Method and system for identifying and predicting convective jolt based on multi-source data fusion
By using a multi-source data fusion method, satellite and radar data are used to filter convective clouds and combined with QAR data. A convective gravity wave drag parameterization scheme is introduced, which solves the problem of insufficient accuracy in the identification and prediction of convective turbulence in the existing technology. This enables efficient identification and prediction of convective turbulence, improving flight safety and flight operation efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
- Filing Date
- 2026-02-24
- Publication Date
- 2026-05-19
AI Technical Summary
Existing technologies are not very accurate in identifying and predicting convective turbulence along routes based on single observation data, and mainstream forecasting methods are not specific enough for diagnosing and forecasting convective turbulence along routes, making it difficult to effectively identify and predict convective turbulence outside clouds, which affects flight safety and flight operation efficiency.
By using a multi-source data fusion method, convective clouds are screened using multiple observational data such as satellite cloud top temperature, lightning location, and radar reflectivity. A set of convective turbulence events along the flight path is constructed by combining QAR data. Furthermore, a parameterization scheme for convective gravity wave drag is introduced to calculate the integrated forecast index, thereby improving the accuracy of identification and prediction.
It improves the accuracy and predictive ability of convective turbulence identification, avoids identification errors caused by single observation data, enhances the targeted prediction ability of convective turbulence, and ensures flight safety and flight operation efficiency.
Smart Images
Figure CN122065254A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of flight safety and aviation meteorological support technology, specifically relating to a method and system for identifying and predicting convective turbulence based on multi-source data fusion. Background Technology
[0002] Turbulence is a significant type of en-route weather affecting aviation safety and flight efficiency. It is primarily caused by atmospheric turbulence (also known as aviation turbulence or aircraft-scale turbulence) with a horizontal scale between 0.1 and 1 km, and can be categorized into convective turbulence and clear-air turbulence. En-route convective turbulence is closely related to convective clouds and thunderstorm activity. Although most turbulence can be identified and avoided by observing visual signs and monitoring airborne radar echoes, in some cases, convective turbulence caused by the propagation and breakup of convective gravity waves can also occur in clear air far from convective clouds. Similar to clear-air turbulence, this type of convective turbulence occurring outside clouds (also known as near-cloud turbulence) is difficult to detect or effectively monitor in advance, representing a weak link in current flight operation management and posing a significant potential threat to flight quality and safety.
[0003] In practical operations, calculating various turbulence indices based on high-resolution numerical model outputs is the mainstream method for turbulence diagnosis and forecasting along flight routes. However, most current turbulence indices (such as the Elrod index and the Dutton index) are constructed based on favorable conditions for clear-sky turbulence formation (such as strong wind shear zones, airflow convergence / divergence zones, horizontal deformation zones, and areas with large horizontal temperature gradients), which are insufficient for the diagnosis and forecasting of convective turbulence along flight routes. Furthermore, existing identification methods are mostly based on single observational data such as satellite or lightning data, which can easily lead to errors or omissions in the identification of convective clouds and the convective turbulence they cause along flight routes. Considering that satellite observations have a wide coverage area, but their resolution (such as the spatial resolution of the FY-4A cloud top temperature product being 4 km) is insufficient to capture smaller-scale individual thunderstorms, while ground-based meteorological radar and lightning observation data can make up for this deficiency to some extent, this invention fully leverages the spatiotemporal complementary advantages of multi-source observation data and introduces the calculation of convective turbulence index based on the convective gravity wave drag parameterization scheme when constructing the integrated forecast index, thereby improving the identification and prediction level of convective turbulence along the route. Summary of the Invention
[0004] To address the issues of low accuracy in identifying convective turbulence along air routes based on single observation data and insufficient consideration of convective turbulence in mainstream forecasting methods, this invention provides a method and system for identifying and predicting convective turbulence based on multi-source data fusion. It filters convective clouds using multiple observational data, including satellite cloud top temperature and shape, lightning location, radar reflectivity, and echo top height. A set of convective turbulence events along the route is constructed by combining QAR data. Simultaneously, a convective turbulence index calculation based on a convective gravity wave drag parameterization scheme is introduced. An integrated forecasting index for convective turbulence along the route is constructed, with historical diagnostic results as the weighting basis, ultimately achieving accurate identification and prediction of convective turbulence along the route.
[0005] To achieve the above objectives, the present invention provides the following solution: A method for identifying and predicting convective turbulence based on multi-source data fusion, the method comprising: Step 1: Collect QAR data for multiple aircraft types along the target route. After preliminary quality control processing, calculate static temperature, vacuum velocity, three-dimensional wind speed, and vortex dissipation rate. To determine the time, location, and intensity of turbulence occurring on the target flight route; Step 2: After acquiring and quality-controlling multi-source observation data covering the target route, atmospheric convective clouds and lightning activity are identified by combining multiple thresholds. The time, location and intensity of aircraft turbulence on the target route are then combined to construct a set of convective turbulence events along the route. Step 3: Obtain reanalysis data covering the spatiotemporal range of QAR data for the target route, calculate various turbulence indices, and construct an integrated forecast index for convective turbulence of the route by combining the route convective turbulence event set; Step 4: Based on the integrated forecast index of convective turbulence along the route, access the real-time results of the operational numerical forecast model, and calculate and output the time, location, and probability forecasts of convective turbulence along the route in real time.
[0006] Preferred methods for inverting and calculating static temperature, vacuum velocity, and three-dimensional wind speed include: ; ; ; ; ; in, Atmospheric quiescent temperature Total atmospheric temperature Mach number, Vacuum speed, It is the specific heat ratio of air. It is the gas constant for dry air. The wind speed is in the zonal direction. For meridional wind speed, Vertical wind speed, The zonal component of the ground velocity. For the meridional component of ground speed, This represents the vertical component of the ground velocity. The pitch angle, Yaw angle For roll angle, Angle of attack.
[0007] Preferably, the vortex dissipation rate is calculated by inversion. The methods include: Based on the acquired vertical wind data, the power spectrum of the vertical wind speed is estimated using Fourier transform. And combined with the power spectrum of the von Karman turbulence theory model To estimate vortex dissipation rate The specific calculation formula is as follows: ; ; in, It takes the vertical wind time series of the sampling window. To represent a complex number, It is the sampling frequency of vertical wind. This represents the vertical wind data acquired within 10 seconds. ; Indicates the deviation correction term. and These represent the low-frequency cutoff frequencies. and high frequency cutoff frequency The index.
[0008] Preferably, the method for identifying atmospheric convective clouds and lightning activity by acquiring multi-source observation data covering the target flight path, performing quality control processing, and then combining multiple thresholds includes: Collect multi-source observation data covering the target flight path, specifically including: cloud top brightness temperature retrieved from satellites. Ground weather radar combined reflectivity factor and echo top height Data, as well as data from satellite lightning imagers and ground-based ADTD lightning locators; Cloud top brightness temperature retrieved from satellite Ground weather radar combined reflectivity factor and echo top height After the data undergoes quality control processing, it is processed according to... And the cloud shape is nearly round, and Threshold conditions are used to determine the time and location information of convective cloud occurrence; After quality control processing of data from satellite lightning imagers and ground-based ADTD lightning locators, the time and location information of lightning activity are determined.
[0009] Preferably, the method for acquiring reanalysis data covering the spatiotemporal range of QAR data for the target route, calculating multiple turbulence indices, and constructing an integrated forecast index for convective turbulence along the route by combining the route's convective turbulence event set includes: Reanalysis data covering the spatiotemporal range of QAR data for the target route were collected, and six convective turbulence indices, i.e., convective gravity wave towing parameterization schemes, were calculated. Minimum Richardson number diffusion coefficient diffusion coefficient Subgrid-scale turbulent kinetic energy and subgrid-scale turbulent kinetic energy And four existing turbulence indices commonly used in aircraft turbulence diagnosis and forecasting, namely the Elrod index. Dutton Index Horizontal temperature gradient index and probability forecast factor index ; Map each turbulence index to value; based on Values, calculated based on QAR data inversion for a set of convective turbulence events along the flight path. Based on this, ROC curves were plotted and the corresponding values for each turbulence index were calculated. Area, according to Area determines the aggregate weight of each turbulence index. This forms an integrated forecast index for convective turbulence along flight routes. .
[0010] Preferably, convective gravity wave drag Minimum Richardson number diffusion coefficient diffusion coefficient Subgrid-scale turbulent kinetic energy and subgrid-scale turbulent kinetic energy The calculation methods include: ; ; ; ; ; ; in, It is the vertical height. It is air density. This represents the wave stress of convective gravity waves. It is the Richardson number. It is a constant related to the background wind field, atmospheric stability, and the upper and lower limits of non-adiabatic forcing. It is a non-linear factor. Indicates horizontal phase velocity, Indicates full wind speed. It is the Brent-Vesara frequency. It is the Smygolinsky constant. It is a length scale. This represents the total deformation of the horizontal wind field. It is Prandtl number. It is the Dürdorf coefficient; Elrod Index Dutton Index Horizontal temperature gradient index and probability forecast factor index The calculation methods include: ; ; ; ; in, It is the vertical shear of horizontal winds. It is horizontal divergence; Horizontal shear representing horizontal wind. For temperature, and These represent latitude and longitude, respectively.
[0011] Preferably, each turbulence index is mapped to... The methods for determining values include: ; ; ; in, It is the first The transformed EDR mapping value It is the first The original turbulence index, , and These are mapping coefficients. The standard deviation is represented by angle brackets, and the expected value is represented by angle brackets.
[0012] Preferably, the ordinate of the ROC curve is This refers to the accuracy rate of forecasts for actual turbulence of mild intensity or higher; the horizontal axis represents the relative false alarm rate. The calculation formula is ,in It is the accuracy rate of forecasts that did not actually experience turbulence; in, and The calculation formula is: ; ; in, and These represent the number of correct and incorrect predictions of the turbulence index relative to the actual occurrence of turbulence. and These represent the number of correct and incorrect predictions of the turbulence index for cases where no turbulence actually occurred.
[0013] Preferred, according to Area determines the aggregate weight of each turbulence index. This forms an integrated forecast index for convective turbulence along flight routes. The methods include: ; ; in, It is the EDR mapping value after the various turbulence indices are converted according to the probability matching mapping relationship. It is to satisfy The number of aircraft turbulence indices under the test conditions, if the first The aircraft turbulence index corresponds to When this happens, the weighting coefficient of the index is set. After discarding that index, the weighting coefficients of the remaining turbulence indices are recalculated. and .
[0014] The present invention also provides a convective turbulence identification and prediction system based on multi-source data fusion. The system is used to implement the aforementioned method and includes: a QAR inversion and real-time identification module, a multi-source fusion and convective event construction module, a turbulence index integration and modeling module, and a real-time forecasting and dissemination module. The QAR inversion and real-time identification module is used to collect QAR data from multiple aircraft types along the target route. After preliminary quality control processing, it inverts and calculates static temperature, vacuum velocity, three-dimensional wind speed, and vortex dissipation rate. To determine the time, location, and intensity of turbulence occurring on the target flight route; The multi-source fusion and convection event construction module is used to acquire multi-source observation data covering the target route and perform quality control processing. After that, it identifies atmospheric convective clouds and lightning activity by integrating multiple thresholds, and constructs a set of convective turbulence events along the target route by combining the time, location and intensity of aircraft turbulence. The turbulence index integration and modeling module is used to acquire reanalysis data covering the spatiotemporal range of QAR data for the target route, calculate various turbulence indices, and construct an integrated forecast index for convective turbulence of the route by combining the route convective turbulence event set. The real-time forecasting and publishing module is used to calculate and output the time, location and probability forecasts of convective turbulence along the route in real time, based on the integrated forecast index of convective turbulence along the route and the results of real-time operational numerical forecasting models.
[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) Improve the ability to identify and diagnose convective turbulence on target routes: Based on high-resolution QAR data and multi-source observation data (including satellite cloud top brightness temperature and cloud shape, lightning imager and lightning locator data, and ground weather radar combined reflectivity factor and echo top height), convective clouds are screened out and convective turbulence is identified, avoiding errors or omissions in convective cloud identification caused by single observation data, and improving the accuracy of convective turbulence identification on target routes; (2) Form an integrated forecast index based on the main physical mechanism of convective turbulence to improve the targeted prediction capability of convective turbulence for target routes: On the basis of considering the commonly used diagnostic indices for existing aircraft turbulence business forecasts, a convective turbulence index calculation based on the parameterization scheme of convective gravity wave drag is introduced, and the set weights are reasonably assigned according to the historical diagnostic effects of each turbulence index, and finally an integrated forecast index of convective turbulence for routes is formed to improve the targeted prediction capability of convective turbulence for target routes. Attached Figure Description
[0016] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0017] Figure 1 This is a schematic diagram of a method for identifying and predicting convective turbulence based on multi-source data fusion, according to an embodiment of the present invention. Detailed Implementation
[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0020] Example 1 like Figure 1 As shown, this invention provides a method for identifying and predicting convective turbulence based on multi-source data fusion, the method comprising: Step 1: Collect QAR data for multiple aircraft types along the target route. After preliminary quality control processing, calculate static temperature, vacuum velocity, three-dimensional wind speed, and vortex dissipation rate. To determine the time, location, and intensity of turbulence occurring on the target flight route; Step 2: After acquiring and quality-controlling multi-source observation data covering the target route, atmospheric convective clouds and lightning activity are identified by combining multiple thresholds. The time, location and intensity of aircraft turbulence on the target route are then combined to construct a set of convective turbulence events along the route. Step 3: Obtain reanalysis data covering the spatiotemporal range of QAR data for the target route, calculate various turbulence indices, and construct an integrated forecast index for convective turbulence of the route by combining the route convective turbulence event set; Step 4: Based on the integrated forecast index of convective turbulence along the route, access the real-time results of the operational numerical forecast model, and calculate and output the time, location, and probability forecasts of convective turbulence along the route in real time.
[0021] The specific implementation process is as follows: Step 1: Target route QAR data collection, quality control and calculate Sub-step 1-A: Target route QAR data collection, quality control, and 3D wind field inversion Collect QAR data for multiple aircraft types along the target route, specifically involving the following parameters: Total Atmospheric Temperature air pressure altitude ,Mach number , ground speed zonal component of ground speed Ground speed meridional component Vertical component of ground velocity Pitch angle Yaw angle Roll angle Angle of attack Outliers and missing values were removed; based on the QAR data after quality control processing, assuming the sideslip angle was zero during the cruise phase, the atmospheric static temperature was calculated by inversion. Vacuum speed and three-dimensional wind speed (i.e., zonal wind speed) Meridional wind speed Vertical wind speed The specific calculation formula is as follows: in, It is the specific heat ratio of air (adjusted according to the target flight path altitude: 1.40 for the troposphere and 1.39 for the stratosphere). It is the gas constant for dry air.
[0022] Sub-step 1-B: Calculation and determination of turbulence events on the flight path Based on the vertical wind obtained through the above steps, the power spectrum of the vertical wind speed is estimated using Fourier transform. And combined with the power spectrum of the von Karman turbulence theory model To estimate The specific calculation formula is as follows: in, It takes the vertical wind time series of the sampling window. To represent a complex number, It is the sampling frequency of vertical wind. The vertical wind data was sampled and acquired within 10 seconds. ; This represents the deviation correction term, and 1.3 is often chosen as an empirical value. and These represent the low-frequency cutoff frequencies. and high frequency cutoff frequency index, and Generally, an empirical parameter (such as 0.2) is selected. and 0.5 ).
[0023] Based on this, the vortex dissipation rate is obtained by inversion calculation using QAR data. Combining the EDR value range for different intensities of aircraft turbulence, that is, when Between 0.15 and 0.21 The interval was judged as mild bumps, when Between 0.22 and 0.34 The interval is considered moderate turbulence. At 0.34 The above conditions are considered severe turbulence, thus determining the time, location, and intensity of turbulence occurring on the target route.
[0024] Step 2: Collection, quality control, and construction of a set of convective turbulence events from multiple sources. Collect multi-source observation data covering the target flight path, specifically including: cloud top brightness temperature retrieved from satellites. Ground weather radar combined reflectivity factor and echo top height Data, as well as data from satellite lightning imagers and ground-based ADTD lightning locators. Regarding the aforementioned satellites... After the data undergoes quality control processing (removing outliers and missing values), according to And the cloud shape is nearly circular (i.e., it satisfies the condition) Convective clouds are filtered based on a threshold condition that the cloud clusters have equidistant extension distances in both the longitudinal and latitudinal directions. Cloud clusters that meet the above conditions (using longitude) are then selected. ,latitude and time (Indicated) Columns are convective cloud sets The elements (i.e.) ,here Indicates satellite-based The total number of convective cloud clusters identified through data filtering). Similarly, after quality control processing (removing outliers and missing values) of the above radar observation data, based on... and Threshold conditions are used to filter convective clouds, and cloud clusters that meet the above conditions (using longitude) are selected. ,latitude and time (Indicated) Columns are convective cloud sets The elements (i.e.) ,here This represents the total number of convective cloud clusters selected based on radar data. After quality control processing (removing outliers and missing values) of satellite lightning imager and ground-based ADTD lightning locator data, the data is then analyzed based on lightning location and time information (using longitude). ,latitude and time (Representation), constructing an ensemble of atmospheric lightning activity. (Right now ,here This represents the total number of atmospheric lightning strikes determined from lightning data. Based on the above results, a final set of atmospheric convection activity is constructed. , can be represented as ,here It is the total amount of atmospheric convection activity determined by satellite, radar, and lightning data.
[0025] Furthermore, combining the time and location information of aircraft turbulence along the target route obtained in step one (using longitude) ,latitude and time It indicates that the subscript ,here (This refers to the total number of aircraft turbulence events along the target route determined in step one). If there is convective cloud or lightning activity within a horizontal distance of 53 km within 30 minutes before and after a certain aircraft turbulence event (i.e., it meets the following conditions), then... and If the turbulence is 50%, it is determined to be en-route convective turbulence. Based on the above conditions, all aircraft turbulence events along the target route are combined with atmospheric convection activity. Each element was compared one by one, and the aircraft turbulence that met the conditions (using longitude) was determined. ,latitude ,time and intensity (Indicated) as a set of convective turbulence The elements (i.e.) ,here Finally, the convective turbulence of the target route was completed. The construction.
[0026] Step 3: Turbulence Index Calculation, Assessment, and Construction of Integrated Forecast Index for En route convective turbulence Sub-step 3-A: Calculation of various aircraft turbulence indices and EDR mapping considering the physical mechanisms of en-route convective turbulence. Collect reanalysis data covering the target route, specifically involving the following variables: temperature. air pressure Zonal wind , direction of wind Vertical wind And calculate the six convective turbulence indices (i.e., convective gravity wave drag) based on the convective gravity wave drag parameterization scheme. Minimum Richardson number diffusion coefficient diffusion coefficient Subgrid-scale turbulent kinetic energy and subgrid-scale turbulent kinetic energy ), and the four turbulence indices commonly used in traditional aircraft turbulence diagnosis and forecasting (i.e., the Erdős index). Dutton Index Horizontal temperature gradient index and probability forecast factor index ).
[0027] The convective turbulence index is mainly derived from the convective gravity wave drag parameterization scheme, and the specific calculation formula is as follows: in, It is air density. It is the vertical height. The wave stress representing convective gravity waves is calculated using the following formula for each layer in the vertical direction: ,here It is the Brent-Vesara frequency. Indicates the horizontal grid spacing; It is a constant, related to the non-adiabatic forced horizontal structure, and its calculation formula is: ,here It is the half-width of the non-adiabatic forced part, and the estimation formula is: , This represents the cloud coverage rate at the sub-grid scale (taken as a fixed value of 0.4). The large-scale cooling required to achieve zero net heating in a viscous, non-rotating, steady-state airflow is estimated using the following formula: ; It is a constant related to the background wind field, atmospheric stability, and the upper and lower limits of non-adiabatic forcing, and is calculated using the following formula: ,here Indicates full wind speed. and These represent the top and bottom heights of the convective cloud, respectively. It is a nonlinear factor, and its calculation formula is: ,here It is gravitational acceleration. It is the atmospheric convection heating rate. It is the specific heat at constant pressure of air.
[0028] in, It is the Richardson number, calculated using the following formula: ,here This represents potential temperature, and its calculation formula is: .when At that time, the convective gravity wave breaks up. Based on the above... and The diffusion coefficients of the two can be calculated, and the specific estimation formula is as follows: in, This represents the horizontal phase velocity (assuming the convective gravity wave is stationary relative to the convective system, and is therefore taken as 0). It is the Smygolinsky constant (with a value of 0.2). It is the length scale (here set to the vertical grid spacing). The total deformation of the horizontal wind field is represented by the following formula: , and These represent latitude (or east-west direction) and longitude (or north-south direction), respectively. This is the Prandtl number (value 1). Based on the diffusion coefficient above, and further combined with the turbulence closure assumption, the subgrid-scale turbulent kinetic energy can be calculated. and .
[0029] in, It is the Dürdorf coefficient (with a value of 0.1).
[0030] In addition, to account for other favorable conditions for aircraft turbulence formation (such as strong wind shear zones, airflow convergence / divergence zones, horizontal deformation zones, and zones with large horizontal temperature gradients), further calculations were performed. , , and The index, specifically calculated using the following formula: in, It is the vertical shear of horizontal winds. It is horizontal divergence; This indicates the horizontal shear of a horizontal wind.
[0031] The above 10 aircraft turbulence indices are converted into corresponding EDR values according to the probability matching mapping relationship. The specific calculation formula is as follows: in, It is the first The transformed EDR mapping value It is the first The original turbulence index, here , and These are mapping coefficients. The calculations obtained in step 1-B are as follows: The standard deviation is represented by angle brackets, and the expected value is represented by angle brackets.
[0032] Sub-step 3-B: Evaluation of the diagnostic effectiveness of convective turbulence along the flight route and construction of integrated forecast index. Based on the set of convective turbulence events along the flight path in step two above. The values were used as the observation benchmark, and the diagnostic effectiveness of the above 10 aircraft turbulence indices was tested by plotting ROC curves. For any turbulence index, the ordinate of the ROC curve is... (That is, the accuracy of forecasting when mild or stronger turbulence actually occurs;) The value ranges from 0 to 1, with a lower false alarm rate as the value approaches 1. The horizontal axis represents the relative false alarm rate. (The calculation formula is) ,in It is the accuracy rate of forecasts that did not actually experience turbulence; The value ranges from 0 to 1; the closer the value is to 1, the lower the false alarm rate. The closer the ROC curve is to the upper left corner, the better the turbulence index distinguishes between areas where turbulence has occurred and areas where it hasn't. and The specific calculation formula is as follows: here, and These represent the number of correct and incorrect predictions of the turbulence index relative to the actual occurrence of turbulence. and These represent the number of correct and incorrect predictions of the turbulence index for cases where no turbulence actually occurred.
[0033] Furthermore, the area under the ROC curve corresponding to each turbulence index is calculated (i.e., ,here Assume the first The ROC curve corresponding to each turbulence index has... There are 1 point, and their coordinates are arranged in ascending order as follows: ),( ), ……, ( If the area under the ROC curve is such that the area under the curve can be expressed according to the trapezoidal rule, then... ),when If so, then the turbulence index is considered to have a certain predictive ability. The closer the value is to 1, the better the prediction performance. By performing normalization, the corresponding weight coefficients can be reasonably allocated. And ultimately form an integrated forecast index for convective turbulence along the route. The specific calculation formula is as follows: in, It is the EDR mapping value after the various turbulence indices are transformed according to the probability matching mapping relationship, which is calculated in step 3-A; It is to satisfy The number of aircraft turbulence indices under this test condition. It should be noted that if the... The aircraft turbulence index corresponds to When this happens, the weighting coefficient of the index is set. After discarding that index, the weighting coefficients of the remaining turbulence indices are recalculated. and .
[0034] Regarding the integrated forecast index constructed above Still based on the set of convective turbulence events along the route in step two. The value is the observation baseline, and the ROC curve is plotted and the area between it and the horizontal axis is calculated (i.e., If there is (here If so, then the integrated forecast index is considered to be... Effective; if the above conditions are not met, more target route QAR data and other multi-source observation data need to be collected to enrich the set of route convective turbulence events, and more aircraft turbulence index calculations should be considered until the conditions are met.
[0035] Sub-step 3-C: Integrate real-time operational numerical weather prediction model results and output the integrated forecast index and probability of occurrence of convective turbulence along the flight route in real time. The system accesses the output results of high-resolution real-time operational numerical weather prediction models, calculates and outputs the integrated forecast index from the above steps in real time. And based on each set member (must meet) The probability of convective turbulence occurring along the flight path is calculated and output in real time based on the value of this condition. (The calculation formula is) ,in Indicates that the members of the set satisfy The number of Indicates satisfaction The total number of members in this set (the condition). Furthermore, based on the continuously accumulating real-time observation data (including route QAR data, as well as satellite cloud top temperature, lightning location, radar reflectivity, and echo top height data), the integrated forecast index for route convective turbulence is optimized according to the above steps for every additional 2000 route convective turbulence events.
[0036] The integrated forecast index for convective turbulence constructed in this invention is combined with the output results of real-time high-resolution operational numerical weather prediction models to generate real-time location, intensity, and probability forecasts for convective turbulence along air routes, providing a powerful reference for air traffic control, flight planning, and decision-making.
[0037] Example 2 The present invention also provides a convective turbulence identification and prediction system based on multi-source data fusion. The system is used to implement the method described in Embodiment 1. The system includes: a QAR inversion and real-time identification module, a multi-source fusion and convective event construction module, a turbulence index integration and modeling module, and a real-time forecasting and dissemination module. The QAR inversion and real-time identification module is used to collect QAR data from multiple aircraft types along the target route. After preliminary quality control processing, it inverts and calculates static temperature, vacuum velocity, three-dimensional wind speed, and... To determine the time, location, and intensity of turbulence occurring on the target flight route; The multi-source fusion and convective event construction module is used to acquire multi-source observation data covering the target route and perform quality control processing. After integrating multiple thresholds, it identifies atmospheric convective clouds and lightning activity, and combines the time, location and intensity of aircraft turbulence on the target route to construct a set of convective turbulence events. The turbulence index integration and modeling module is used to acquire reanalysis data covering the spatiotemporal range of QAR data for the target route, calculate various turbulence indices, and construct an integrated forecast index for convective turbulence of the route by combining the route convective turbulence event set. The real-time forecasting and publishing module is used to integrate the forecast index of convective turbulence along the route, access the results of real-time operational numerical forecasting models, and calculate and output the time, location, and probability forecasts of convective turbulence along the route in real time.
[0038] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.
Claims
1. A method for identifying and predicting convective turbulence based on multi-source data fusion, characterized in that, The method includes: Step 1: Collect QAR data for multiple aircraft types along the target route. After preliminary quality control processing, calculate static temperature, vacuum velocity, three-dimensional wind speed, and vortex dissipation rate. To determine the time, location, and intensity of turbulence occurring on the target flight route; Step 2: After acquiring and quality-controlling multi-source observation data covering the target route, atmospheric convective clouds and lightning activity are identified by combining multiple thresholds. The time, location and intensity of aircraft turbulence on the target route are then combined to construct a set of convective turbulence events along the route. Step 3: Obtain reanalysis data covering the spatiotemporal range of QAR data for the target route, calculate various turbulence indices, and construct an integrated forecast index for convective turbulence of the route by combining the route convective turbulence event set; Step 4: Based on the integrated forecast index of convective turbulence along the route, access the real-time results of the operational numerical forecast model, and calculate and output the time, location, and probability forecasts of convective turbulence along the route in real time.
2. The method according to claim 1, characterized in that, Methods for inverting and calculating static temperature, vacuum velocity, and three-dimensional wind speed include: ； ； ； ； ； in, Atmospheric quiescent temperature Total atmospheric temperature Mach number, Vacuum speed, It is the specific heat ratio of air. It is the gas constant for dry air. The wind speed is in the zonal direction. For meridional wind speed, Vertical wind speed, The zonal component of the ground velocity. For the meridional component of ground speed, This represents the vertical component of the ground velocity. The pitch angle, Yaw angle For roll angle, Angle of attack.
3. The method according to claim 2, characterized in that, Inversion calculation of vortex dissipation rate The methods include: Based on the acquired vertical wind data, the power spectrum of the vertical wind speed is estimated using Fourier transform. And combined with the power spectrum of the von Karman turbulence theory model To estimate vortex dissipation rate The specific calculation formula is as follows: ； ； in, It takes the vertical wind time series of the sampling window. To represent a complex number, It is the sampling frequency of vertical wind. This represents the vertical wind data acquired within 10 seconds. ; Indicates the deviation correction term. and These represent the low-frequency cutoff frequencies. and high frequency cutoff frequency The index.
4. The method according to claim 1, characterized in that, After acquiring and quality-controlled multi-source observation data covering the target flight path, methods for identifying atmospheric convective clouds and lightning activity by integrating multiple thresholds include: Collect multi-source observation data covering the target flight path, specifically including: cloud top brightness temperature retrieved from satellites. Ground weather radar combined reflectivity factor and echo top height Data, as well as data from satellite lightning imagers and ground-based ADTD lightning locators; Cloud top brightness temperature retrieved from satellite Ground weather radar combined reflectivity factor and echo top height After the data undergoes quality control processing, it is processed according to... And the cloud shape is nearly round, and Threshold conditions are used to determine the time and location information of convective cloud occurrence; After quality control processing of data from satellite lightning imagers and ground-based ADTD lightning locators, the time and location information of lightning activity are determined.
5. The method according to claim 1, characterized in that, Methods for acquiring reanalysis data covering the spatiotemporal range of QAR data for target routes, calculating various turbulence indices, and constructing integrated forecast indices for convective turbulence along routes by combining these indices with a set of convective turbulence events include: Reanalysis data covering the spatiotemporal range of QAR data for the target route were collected, and six convective turbulence indices, i.e., convective gravity wave towing parameterization schemes, were calculated. Minimum Richardson number diffusion coefficient diffusion coefficient Subgrid-scale turbulent kinetic energy and subgrid-scale turbulent kinetic energy And four existing turbulence indices commonly used in aircraft turbulence diagnosis and forecasting, namely the Elrod index. Dutton Index Horizontal temperature gradient index and probability forecast factor index ; Map each turbulence index to value; based on Values, calculated based on QAR data inversion for a set of convective turbulence events along the flight path. Based on this, ROC curves were plotted and the corresponding values for each turbulence index were calculated. Area, according to Area determines the aggregate weight of each turbulence index. This forms an integrated forecast index for convective turbulence along flight routes. .
6. The method according to claim 5, characterized in that, Convective gravitational wave drag Minimum Richardson number diffusion coefficient diffusion coefficient Subgrid-scale turbulent kinetic energy and subgrid-scale turbulent kinetic energy The calculation methods include: ； ； ； ； ； ； in, It is the vertical height. It is air density. This represents the wave stress of convective gravity waves. It is the Richardson number. It is a constant related to the background wind field, atmospheric stability, and the upper and lower limits of non-adiabatic forcing. It is a non-linear factor. Indicates horizontal phase velocity, Indicates full wind speed. It is the Brent-Vesara frequency. It is the Smygolinsky constant. It is a length scale. This represents the total deformation of the horizontal wind field. It is Prandtl number. It is the Dürdorf coefficient; Elrod Index Dutton Index Horizontal temperature gradient index and probability forecast factor index The calculation methods include: ； ； ； ； in, It is the vertical shear of horizontal winds. It is horizontal divergence; Horizontal shear representing horizontal wind. For temperature, and These represent latitude and longitude, respectively.
7. The method according to claim 6, characterized in that, Map each turbulence index to The methods for determining values include: ； ； ； in, It is the first The transformed EDR mapping value It is the first The original turbulence index, , and These are mapping coefficients. The standard deviation is represented by angle brackets, and the expected value is represented by angle brackets.
8. The method according to claim 6, characterized in that, The ordinate of the ROC curve is That is, the accuracy rate of forecasting when turbulence of mild intensity or above actually occurs; The horizontal axis represents the relative false alarm rate. The calculation formula is ,in It is the accuracy rate of forecasts that did not actually experience turbulence; in, and The calculation formula is: ； ； in, and These represent the number of correct and incorrect predictions of the turbulence index relative to the actual occurrence of turbulence. and These represent the number of correct and incorrect predictions of the turbulence index for cases where no turbulence actually occurred.
9. The method according to claim 6, characterized in that, according to Area determines the aggregate weight of each turbulence index. This forms an integrated forecast index for convective turbulence along flight routes. The methods include: ； ； in, It is the EDR mapping value after the various turbulence indices are converted according to the probability matching mapping relationship. It is to satisfy The number of aircraft turbulence indices under the test conditions, if the first The aircraft turbulence index corresponds to When this happens, the weighting coefficient of the index is set. After discarding that index, the weighting coefficients of the remaining turbulence indices are recalculated. and .
10. A convective turbulence identification and prediction system based on multi-source data fusion, the system being used to implement the method described in any one of claims 1-9, characterized in that, The system includes: a QAR inversion and real-time identification module, a multi-source fusion and convection event construction module, a turbulence index integration and modeling module, and a real-time forecasting and dissemination module; The QAR inversion and real-time identification module is used to collect QAR data from multiple aircraft types along the target route. After preliminary quality control processing, it inverts and calculates static temperature, vacuum velocity, three-dimensional wind speed, and vortex dissipation rate. To determine the time, location, and intensity of turbulence occurring on the target flight route; The multi-source fusion and convection event construction module is used to acquire multi-source observation data covering the target route and perform quality control processing. After that, it identifies atmospheric convective clouds and lightning activity by integrating multiple thresholds, and constructs a set of convective turbulence events along the target route by combining the time, location and intensity of aircraft turbulence. The turbulence index integration and modeling module is used to acquire reanalysis data covering the spatiotemporal range of QAR data for the target route, calculate various turbulence indices, and construct an integrated forecast index for convective turbulence of the route by combining the route convective turbulence event set. The real-time forecasting and publishing module is used to calculate and output the time, location and probability forecasts of convective turbulence along the route in real time, based on the integrated forecast index of convective turbulence along the route and the results of real-time operational numerical forecasting models.