index and the radar echo data.
[0005] Optionally, the near-surface wet layer enhancement constraint based on the GNSS observation data and the meteorological station surface observation data specifically includes: The initial tomography matrix is obtained by performing calculations based on the GNSS observation data and the preset three-dimensional hierarchical encrypted grid. The initial wet delay observation vector is obtained by performing calculations based on the GNSS observation data and the meteorological station surface observation data. The signal-to-noise ratio in the GNSS observation data is inverted to calculate the change in wet layer thickness, thus obtaining the near-surface wet layer thickness disturbance. Based on the initial tomography matrix and the near-surface wet layer thickness disturbance, the low-altitude constraint matrix is obtained by calculation using the first function. Based on the initial wet delay observation vector and the near-ground wet layer thickness disturbance, the second function is used to calculate and process the low-altitude constraint vector. Constraint conditions are constructed based on the low-altitude constraint matrix and the low-altitude constraint vector to obtain the near-ground wet layer enhancement constraint.
[0006] Optionally, the step of inverting the signal-to-noise ratio in the GNSS observation data to calculate the wet layer thickness variation and obtain the near-surface wet layer thickness disturbance specifically includes: The signal-to-noise ratio of data collected from different satellites within a predetermined elevation angle range is analyzed to obtain the interference frequency. The near-surface wet layer thickness disturbance is obtained by inverting the frequency change between two adjacent moments based on the interference frequency.
[0007] Optionally, the step of reconstructing the water vapor field within the preset three-dimensional hierarchical grid using an enhanced tomography algorithm based on the near-surface moist layer enhancement constraint and the multi-source data to obtain a target three-dimensional water vapor field that meets preset conditions specifically includes: Based on the initial tomography matrix and the low-altitude constraint matrix, an enhancement coefficient matrix is obtained through calculation. The enhanced observation vector is obtained by performing calculations based on the initial wet delay observation vector and the low-altitude constraint vector. Based on the enhancement coefficient matrix and the enhancement observation vector, a multi-constraint hybrid iterative method is used to reconstruct the water vapor field within the preset three-dimensional hierarchical encrypted grid, thereby obtaining the target three-dimensional water vapor field with the objective of minimizing the residual between the predicted wet delay and the observed wet delay.
[0008] Optionally, the calculation and processing based on the target three-dimensional water vapor field to obtain the comprehensive water vapor burst index specifically includes: The precipitable water volume is calculated by integrating the three-dimensional water vapor density of the target three-dimensional water vapor field in the vertical direction at different time periods, and the rapid accumulation rate of water vapor is obtained. Spatial difference calculations are performed on the different directional dimensions of the target three-dimensional water vapor field to obtain the three-dimensional water vapor gradient. The water vapor density in the near-surface layer of the target three-dimensional water vapor field is integrated to obtain the enhanced water vapor quantity in the lower atmosphere. The vertical time difference calculation process is performed on the target three-dimensional water vapor field to obtain the vertical water vapor rise velocity. The comprehensive water vapor burst index is obtained by fusing three-dimensional water vapor information based on the rapid water vapor accumulation rate, the three-dimensional water vapor gradient, the enhanced water vapor volume at low altitudes, and the vertical water vapor rise velocity.
[0009] Optionally, the step of performing short-term rainfall prediction based on the water vapor burst composite index and the radar echo data to obtain short-term rainfall prediction results specifically includes: The water vapor burst threshold is determined based on the historical sequence of the comprehensive water vapor burst index of the target area and the actual rainfall label corresponding to the historical sequence of the comprehensive water vapor burst index. Determine the radar echo increment threshold based on historical radar echo data of the target area; Short-term rainfall prediction is performed based on the comprehensive water vapor burst index, the radar echo increment in the radar echo data of the target area acquired in real time, the water vapor burst threshold, and the radar echo increment threshold, to obtain the short-term rainfall prediction result.
[0010] Optionally, the step of performing short-term rainfall prediction based on the comprehensive water vapor burst index, real-time acquired radar echo data of the target area, the water vapor burst threshold, and the radar echo increment threshold to obtain short-term rainfall prediction results specifically includes: When the time interval between the first moment when the comprehensive water vapor burst index is greater than or equal to the water vapor burst threshold and the second moment when the radar echo increment is greater than or equal to the radar echo increment threshold is less than or equal to a preset time interval threshold, a short-term heavy rainfall warning is triggered.
[0011] To address the aforementioned problems, this application provides a short-term rainfall prediction device based on three-dimensional GNSS water vapor tomography, comprising: The acquisition module is used to acquire multi-source data for short-term rainfall prediction, including GNSS observation data, meteorological station surface observation data, and radar echo data. The constraint construction module is used to construct near-ground wet layer enhanced constraints of a preset three-dimensional layered densified grid based on the GNSS observation data and the meteorological station surface observation data. The reconstruction module is used to reconstruct the water vapor field within the preset three-dimensional hierarchical densified grid based on the enhanced constraint of the near-surface wet layer and the multi-source data using an enhanced tomography algorithm, so as to obtain the target three-dimensional water vapor field that meets the preset conditions. The calculation module is used to perform calculations based on the target three-dimensional water vapor field to obtain the comprehensive water vapor burst index; The prediction module is used to predict short-term rainfall based on the water vapor burst index and the radar echo data, and obtain the short-term rainfall prediction result.
[0012] To address the aforementioned problems, this application provides a storage medium storing a computer program that, when executed by a processor, implements the steps of the aforementioned short-term rainfall prediction method based on three-dimensional GNSS water vapor tomography.
[0013] To address the aforementioned problems, this application provides an electronic device, comprising at least a memory and a processor. The memory stores a computer program, and the processor, when executing the computer program in the memory, implements the steps of the aforementioned short-term rainfall prediction method based on three-dimensional GNSS water vapor tomography.
[0014] The beneficial effects of this application are as follows: By synergistically analyzing the three-dimensional structure of water vapor and radar echoes, this application can capture signals of water vapor accumulation and upward development 10–30 minutes before rainfall occurs, providing a significantly better early warning lead than traditional methods that rely solely on echoes. This method is more adaptable to traditional observation blind spots, such as urban areas obscured by tall buildings and areas with weak radar coverage at long distances. Furthermore, it can improve the accuracy and stability of short-term urban rainstorm forecasts without requiring additional hardware, effectively supporting rapid early warning and management of urban stormwater risks.
[0015] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description
[0016] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating a short-term rainfall prediction method based on three-dimensional GNSS water vapor tomography provided in an embodiment of this application is shown. Figure 2 A flowchart illustrating a short-term rainfall prediction method based on three-dimensional GNSS water vapor tomography, according to another embodiment of this application, is shown. Figure 3 A structural block diagram of a short-term rainfall prediction device based on three-dimensional GNSS water vapor tomography, provided in another embodiment of this application, is shown. Detailed Implementation
[0017] Various embodiments and features of this application are described herein with reference to the accompanying drawings.
[0018] It should be understood that various modifications can be made to the embodiments described herein. Therefore, the above description should not be considered as limiting, but merely as an example of embodiments. Other modifications within the scope and spirit of this application will be apparent to those skilled in the art.
[0019] The accompanying drawings, which are included in and form part of this specification, illustrate embodiments of the present application and, together with the general description of the present application given above and the detailed description of the embodiments given below, serve to explain the principles of the present application.
[0020] These and other features of this application will become apparent from the following description of preferred forms of embodiments given as non-limiting examples, with reference to the accompanying drawings.
[0021] It should also be understood that although this application has been described with reference to some specific examples, those skilled in the art can certainly implement many other equivalent forms of this application.
[0022] The above and other aspects, features and advantages of this application will become more apparent when taken in conjunction with the accompanying drawings and in view of the following detailed description.
[0023] Specific embodiments of this application are described thereafter with reference to the accompanying drawings; however, it should be understood that the claimed embodiments are merely examples of this application, which can be implemented in various ways. Well-known and / or repeated functions and structures are not described in detail to avoid unnecessary or redundant details that could obscure the application. Therefore, the specific structural and functional details claimed herein are not intended to be limiting, but merely to teach those skilled in the art to use this application in a variety of substantially any suitable detailed structures.
[0024] This specification may use the phrases “in one embodiment,” “in another embodiment,” “in yet another embodiment,” or “in other embodiments,” all of which may refer to one or more of the same or different embodiments according to this application.
[0025] This application provides a short-term rainfall prediction method based on three-dimensional GNSS water vapor tomography, such as... Figure 1 As shown, it includes: Step S101: Obtain multi-source data for short-term rainfall prediction, including GNSS observation data, meteorological station surface observation data, and radar echo data; In this step, the GNSS observation data includes real-time acquisition by multiple GNSS receivers deployed in the urban area, including: carrier phase observations, pseudorange observations, signal-to-noise ratio, satellite position, and satellite rise and fall angles (calculated from broadcast ephemeris). The sampling frequency can be 1 Hz, 5 Hz, or 30 s. Reference station correction data includes RTCM format data from adjacent reference stations, used to correct the GNSS observation data, stabilize relative positioning observations, and eliminate system biases. The meteorological station surface observation data includes temperature, humidity, air pressure, wind speed, and wind direction, used for calibrating meteorological parameters for water vapor inversion. The radar echo data is acquired through weather radar, including echo intensity in dBZ, liveness development trend, and vertical echo profile. The radar data time resolution is generally 6 minutes.
[0026] Step S102: Construct a near-surface wet layer enhancement constraint based on the GNSS observation data and the meteorological station surface observation data; In this step, the initial tomography matrix is obtained by performing calculations based on the GNSS observation data and the preset three-dimensional layered encrypted grid; the initial wet delay observation vector is obtained by performing calculations based on the GNSS observation data and the meteorological station surface observation data; the wet layer thickness variation is inverted and calculated based on the signal-to-noise ratio in the GNSS observation data to obtain the near-surface wet layer thickness perturbation; the low-altitude constraint matrix is obtained by performing calculations based on the initial tomography matrix and the near-surface wet layer thickness perturbation using a first function; the low-altitude constraint vector is obtained by performing calculations based on the initial wet delay observation vector and the near-surface wet layer thickness perturbation using a second function; and the near-surface wet layer enhanced constraint is obtained by constructing constraint conditions based on the low-altitude constraint matrix and the low-altitude constraint vector.
[0027] Step S103: Based on the enhanced constraint of the near-surface moist layer and the multi-source data, the enhanced tomography algorithm is used to reconstruct the water vapor field within the preset three-dimensional hierarchical densified grid to obtain the target three-dimensional water vapor field that meets the preset conditions; In this step, the initial tomography matrix and the low-altitude constraint matrix are used to calculate and process the enhancement coefficient matrix; the initial wet delay observation vector and the low-altitude constraint vector are used to calculate and process the enhanced observation vector; and the enhanced coefficient matrix and the enhanced observation vector are used to reconstruct the water vapor field within the preset three-dimensional hierarchical densified grid using a multi-constraint hybrid iterative method to obtain the target three-dimensional water vapor field with the objective of minimizing the residual between the predicted wet delay and the observed wet delay.
[0028] Step S104: Perform calculations based on the target three-dimensional water vapor field to obtain the comprehensive water vapor burst index; In this step, the precipitable water volume is calculated based on the vertical integration of the three-dimensional water vapor density of the target three-dimensional water vapor field at different time periods to obtain the rapid water vapor accumulation rate; spatial difference calculation is performed on the target three-dimensional water vapor field in different directional dimensions to obtain the three-dimensional water vapor gradient; the water vapor density of the target near-surface layer of the target three-dimensional water vapor field is integrated to obtain the low-altitude enhanced water vapor volume; vertical time difference calculation is performed on the target three-dimensional water vapor field to obtain the vertical water vapor rise velocity; and three-dimensional water vapor information is fused based on the rapid water vapor accumulation rate, the three-dimensional water vapor gradient, the low-altitude enhanced water vapor volume, and the vertical water vapor rise velocity to obtain the comprehensive water vapor burst index.
[0029] Step S105: Based on the water vapor burst comprehensive index and the radar echo data, perform short-term rainfall prediction to obtain short-term rainfall prediction results.
[0030] In the specific implementation of this step, when the time interval between the first moment when the comprehensive water vapor burst index is greater than or equal to the water vapor burst threshold and the second moment when the radar echo increment is greater than or equal to the radar echo increment threshold is less than or equal to a preset time interval threshold, a short-term heavy rainfall warning is triggered.
[0031] This application utilizes the synergistic analysis of water vapor three-dimensional structure and radar echoes to capture water vapor accumulation and rising development signals 10–30 minutes before rainfall, providing a significantly better early warning lead than traditional methods relying solely on echoes. This method is more adaptable to traditional observation blind spots, such as urban areas obscured by tall buildings and areas with weak radar coverage at long distances. Furthermore, it improves the accuracy and stability of short-term urban rainstorm forecasts without requiring additional hardware, effectively supporting rapid early warning and management of urban stormwater risks.
[0032] Another embodiment of this application provides a different method for short-term rainfall prediction based on three-dimensional GNSS water vapor tomography, such as... Figure 2 As shown, it includes: Step S201: Obtain multi-source data for short-term rainfall prediction, including GNSS observation data, meteorological station surface observation data, and radar echo data; In this step, the GNSS observation data includes real-time acquisition by multiple GNSS receivers deployed in the urban area, including: carrier phase observations, pseudorange observations, signal-to-noise ratio, satellite position, and satellite rise and fall angles (calculated from broadcast ephemeris). The sampling frequency can be 1 Hz, 5 Hz, or 30 s. Reference station correction data includes RTCM format data from adjacent reference stations, used to correct the GNSS observation data, stabilize relative positioning observations, and eliminate system biases. The meteorological station surface observation data includes temperature, humidity, air pressure, wind speed, and wind direction, used for calibrating meteorological parameters for water vapor inversion. The radar echo data is acquired through weather radar, including echo intensity in dBZ, liveness development trend, and vertical echo profile. The radar data time resolution is generally 6 minutes.
[0033] Step S202: Perform calculations based on the GNSS observation data and the preset three-dimensional hierarchical encrypted grid to obtain the initial tomography matrix; In this step, calculations are performed based on the satellite position, receiver position, and a pre-defined three-dimensional hierarchical encrypted grid to obtain the initial tomography matrix. Specifically, each satellite signal propagates along a ray from the satellite to the receiver, and each element in the initial tomography matrix represents the first ray. i The ray in the preset three-dimensional layered encrypted mesh j The length of the signal passing through each small grid (voxel) is 0 if it does not pass through. In other words, the initial tomography matrix A completely reflects the geometric relationship between the satellite signal and the three-dimensional grid.
[0034] Step S203: Based on the GNSS observation data and the meteorological station surface observation data, perform calculations to obtain the initial wet delay observation vector; In this step, based on GNSS observation data (carrier phase, pseudorange) and meteorological parameters (temperature, air pressure, humidity, etc.), the wet delay or water vapor delay of each satellite signal ray is calculated. This represents the total delay of the satellite signal after passing through atmospheric water vapor, thus obtaining the initial wet delay observation vector. L .
[0035] Step S204: Perform inversion calculation of the wet layer thickness variation in the signal-to-noise ratio of the GNSS observation data to obtain the near-surface wet layer thickness disturbance; In this step, spectral analysis is performed on the SNR (signal-to-noise ratio) of satellites collected from different satellites within a predetermined elevation angle range to obtain the interference frequency. The predetermined elevation angle range is the range of 5° to 25° low elevation angle; the mathematical expression is as follows:
[0036] in, h This is the equivalent height of the wet layer. The carrier wavelength of the GNSS signal is given. Based on the interference frequency, the frequency change between two adjacent moments is inverted to calculate the near-surface wet layer thickness disturbance, expressed mathematically as follows:
[0037] in, represent k Interference frequency at time, represent k-1 Interference frequency at time; This represents the disturbance of the near-surface wet layer thickness.
[0038] Step S205: Based on the initial tomography matrix and the near-surface wet layer thickness disturbance, the low-altitude constraint matrix is obtained by calculation using the first function. ; In the specific implementation of this step, the mathematical expression of the first function is as follows:
[0039] in, This indicates the incremental effect of a unit increase in height on the amplitude; This indicates that even without changes in altitude, the signal still has a fundamental amplitude; Step S206: Based on the initial wet delay observation vector and the near-surface wet layer thickness disturbance, the low-altitude constraint vector is obtained by calculation using the second function. ; In the specific implementation of this step, the mathematical expression of the second function is as follows:
[0040] in, The conversion factor for wet layer thickness; This is the correction factor for the surface layer PWV; The atmospheric precipitable water content is calculated using GNSS data to characterize the overall water vapor content level at this location.
[0041] Step S207: Based on the low-altitude constraint matrix and the low-altitude constraint vector, construct the constraint conditions to obtain the near-ground wet layer enhancement constraint; In the specific implementation process of this step, the mathematical expression of the constraint conditions is as follows:
[0042] in, x Let be the three-dimensional water vapor density vector to be solved.
[0043] Step S208: Perform calculations based on the initial tomography matrix and the low-altitude constraint matrix to obtain the enhancement coefficient matrix; In the specific implementation process of this step, based on the initial tomography matrix A and the low-altitude constraint matrix... The enhancement coefficient matrix is obtained through calculation. The mathematical expression is as follows:
[0044] Step S209: Calculate and process the initial wet delay observation vector and the low-altitude constraint vector to obtain the enhanced observation vector; In the specific implementation process, this step involves calculation based on the initial wet delay observation vector and the low-altitude constraint vector to obtain the enhanced observation vector, the mathematical expression of which is as follows:
[0045] Step S210: Based on the enhancement coefficient matrix and the enhancement observation vector, a multi-constraint hybrid iterative method is used to reconstruct the water vapor field within the preset three-dimensional hierarchical densified grid, so as to obtain the target three-dimensional water vapor field with the objective of minimizing the residual between the predicted wet delay and the observed wet delay; In the specific implementation process of this step, a multi-constraint hybrid iterative equation is constructed based on the enhancement coefficient matrix and the enhancement observation vector; the mathematical expression is as follows:
[0046] in, For the first k The three-dimensional water vapor field is estimated in the next iteration, where α is the relaxation factor. To enhance the transpose of the coefficient matrix, the residuals are back-projected onto a 3D grid. This method continuously reduces the residual between the predicted and observed wet delays, gradually approximating the 3D water vapor distribution to the actual atmospheric water vapor structure, achieving high-precision inversion, especially in the low-altitude region. The preset 3D hierarchical densified grid uses a non-uniform network in the vertical direction: 0–300 m: 50 m or 100 m resolution (high density); 300–1500 m: 200–300 m resolution; above 1500 m: 300–500 m. The grid structure fully reflects the details of the urban boundary layer, improving the ability to invert low-altitude water vapor gradients and identify water vapor fronts.
[0047] Step S211: Perform calculations based on the target three-dimensional water vapor field to obtain the comprehensive water vapor burst index; In this step, the precipitable water content is calculated based on the vertical integration of the three-dimensional water vapor density of the target three-dimensional water vapor field at different time periods, to obtain the rapid water vapor accumulation rate. The mathematical expression is as follows:
[0048] in, PWV This represents the precipitable water volume obtained by integrating the three-dimensional water vapor density of the region in the vertical direction. This indicator reflects the region's ability to rapidly accumulate water vapor and is a typical early signal before a convective outbreak. represent The precipitable water content after integrating the three-dimensional water vapor density in the vertical direction at any given time; represent The precipitable water content after integrating the three-dimensional water vapor density in the vertical direction at any given time; represent Time and The time difference between moments.
[0049] Spatial difference calculations are performed on the target three-dimensional water vapor field in different directional dimensions to obtain the three-dimensional water vapor gradient. The mathematical expression is as follows:
[0050] The three-dimensional water vapor gradient is obtained by spatially differencing the enhanced tomography results in the x, y, and z directions. Regions with significantly increased gradients correspond to zones of rapid humidity changes and are used to identify key convection triggering structures such as water vapor fronts and humidity shear zones.
[0051] The water vapor density in the near-surface layer of the target three-dimensional water vapor field is integrated to obtain the enhanced water vapor quantity in the lower atmosphere. The mathematical expression is as follows:
[0052] Because enhanced tomography introduces a low-altitude constraint matrix This significantly improves the accuracy of humidity inversion in the 0–300 m near-surface layer. This index directly reflects the absolute humidity level of the near-surface layer and is crucial in determining the triggering conditions for strong convection.
[0053] The vertical time-difference calculation was performed on the target three-dimensional water vapor field to obtain the vertical water vapor rise velocity. The mathematical formula for calculation is as follows:
[0054] The upward velocity of water vapor was derived by considering the vertical variations of the three-dimensional water vapor field over multiple time points. This indicator characterizes the rate at which water vapor is transported to the upper levels and is an important precursor to the release of convective uplift energy (CAPE). Regions often correspond to central areas where flow is about to develop.
[0055] The comprehensive water vapor burst index is obtained by fusing three-dimensional water vapor information based on the rapid water vapor accumulation rate, the three-dimensional water vapor gradient, the enhanced water vapor volume at low altitudes, and the vertical water vapor ascent velocity. The mathematical expression is as follows:
[0056] in, For rapid water vapor accumulation rate; A three-dimensional water vapor gradient; This refers to the vertical velocity of rising water vapor. To increase water vapor levels at lower altitudes; , , , These are weighting coefficients, which are weight values obtained from statistical analysis of historical data of the target area.
[0057] Step S212: Based on the water vapor burst comprehensive index and the radar echo data, perform short-term rainfall prediction to obtain short-term rainfall prediction results.
[0058] In this step, the water vapor burst threshold is determined based on the historical sequence of the comprehensive water vapor burst index of the target area and the actual rainfall labels corresponding to the historical sequence of the comprehensive water vapor burst index. A radar echo increment threshold is determined based on historical radar echo data of the target area. Short-term rainfall prediction is performed based on the comprehensive water vapor burst index, the radar echo increment in the real-time acquired radar echo data of the target area, the water vapor burst threshold, and the radar echo increment threshold, yielding a short-term rainfall prediction result. A short-term heavy rainfall warning is triggered when the time interval between the first moment when the comprehensive water vapor burst index is greater than or equal to the water vapor burst threshold and the second moment when the radar echo increment is greater than or equal to the radar echo increment threshold is less than or equal to a preset time interval threshold. The preset time interval threshold can be 10 minutes and can be set according to actual needs. Subsequently, a rainfall risk distribution map can be generated based on the forecast results; the WECI radar results can be visualized as a gridded heat map, marking high-risk, medium-risk, and low-risk areas; early warning of urban waterlogging sensitive points can be issued, and combined with the specific situation of the urban drainage network, drainage network pressure, potential water accumulation points, and expected water accumulation depth can be obtained; urban emergency measures can be automatically linked, and instructions can be automatically generated according to the warning level: yellow warning (remind the public, increase patrols), orange warning (drain air conditioning reservoirs in advance, dispatch pumping stations), red warning (shut down traffic in low-lying areas, issue emergency drainage instructions); real-time rolling updates, recalculating water and air field changes, radar echo trends, and precipitation risk levels every 1-5 minutes; achieving continuous rolling early warning.
[0059] This application utilizes the synergistic analysis of water vapor three-dimensional structure and radar echoes to capture water vapor accumulation and rising development signals 10–30 minutes before rainfall, providing a significantly better early warning lead than traditional methods relying solely on echoes. This method is more adaptable to traditional observation blind spots (such as urban areas obscured by tall buildings and areas with weak radar coverage at long distances) and can improve the accuracy and stability of short-term urban rainstorm forecasts without requiring additional hardware, effectively supporting rapid early warning and management of urban stormwater risks.
[0060] Another embodiment of this application provides a short-term rainfall prediction device based on three-dimensional GNSS water vapor tomography, such as... Figure 3 As shown, it includes: Acquisition module 1 is used to acquire multi-source data for short-term rainfall prediction, including GNSS observation data, meteorological station surface observation data, and radar echo data; Constraint construction module 2 is used to construct near-ground wet layer enhanced constraints of a preset three-dimensional layered densified grid based on the GNSS observation data and the meteorological station surface observation data; Reconstruction module 3 is used to reconstruct the water vapor field within the preset three-dimensional hierarchical densified grid based on the enhanced constraint of the near-ground wet layer and the multi-source data using an enhanced tomography algorithm, so as to obtain the target three-dimensional water vapor field that meets the preset conditions. Calculation module 4 is used to perform calculations based on the target three-dimensional water vapor field to obtain the comprehensive water vapor burst index; Prediction module 5 is used to predict short-term rainfall based on the water vapor burst comprehensive index and the radar echo data, and obtain the short-term rainfall prediction result.
[0061] In the specific implementation process, the constraint construction module 2 is specifically used to perform calculations based on the GNSS observation data and the preset three-dimensional layered encrypted grid to obtain an initial tomography matrix; perform calculations based on the GNSS observation data and the meteorological station surface observation data to obtain an initial wet delay observation vector; perform wet layer thickness variation inversion calculation on the signal-to-noise ratio in the GNSS observation data to obtain near-ground wet layer thickness perturbation; perform calculations based on the initial tomography matrix and the near-ground wet layer thickness perturbation using a first function to obtain a low-altitude constraint matrix; perform calculations based on the initial wet delay observation vector and the near-ground wet layer thickness perturbation using a second function to obtain a low-altitude constraint vector; and construct constraint conditions based on the low-altitude constraint matrix and the low-altitude constraint vector to obtain the near-ground wet layer enhanced constraint.
[0062] In the specific implementation process, the constraint construction module 2 is also used to perform spectrum analysis on the signal-to-noise ratio of different satellites in the predetermined elevation angle range to obtain the interference frequency; based on the interference frequency, the frequency change at two adjacent moments is inverted to obtain the near-ground wet layer thickness disturbance.
[0063] In the specific implementation process, the reconstruction module 3 is specifically used to perform calculations based on the initial tomography matrix and the low-altitude constraint matrix to obtain the enhancement coefficient matrix; to perform calculations based on the initial wet delay observation vector and the low-altitude constraint vector to obtain the enhanced observation vector; and to reconstruct the water vapor field within the preset three-dimensional hierarchical densified grid using a multi-constraint hybrid iterative method based on the enhancement coefficient matrix and the enhanced observation vector, so as to obtain the target three-dimensional water vapor field with the objective of minimizing the residual between the predicted wet delay and the observed wet delay.
[0064] In the specific implementation process, the calculation module 4 is specifically used to calculate the precipitable water volume after integrating the three-dimensional water vapor density of the target three-dimensional water vapor field in the vertical direction at different time periods, to obtain the rapid water vapor accumulation rate; to perform spatial difference calculation on the different directional dimensions of the target three-dimensional water vapor field, to obtain the three-dimensional water vapor gradient; to perform integral calculation on the water vapor density of the target near-surface layer of the target three-dimensional water vapor field, to obtain the low-altitude enhanced water vapor volume; to perform vertical time difference calculation on the target three-dimensional water vapor field, to obtain the vertical water vapor rise velocity; and to perform three-dimensional water vapor information fusion based on the rapid water vapor accumulation rate, the three-dimensional water vapor gradient, the low-altitude enhanced water vapor volume, and the vertical water vapor rise velocity, to obtain the comprehensive water vapor burst index.
[0065] In the specific implementation process, the prediction module 5 is specifically used to determine the water vapor burst threshold based on the historical sequence of the water vapor burst comprehensive index of the target area and the actual rainfall label corresponding to the historical sequence of the water vapor burst comprehensive index; determine the radar echo increment threshold based on the historical radar echo data of the target area; and perform short-term rainfall prediction based on the water vapor burst comprehensive index, the radar echo increment in the radar echo data of the target area acquired in real time, the water vapor burst threshold, and the radar echo increment threshold to obtain the short-term rainfall prediction result.
[0066] In the specific implementation process, the prediction module 5 is also used to trigger a short-term heavy rainfall warning when the time interval between the first moment when the comprehensive water vapor burst index is greater than or equal to the water vapor burst threshold and the second moment when the radar echo increment is greater than or equal to the radar echo increment threshold is less than or equal to a preset time interval threshold.
[0067] This application utilizes the synergistic analysis of water vapor three-dimensional structure and radar echoes to capture water vapor accumulation and rising development signals 10–30 minutes before rainfall, providing a significantly better early warning lead than traditional methods relying solely on echoes. This method is more adaptable to traditional observation blind spots (such as urban areas obscured by tall buildings and areas with weak radar coverage at long distances) and can improve the accuracy and stability of short-term urban rainstorm forecasts without requiring additional hardware, effectively supporting rapid early warning and management of urban stormwater risks.
[0068] Another embodiment of this application provides a storage medium storing a computer program, which, when executed by a processor, implements the following method steps: Step 1: Acquire multi-source data for short-term rainfall forecasting, including GNSS observation data, meteorological station surface observation data, and radar echo data; Step 2: Construct a near-surface wet layer enhancement constraint based on the GNSS observation data and the meteorological station surface observation data; Step 3: Based on the enhanced constraint of the near-surface wet layer and the multi-source data, the enhanced tomography algorithm is used to reconstruct the water vapor field within the preset three-dimensional hierarchical densified grid to obtain the target three-dimensional water vapor field that meets the preset conditions; Step 4: Perform calculations based on the target three-dimensional water vapor field to obtain the comprehensive water vapor burst index; Step 5: Based on the comprehensive water vapor burst index and the radar echo data, perform short-term rainfall prediction to obtain the short-term rainfall prediction result.
[0069] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0070] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.
[0071] The specific implementation process of the above method steps can be found in any of the above embodiments of the short-term rainfall prediction method based on three-dimensional GNSS water vapor tomography, and will not be repeated here.
[0072] This application utilizes the synergistic analysis of water vapor three-dimensional structure and radar echoes to capture water vapor accumulation and rising development signals 10–30 minutes before rainfall, providing a significantly better early warning lead than traditional methods relying solely on echoes. This method is more adaptable to traditional observation blind spots (such as urban areas obscured by tall buildings and areas with weak radar coverage at long distances) and can improve the accuracy and stability of short-term urban rainstorm forecasts without requiring additional hardware, effectively supporting rapid early warning and management of urban stormwater risks.
[0073] Another embodiment of this application provides an electronic device, which can be a server. The electronic device includes a processor, a memory, a network interface, and a database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the processor executes the program, it implements the functions or steps of a short-term rainfall prediction method based on three-dimensional GNSS water vapor tomography on the server side.
[0074] In one embodiment, an electronic device is provided, which can be a client. The electronic device includes a processor, memory, a network interface, a display screen, and an input device connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface is used to communicate with an external server via a network connection. When the processor executes the program of the electronic device, it implements the functions or steps of a client-side method for short-term rainfall prediction based on three-dimensional GNSS water vapor tomography.
[0075] Another embodiment of this application provides an electronic device, including at least a memory and a processor. The memory stores a computer program, and the processor, when executing the computer program in the memory, performs the following method steps: Step 1: Acquire multi-source data for short-term rainfall forecasting, including GNSS observation data, meteorological station surface observation data, and radar echo data; Step 2: Construct a near-surface wet layer enhancement constraint based on the GNSS observation data and the meteorological station surface observation data; Step 3: Based on the enhanced constraint of the near-surface wet layer and the multi-source data, the enhanced tomography algorithm is used to reconstruct the water vapor field within the preset three-dimensional hierarchical densified grid to obtain the target three-dimensional water vapor field that meets the preset conditions; Step 4: Perform calculations based on the target three-dimensional water vapor field to obtain the comprehensive water vapor burst index; Step 5: Based on the comprehensive water vapor burst index and the radar echo data, perform short-term rainfall prediction to obtain the short-term rainfall prediction result.
[0076] The specific implementation process of the above method steps can be found in any of the above embodiments of the short-term rainfall prediction method based on three-dimensional GNSS water vapor tomography, and will not be repeated here.
[0077] This application utilizes the synergistic analysis of water vapor three-dimensional structure and radar echoes to capture water vapor accumulation and rising development signals 10–30 minutes before rainfall, providing a significantly better early warning lead than traditional methods relying solely on echoes. This method is more adaptable to traditional observation blind spots, such as urban areas obscured by tall buildings and areas with weak radar coverage at long distances. Furthermore, it improves the accuracy and stability of short-term urban rainstorm forecasts without requiring additional hardware, effectively supporting rapid early warning and management of urban stormwater risks.
[0078] The above embodiments are merely exemplary embodiments of this application and are not intended to limit this application. Those skilled in the art can make various modifications or equivalent substitutions to this application within the scope and nature of this application, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of this application.
Claims
1. A short-duration rainfall prediction method based on three-dimensional GNSS water vapor tomography, characterized in that, include: Acquire multi-source data for short-term rainfall forecasting, including GNSS observation data, meteorological station surface observation data, and radar echo data; Near-ground wet layer enhancement constraint is constructed based on the GNSS observation data and the meteorological station surface observation data to build a preset three-dimensional layered and densified grid. Based on the enhanced constraint of the near-surface wet layer and the multi-source data, the enhanced tomography algorithm is used to reconstruct the water vapor field within the preset three-dimensional hierarchical densified grid to obtain the target three-dimensional water vapor field that meets the preset conditions. Based on the target three-dimensional water vapor field, a comprehensive water vapor burst index is obtained through calculation and processing. Short-term rainfall prediction results are obtained by performing short-term rainfall prediction based on the comprehensive water vapor burst index and the radar echo data.
2. The method as described in claim 1, characterized in that, The near-surface wet layer enhancement constraint, which constructs a preset three-dimensional layered and encrypted grid based on the GNSS observation data and the meteorological station surface observation data, specifically includes: The initial tomography matrix is obtained by performing calculations based on the GNSS observation data and the preset three-dimensional hierarchical encrypted grid. The initial wet delay observation vector is obtained by performing calculations based on the GNSS observation data and the meteorological station surface observation data. The signal-to-noise ratio in the GNSS observation data is inverted to calculate the change in wet layer thickness, thus obtaining the near-surface wet layer thickness disturbance. Based on the initial tomography matrix and the near-ground wet layer thickness disturbance, the low-altitude constraint matrix is obtained by calculation using the first function. Based on the initial wet delay observation vector and the near-ground wet layer thickness disturbance, the second function is used to calculate and process the low-altitude constraint vector. Constraint conditions are constructed based on the low-altitude constraint matrix and the low-altitude constraint vector to obtain the near-ground wet layer enhancement constraint.
3. The method as described in claim 2, characterized in that, The step of inverting the signal-to-noise ratio in the GNSS observation data to calculate the variation in wet layer thickness, and obtaining the near-surface wet layer thickness disturbance, specifically includes: The signal-to-noise ratio of data collected from different satellites within a predetermined elevation angle range is analyzed to obtain the interference frequency. The near-surface wet layer thickness disturbance is obtained by inverting the frequency change between two adjacent moments based on the interference frequency.
4. The method as described in claim 2, characterized in that, The process of reconstructing the water vapor field within the preset three-dimensional hierarchical mesh based on the enhanced constraint of the near-surface moist layer and the multi-source data using an enhanced tomography algorithm to obtain a target three-dimensional water vapor field that meets preset conditions specifically includes: Based on the initial tomography matrix and the low-altitude constraint matrix, an enhancement coefficient matrix is obtained through calculation. The enhanced observation vector is obtained by performing calculations based on the initial wet delay observation vector and the low-altitude constraint vector. Based on the enhancement coefficient matrix and the enhancement observation vector, a multi-constraint hybrid iterative method is used to reconstruct the water vapor field within the preset three-dimensional hierarchical encrypted grid, thereby obtaining the target three-dimensional water vapor field with the objective of minimizing the residual between the predicted wet delay and the observed wet delay.
5. The method as described in claim 1, characterized in that, The calculation and processing based on the target three-dimensional water vapor field to obtain the comprehensive water vapor burst index specifically includes: The precipitable water volume is calculated by integrating the three-dimensional water vapor density of the target three-dimensional water vapor field in the vertical direction at different time periods, and the rapid accumulation rate of water vapor is obtained. Spatial difference calculations are performed on the different directional dimensions of the target three-dimensional water vapor field to obtain the three-dimensional water vapor gradient. The water vapor density in the near-surface layer of the target three-dimensional water vapor field is integrated to obtain the enhanced water vapor quantity in the lower atmosphere. The vertical time difference calculation process is performed on the target three-dimensional water vapor field to obtain the vertical water vapor rise velocity. The comprehensive water vapor burst index is obtained by fusing three-dimensional water vapor information based on the rapid water vapor accumulation rate, the three-dimensional water vapor gradient, the enhanced water vapor volume at low altitudes, and the vertical water vapor rise velocity.
6. The method as described in claim 1, characterized in that, The step of performing short-term rainfall prediction based on the comprehensive water vapor burst index and the radar echo data to obtain short-term rainfall prediction results specifically includes: The water vapor burst threshold is determined based on the historical sequence of the comprehensive water vapor burst index of the target area and the actual rainfall label corresponding to the historical sequence of the comprehensive water vapor burst index. Determine the radar echo increment threshold based on historical radar echo data of the target area; Short-term rainfall prediction is performed based on the comprehensive water vapor burst index, the radar echo increment in the radar echo data of the target area acquired in real time, the water vapor burst threshold, and the radar echo increment threshold, to obtain the short-term rainfall prediction result.
7. The method as described in claim 6, characterized in that, The method of performing short-term rainfall prediction based on the comprehensive water vapor burst index, real-time acquired radar echo data of the target area, the water vapor burst threshold, and the radar echo increment threshold, to obtain short-term rainfall prediction results, specifically includes: When the time interval between the first moment when the comprehensive water vapor burst index is greater than or equal to the water vapor burst threshold and the second moment when the radar echo increment is greater than or equal to the radar echo increment threshold is less than or equal to a preset time interval threshold, a short-term heavy rainfall warning is triggered.
8. A short-term rainfall prediction device based on three-dimensional GNSS water vapor tomography, characterized in that, include: The acquisition module is used to acquire multi-source data for short-term rainfall prediction, including GNSS observation data, meteorological station surface observation data, and radar echo data. The constraint construction module is used to construct near-ground wet layer enhanced constraints of a preset three-dimensional layered densified grid based on the GNSS observation data and the meteorological station surface observation data. The reconstruction module is used to reconstruct the water vapor field within the preset three-dimensional hierarchical densified grid based on the enhanced constraint of the near-surface wet layer and the multi-source data using an enhanced tomography algorithm, so as to obtain the target three-dimensional water vapor field that meets the preset conditions. The calculation module is used to perform calculations based on the target three-dimensional water vapor field to obtain the comprehensive water vapor burst index; The prediction module is used to predict short-term rainfall based on the water vapor burst index and the radar echo data, and obtain the short-term rainfall prediction result.
9. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the steps of the short-term rainfall prediction method based on three-dimensional GNSS water vapor tomography as described in any one of claims 1-7.
10. An electronic device, characterized in that, It includes at least a memory and a processor, wherein the memory stores a computer program, and the processor, when executing the computer program in the memory, implements the steps of the short-term rainfall prediction method based on three-dimensional GNSS water vapor tomography as described in any one of claims 1-7.