Three-dimensional dynamic simulation method and system based on multi-modal weather data

Abstract

The invention relates to the technical field of weather three-dimensional simulation, in particular to a three-dimensional dynamic simulation method and system based on multi-modal weather data. The method comprises the following steps: acquiring multi-modal meteorological observation data including ground observation station information, radar observation information and satellite observation information, and calculating a time difference value between observation time of each data source; taking the ratio of the time difference value to the spatial distance between the key observation points as a disturbance propagation speed; establishing a space-time correlation window based on the disturbance propagation speed, calculating a deviation correlation coefficient between the observation sources in the space-time correlation window, and judging an atmospheric state type through the deviation correlation coefficient; and mapping the atmospheric state type into a multi-source coordination parameter. According to the method, a proper range can be set according to propagation characteristics of a weather system through the space-time correlation window, observation data intensive with physical consistency is realized, and the problem of dislocation of different observation source data on timestamps or coverage ranges is effectively avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional weather simulation, and in particular to a three-dimensional dynamic simulation method and system based on multimodal weather data. Background Art

[0002] Modern meteorological monitoring systems have formed a diversified observation network consisting of ground-based automatic weather stations, weather radars, and meteorological satellites. These different types of observation equipment each offer advantages in terms of spatial coverage, temporal resolution, observational elements, and measurement accuracy. Ground-based observation stations provide high-precision point-by-point observation data, weather radars can detect precipitation and wind field structures in three dimensions, and meteorological satellites provide continuous observation capabilities over large areas. Three-dimensional dynamic simulations based on multimodal weather data represent the evolution of weather phenomena in real time in three dimensions, encompassing not only horizontal distribution but also vertical structure and temporal evolution.

[0003] Different observation sources often experience inconsistent observation times due to factors such as equipment operating cycles, data transmission delays, and geographical differences. Existing technologies typically treat this as a source of error that needs to be eliminated, employing methods such as time interpolation or forced synchronization. However, this approach often loses important meteorological dynamics. Summary of the Invention

[0004] Based on this, it is necessary for the present invention to provide a three-dimensional dynamic simulation method and system based on multimodal weather data to solve at least one of the above technical problems.

[0005] To achieve the above objectives, a three-dimensional dynamic simulation method based on multimodal weather data includes the following steps: Step S1: Acquire multimodal meteorological observation data including ground observation station information, radar observation information, and satellite observation information, and calculate the time difference between the observation times of each data source; Step S2: The ratio of the time difference to the spatial distance between the key observation points is used as the disturbance propagation speed; Step S3: Establish a spatiotemporal correlation window based on the disturbance propagation speed, calculate the deviation correlation coefficient between each observation source within the spatiotemporal correlation window, and determine the atmospheric state type based on the deviation correlation coefficient; map the atmospheric state type to a multi-source coordination parameter; Step S4: Calculate the spatial gradient consistency index of different observation sources in the same area, and mark the observation abnormal area when the gradient direction or amplitude diverges; Step S5: Optimize the geometric complementary weights of the multimodal meteorological observation data using multi-source coordination parameters and observation anomaly areas, generate a three-dimensional meteorological field, and realize visualization output.

[0006] The present invention further provides a three-dimensional dynamic simulation system based on multimodal weather data, which is used to execute the above-mentioned three-dimensional dynamic simulation method based on multimodal weather data. The three-dimensional dynamic simulation system based on multimodal weather data includes: Multi-source data acquisition module, used to obtain multi-modal meteorological observation data including ground observation station information, radar observation information and satellite observation information, and calculate the time difference between the observation time of each data source; The propagation speed calculation module is used to take the ratio of the time difference to the spatial distance between the key observation points as the disturbance propagation speed; The state recognition weight module is used to establish a spatiotemporal correlation window based on the disturbance propagation speed, calculate the deviation correlation coefficient between each observation source within the spatiotemporal correlation window, and determine the atmospheric state type based on the deviation correlation coefficient; the atmospheric state type is mapped to a multi-source coordination parameter; The abnormal region detection module is used to calculate the spatial gradient consistency index of different observation sources in the same area, and mark the observation abnormal region when the gradient direction or amplitude diverges; The three-dimensional dynamic reconstruction module is used to optimize the geometric complementary weights of multi-modal meteorological observation data using multi-source coordination parameters and observation anomaly areas, generate a three-dimensional meteorological field and realize visual output.

[0007] The present invention significantly improves the accuracy and reliability of multimodal meteorological data fusion by adjusting the fusion weight coefficients based on observed anomaly areas. During actual observations, due to differences in sensor equipment, observation time, and spatial resolution between different observation sources, inconsistent or abnormal data often occur in some areas. By reducing the weight of the corresponding observation source in the abnormal area, the system effectively weakens the negative impact of abnormal observation data on the fusion results, avoiding misjudgments and information distortion caused by the anomaly of a single data source, thereby ensuring the scientific nature and stability of the fused data. At the same time, normal areas maintain their original weights, fully leveraging the advantages of the data from each observation source and achieving complementary advantages. This dynamic weight adjustment mechanism enhances the adaptability and robustness of multi-source data fusion, enabling the fusion model to adapt to complex and changing meteorological environments. The adjusted fusion weights are used to perform weighted fusion on multimodal meteorological observation data. The fused data comprehensively reflects the spatial and temporal information of multi-source observations such as ground stations, radars, and satellites, covering the multi-level characteristics of point, surface, and volume observations, effectively compensating for the shortcomings of a single data source. Through weighted fusion, the spatial coverage and accuracy characteristics of data from different observation sources are reasonably balanced, which improves the integrity and spatial continuity of meteorological field data. The fused data can more realistically reflect the detailed characteristics of atmospheric physical processes, which is helpful for subsequent meteorological analysis and forecast accuracy improvement. A three-dimensional grid structure is constructed based on the spatial range of the spatiotemporal correlation window, and the fused meteorological data is interpolated to the grid nodes, which greatly improves the regularization and systematicness of the data. Regularized three-dimensional meteorological field data not only facilitates computer processing and model input, but also improves the operability and application scope of the data. Discrete observation data are converted into a continuous three-dimensional field through interpolation algorithms, achieving smooth transitions and detailed representation in space, making meteorological field data more scientific and representative of observations. This three-dimensional grid structure can cover complex terrain and multi-scale meteorological phenomena, laying a solid foundation for refined meteorological simulation and analysis. Overall, this step builds a complete multimodal meteorological data fusion and three-dimensional dynamic visualization system. Through the dynamic adjustment of fusion weights, it effectively resists the interference of abnormal data and ensures data quality and fusion accuracy; through three-dimensional grid construction and interpolation, it realizes the continuity and regularization of observation data and enhances the scientific expression ability of data; through advanced volume rendering technology, it improves the display effect and analysis convenience of meteorological information. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments thereof made with reference to the following drawings: Figure 1 A schematic flow chart of the steps of a three-dimensional dynamic simulation method based on multimodal weather data according to the present invention; Figure 2This is a module diagram of a three-dimensional dynamic simulation system based on multimodal weather data according to the present invention; Figure 3 Schematic diagram of establishing a spatiotemporal correlation window based on the disturbance propagation speed in the present invention; Figure 4 This is a decision flow chart for judging the atmospheric state type through the deviation correlation coefficient. DETAILED DESCRIPTION

[0009] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. It is obvious that the embodiments described are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts are within the scope of protection of the present invention.

[0010] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0011] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0012] To achieve this, please refer to Figures 1 to 4 The present invention provides a three-dimensional dynamic simulation method based on multimodal weather data, the method comprising the following steps: Step S1: Acquire multimodal meteorological observation data including ground observation station information, radar observation information, and satellite observation information, and calculate the time difference between the observation times of each data source; Furthermore, step S1 includes the following steps: At the same time, it receives point-like meteorological data from ground automatic weather stations, spatial scanning data from weather radars, and spatial scanning data from meteorological satellites as original observation data sets; Convert the original observation data set into a standardized format to obtain multimodal meteorological observation data; Calculate the time difference between different data sources in the multimodal meteorological observation data within the same observation period, where the time difference includes the time difference between ground station and satellite observation, the time difference between ground station and radar observation, and the time difference between radar and satellite observation.

[0013] In some embodiments, point meteorological data collected by ground automatic weather stations, multi-elevation volumetric spatial scanning data obtained by meteorological radars, and high-resolution visible light / infrared remote sensing image data downlinked by meteorological satellites. Since the three types of data have significant differences in collection methods, spatial resolutions, and temporal resolutions, they need to be formatted and standardized before entering the fusion calculation process. Specifically, the temperature, humidity, air pressure, wind speed and other elements recorded hourly or minute by hour by the ground meteorological station are constructed as observation point data with longitude and latitude marks; the radar reflectivity factor (Z), radial velocity (V) and other data of the meteorological radar are resampled to a unified spatial grid structure, and the scanning time and corresponding elevation angle information are marked; the pixel brightness temperature value or reflectivity data in the meteorological satellite data is projected into a geographic reference format according to the orbital parameters and observation time, and a two-dimensional layer consisting of pixel position and timestamp is constructed. After the above preprocessing is completed, a multimodal meteorological observation dataset with a unified structure is obtained, which has the ability to describe spatiotemporal labels and physical elements. The time synchronization between the three types of observation sources is quantified within the same observation period. To this end, it is necessary to calculate separately: (1) The time difference between ground station and satellite observations; (2) The time difference between the ground station and radar observations; (3) The time difference between radar and satellite observations.

[0014] The specific implementation method is as follows: extract the radar scanning point closest to each ground automatic weather station from the meteorological radar scanning data, and record the scanning timestamp of the scanning point; extract the satellite pixels covering the location of each ground station from the satellite remote sensing image, and record its image acquisition timestamp; compare the above two timestamps with the observation timestamp of the ground station respectively to obtain the ground-radar time difference and the ground-satellite time difference respectively.

[0015] Furthermore, a target area containing several observation points is selected, and the average observation time of all coverage points of the meteorological radar and the average observation time of all satellite pixels are calculated in this area. The difference between the two is the radar-satellite time difference of this area.

[0016] It should be noted that due to the fact that radar scanning is usually carried out in a body scan with a period of several minutes, and the satellite transit frequency is also affected by the orbit control, there is natural asynchrony in the observation time. In order to improve the calculation accuracy of the subsequent disturbance propagation speed, the combination of observation points with the smallest time difference is preferentially selected for the estimation of the propagation speed, and the observation data of all types is unified to a high-precision time stamp system (such as UTC) to avoid the distortion of the difference value caused by time zone conversion or time format error.

[0017] Further, the calculation of the time difference between different data sources in the multi-modal meteorological observation data in the same observation period comprises: extracting the effective scanning point closest to the spatial position of the target ground station from the spatial scanning data of the meteorological radar; comparing the actual observation time stamp of the target ground station with the scanning time stamp of the effective scanning point, calculating the absolute difference value as the ground station-radar time difference of the target ground station; According to the pixel unit information corresponding to the geographical position of the target ground station in the spatial scanning data of the meteorological satellite, the accurate observation time stamp of the pixel unit is read to obtain the pixel time stamp; Comparing the pixel time stamp with the actual observation time stamp of the target ground station, the absolute time interval between the two time stamps is taken as the ground station-satellite time difference of the target ground station; Based on the target area range, the time difference between the meteorological radar and the meteorological satellite observation data is calculated to obtain the radar-satellite time difference of the target area; The ground station-radar time difference, ground station-satellite time difference and radar-satellite time difference are merged as the time difference value.

[0018] In some embodiments, a certain target ground automatic weather station is taken as a reference object, and the effective radar scanning point closest to the position of the ground station is searched in the spatial scanning data of the meteorological radar starting from the known latitude and longitude coordinates of the ground station. The effective scanning point usually refers to the point with reflectivity or velocity observation data in the vertical height range, and the data on the lowest elevation section is preferentially selected to reduce height interference. After positioning to the nearest point, the time stamp information (i.e. the specific time of this radar scanning at this position) of the scanning point is extracted and compared with the observation time stamp recorded by the target ground station, the absolute time difference between the two is calculated to obtain the ground station-radar time difference of the station. This time difference reflects the degree of time offset of the two types of sensors when observing the same space region.

[0019] From the meteorological satellite's spatial scan data, the satellite pixel unit corresponding to the geographic coordinates of the target ground station is extracted. The row and column numbers or actual coverage locations of these pixel units can be obtained by inverse calculation using geographic projection. In most polar-orbiting or geostationary meteorological satellites, each pixel typically carries a unique observation timestamp, representing the precise moment the location was scanned. This pixel timestamp is then compared with the observation timestamp of the target ground station, and the absolute time interval is calculated to obtain the ground station-satellite time difference for that ground station.

[0020] To obtain a more regionally representative estimate of the time difference between radar and satellite, a target region must be constructed. This target region can be defined as a spatial grid cell containing several observation points or as the area covered by a known weather system. Within this region, the observation times of all relevant scan points are extracted from the weather radar data, and their average value is calculated as the "regional radar average time." Simultaneously, the observation timestamps of all satellite pixels covering the region are extracted from the weather satellite imagery, and their average value is calculated as the "regional satellite average time." The absolute time difference between these two values ​​constitutes the radar-satellite time difference for the target region.

[0021] After the three time differences are calculated separately, they are combined to form a complete set of time difference values, corresponding to: (1) Ground station-radar time difference; (2) Ground station-satellite time difference; (3) Radar-satellite time difference.

[0022] This time difference set will serve as an important input for the subsequent calculation of disturbance propagation speed and the construction of spatiotemporal correlation windows.

[0023] It should be noted that in practice, data update intervals vary between different observation sources. Radar scans may occur every 5-10 minutes, satellite data acquisition intervals may be as high as 15-30 minutes, and ground stations typically update data at a rate of 1 minute or more. To minimize the propagation of time errors, pairs of observations with the smallest time intervals are prioritized for pairing calculations, and all timestamps are uniformly converted to Coordinated Universal Time (UTC) or GPS time to avoid errors caused by inconsistent time zones or formats. Furthermore, in areas with extreme weather conditions or data gaps, some observation points may not have corresponding radar or satellite observations. In these cases, these points can be skipped or interpolated using surrounding observations.

[0024] Furthermore, the calculation of the time difference between the meteorological radar and meteorological satellite observation data based on the target area range includes: Determine a target area, wherein the target area is specifically a spatial grid unit or a weather characteristic area; The average time of all scanning points constituting the observation of the target area is extracted from the spatial scanning data of the weather radar, denoted as the area radar average time; The average time of all satellite pixels covering the target area is extracted from the spatial scanning data of the weather satellite, denoted as the area satellite average time; The absolute difference between the area radar average time and the area satellite average time is calculated to obtain the radar-satellite time difference of the target area.

[0025] In some embodiments, to further enhance the registration accuracy between different observation sources in the time dimension, the time difference between the weather radar and weather satellite observation data needs to be calculated based on the range of the target area. This step mainly targets the two data sources with spatial scanning characteristics, i.e., weather radar and weather satellite, extracts their observation time characteristics in a certain common area, and thus quantifies the degree of time asynchronization when observing the same spatial target.

[0026] The spatial range of the target area needs to be determined, which can be flexibly determined according to specific business requirements or weather process characteristics. Common methods include: selecting one or more grid cells as the target area in a spatial grid cell divided at a fixed resolution (such as a 1 km x 1 km grid); extracting a continuous pixel area corresponding to the spatial boundary of a weather system (such as a convective cell, frontal zone, or typhoon cloud system) as the target weather feature area. After determining the area boundary, all observation points corresponding to the area can be retrieved in the radar and satellite data.

[0027] According to the geographic projection parameters of the radar volume scanning data, determine which radar return cell (such as reflectivity pixel) falls within the target area, and collect the timestamps of these scanning points. It should be noted that, since the radar forms three-dimensional volume scanning data by scanning at different elevations, the scanning time at different elevations may differ by several seconds to tens of seconds, so in this step, the scanning time at the same elevation (such as the lowest elevation) is usually selected, or the scanning points on a certain fixed elevation section are selected to maintain time consistency. The timestamps of all valid scanning points are averaged to obtain the radar average time of the area.

[0028] Simultaneously, all satellite pixel units covering the target area are extracted from the spatial scanning data of the weather satellite. Satellite observations have scanning strip or scanning line structures, and the observation time of different pixels changes as the scanning trajectory advances line by line. Through geographic projection or orbit registration, the target area is corresponded to the pixel area in the satellite image, and the observation timestamps of these pixels are extracted. Similarly, the timestamps of all covering pixels are averaged to obtain the satellite average time.

[0029] The radar-satellite time difference within the target area is calculated by subtracting the regional radar average time from the regional satellite average time and taking its absolute value. This value reflects the temporal asynchrony between the radar and satellite when observing the area and is a key input parameter for subsequent disturbance propagation velocity estimation and establishment of the spatiotemporal correlation window.

[0030] It should be noted that due to differences in observation frequency and scanning mechanisms between radar and satellites, the size of the selected region can affect the accuracy of the time difference calculation. A larger region may include points with longer scanning time spans, resulting in a shift in the average value; while a too small region may result in unstable results due to an insufficient number of observation points. Therefore, in practical applications, an appropriate region size should be selected based on the resolution and temporal distribution of the observation data, and the observation timestamps should be appropriately weighted or filtered to improve the representativeness and robustness of the regional time difference estimate.

[0031] Step S2: The ratio of the time difference to the spatial distance between the key observation points is used as the disturbance propagation speed; Furthermore, step S2 includes: Select two different types of observation points with the smallest observation time difference from the multimodal meteorological observation data as key observation point pairs, including the combination of ground station and radar station, the combination of ground station and satellite pixel, or the combination of radar station and satellite pixel; Calculate the actual spatial distance between key observation point pairs; Determine the propagation time based on the key observation point pairs and time differences. When there are multiple time differences, the time difference with the smallest value is selected as the propagation time. The actual spatial distance between the key observation point pairs is divided by the propagation time to obtain the dynamic propagation velocity.

[0032] In some embodiments, from the standardized and time-stamped multi-modal meteorological observation data, two different types of observation points with the smallest time difference are selected to form a "key observation point pair". The point pair can be derived from one of the following three combinations: a ground automatic weather station and a weather radar scanning point; a ground automatic weather station and a weather satellite pixel; a weather radar scanning point and a weather satellite pixel. In actual implementation, the system will calculate the time stamp difference of all possible multi-source observation point pairs, and preferentially select the heterogenous observation point pair with the smallest absolute time difference as the key observation point pair, in order to improve the closeness of the disturbance occurrence time and enhance the credibility of the propagation speed estimation. The spatial distance between the key observation point pair is calculated. The spatial distance here refers to the actual distance between the two observation points in three-dimensional geographical space, which can be calculated as follows: for ground stations, satellite pixels, radar scanning points, etc., the spherical distance (great circle distance) between the latitudes and longitudes is calculated using the WGS84 ellipsoid model, and the three-dimensional spatial distance is estimated by combining the vertical height difference. For example, if the radar scanning point is at an altitude of 2 km and the ground station is on the ground, the horizontal distance and the height distance need to be combined to obtain a more realistic spatial path length. Combined with the time stamp information of the observation point pair, the time span of the disturbance propagation, i.e. the propagation time, is determined. When two observation points record the same or related meteorological disturbances (such as the same wind field change or cloud movement) at similar times, the time difference between them is the propagation time. In some cases, the system may detect multiple time difference values (for example, multiple radar elevation scanning points and satellite pixel time alignment combinations), in which case the smallest time difference should be selected as the propagation time, because the smaller the time difference, the more likely it is that the observation is of the same disturbance event in space, rather than an error caused by the superposition of multiple events.

[0033] The spatial distance calculated above is divided by the propagation time to obtain the propagation speed of the disturbance. This speed value represents the average propagation rate of meteorological disturbances (such as wind field mutations, echo movements, cloud drift, etc.) under certain spatio-temporal conditions, with a unit of meters per second (m / s) or kilometers per hour (km / h), which will be used as a key reference factor when constructing the spatio-temporal correlation window.

[0034] It should be noted that, in order to avoid speed anomalies caused by observation errors or differences in resolution of different types of data, a physically reasonable range threshold (such as 5-300 m / s) can be set for the propagation speed in the actual system, and observation point pairs outside this range can be excluded or a sub-optimal key observation point pair can be selected for calculation. At the same time, if the meteorological elements recorded by the two observation points have essential differences (such as one being wind speed and the other being brightness temperature), the correlation between them in the disturbance response needs to be confirmed through a feature matching algorithm before they can be used as an effective key observation point pair.

[0035] It is particularly important that the actual spatial distance between the key observation point pairs is calculated as follows: For the combination of a ground station and a radar station, the spherical distance between the geographical coordinates of the ground station and the geographical coordinates of the radar station is calculated as the actual spatial distance; For a combination of ground station and satellite pixels, the actual spatial distance is calculated as the distance between the geographical coordinates of the ground station and the position of the ground station under the satellite at the moment of satellite transit; For the combination of radar station and satellite pixels, the distance between the geographic coordinates of the radar station and the coordinates of the geometric center point of the target area is calculated as the actual spatial distance, where the geometric center point of the target area is the center position of the overlapping part of the radar scanning range and the satellite observation area.

[0036] In some embodiments, in order to accurately estimate the propagation speed of meteorological disturbances between different observation platforms, a particularly important step is to accurately calculate the actual spatial distance between pairs of key observation points. Since multimodal observation data comes from different observation platforms - including ground-based automatic weather stations, weather radars, and meteorological satellites, there are significant differences in spatial distribution, observation angles, and observation area ranges among the three. Therefore, it is necessary to use corresponding distance calculation methods based on different types of key observation point combinations to ensure that the distance estimation has physical meaning and geographical accuracy. Specifically, there are three situations: For a ground station and radar station combination, since both are located on the ground, or the radar station is at a relatively low tower altitude, their spatial locations can be simplified to two points on a sphere. In this case, the Great-Circle Distance model is used to calculate the spherical distance between the two points using the Haversine or Vincenty formulas based on their latitude and longitude coordinates. This is the actual spatial distance of the combination. For example, if the coordinates of ground station A are (30.00°N, 114.00°E) and the coordinates of radar station B are (30.05°N, 114.10°E), the spherical distance between the two is approximately 11.2 km, which is the spatial distance of this key observation point pair.

[0037] The combination of ground stations and satellite pixels requires consideration of the unique characteristics of satellite observations. Since satellites are orbiting platforms, their observations are vertically or obliquely directed toward the Earth, while the ground station is actually located at the geographic location corresponding to a pixel in the observed image. Therefore, to maintain a consistent calculation standard, this step uses the "sub-satellite point position of the ground station at the moment of satellite transit" as the geographic location of the corresponding satellite point. The sub-satellite point refers to the position of the satellite directly above the Earth's surface at a given moment and is the projection of the satellite's trajectory onto the Earth's surface. By consulting satellite orbital data (such as TLE data) and imaging parameters, the latitude and longitude of the sub-satellite point can be accurately determined. The spherical distance between this point and the ground station's latitude and longitude is then calculated as the actual spatial distance between the ground station and the satellite pixel. For example, if the ground station is located at (30.00°N, 114.00°E) and the sub-satellite point at the corresponding moment is (30.01°N, 113.98°E), the distance is approximately 2.6 km.

[0038] For the combination of radar station and satellite pixels, since both observe a wide area and have different viewing angles, this step uses a regional center point as the spatial reference point to minimize error. The specific method is to first determine the spatial overlap between the current radar scan range and the satellite imagery—the intersection of their observation coverage. Then, from this overlap region, extract the latitude and longitude coordinates of all valid radar echo pixels and satellite imagery pixels and calculate their geometric center point (the latitude and longitude coordinates of the center of the overlap region). Finally, the spherical distance between this center point and the fixed geographic location of the radar station is used as the actual spatial distance. For example, if the radar station coordinates are (30.00°N, 114.00°E) and the center point of the overlap region is (30.10°N, 114.05°E), the calculated spherical distance is 12.6 km.

[0039] It should be noted that there may be resolution differences or geometric distortion between satellite pixels and radar echo units. Therefore, when determining the geometric center point, the registered geographic coordinate data should be used, and invalid observations (such as cloud top obscuration and strongly attenuated radar points) should be eliminated to improve the physical representativeness of the center point. In addition, for geostationary satellites, their sub-satellite point is basically fixed above the equator, and the spatial projection method is different from that of low-orbit polar satellites. Therefore, in the distance calculation, the satellite type should be treated separately to avoid estimation errors caused by different altitudes and scanning modes.

[0040] Step S3: Establish a spatiotemporal correlation window based on the disturbance propagation speed, calculate the deviation correlation coefficient between each observation source within the spatiotemporal correlation window, and determine the atmospheric state type based on the deviation correlation coefficient; map the atmospheric state type to a multi-source coordination parameter; Furthermore, in step S3, establishing a spatiotemporal correlation window based on the disturbance propagation speed includes: The geometric midpoint of the key observation point pair is taken as the spatial center coordinate, and the average observation time of the two observation points is taken as the time center coordinate; The preset reference time constant is divided by the disturbance propagation speed to obtain a time radius value, and the disturbance propagation speed is multiplied by the time radius value to obtain a space radius value; The azimuth angle of the line between the key observation point pair is calculated as the main propagation direction angle; When the disturbance propagation speed is greater than the preset speed threshold, it is determined that it is a directional propagation phenomenon, and the elliptical shape parameter is set; when the disturbance propagation speed is less than or equal to the preset speed threshold, it is determined that it is a diffusion propagation phenomenon, and the circular shape parameter is set, to obtain the geometric shape parameter; The spatial center coordinate and the time center coordinate are taken as the positioning center of the space-time correlation window, and the space radius value and the time radius value are used to determine the spatial boundary range and the time extension range of the window respectively, and the circular or elliptical structure of the window is set according to the geometric shape parameter, and when it is an ellipse, the main propagation direction angle is used to determine the direction of the long axis of the ellipse, to obtain the space-time correlation window.

[0041] In some embodiments, a space-time correlation window for identifying the atmospheric disturbance region of space-time consistency among multi-source observation data is constructed; thereby identifying the atmospheric state type in the window, and further mapping it to the weight coefficient of multi-source data fusion. This step not only needs to refer to the space-time position of the key observation point pair, but also needs to comprehensively consider the disturbance propagation speed, propagation form and other elements to form a dynamic correlation window with physical meaning. The specific implementation method includes the following: The geometric midpoint of the key observation point pair is taken as the spatial center coordinate, and the average observation time of the two observation points is taken as the time center coordinate, to ensure the symmetry of the window area to the disturbance propagation path. The geometric midpoint is obtained by calculating the average of the longitude and latitude of the two points, representing the middle line position of the disturbance propagation between the current two observation points. At the same time, the average observation time of the two observation points is taken as the time center coordinate, which is used to determine the center reference point of the disturbance on the time axis. This combination constitutes the "anchor point" of the space-time window. According to the disturbance propagation speed and a preset time reference constant (for example, 300 seconds), the time and space scales of the window are calculated. Specifically, the time reference constant is divided by the disturbance propagation speed to obtain a time radius value (i.e. the time coverage range of the window); then the disturbance propagation speed is multiplied by the time radius to obtain a space radius value (i.e. the maximum distance that the disturbance can propagate within the time period). For example, if the disturbance propagation speed is 50 m / s and the time reference constant is 300 seconds, the time radius is 6 minutes and the space radius is 15 km. The azimuth angle of the line between the key observation point pair is calculated, which is the geographical direction angle from one point to another point (usually starting from 0° north clockwise to 360°), which represents the main propagation direction of the disturbance. The spherical triangle algorithm or simplified plane triangle calculation can be used, depending on the map projection method used.

[0042] The propagation pattern of the disturbance is determined by its velocity, and the geometric shape of the window is determined accordingly. When the disturbance propagation velocity exceeds a certain threshold (e.g., 30 m / s), the disturbance is considered to have a clear directionality and is classified as a "directional propagation phenomenon." In this case, an elliptical window is generated with the main propagation direction as the major axis. Conversely, when the disturbance velocity is less than or equal to the threshold, the disturbance is considered to have a non-directional expansion characteristic and is classified as a "diffuse propagation phenomenon." In this case, a circular window is generated with the spatial radius as the radius. The major axis of the ellipse is set to the spatial radius, and the minor axis can be set to a certain ratio of the major axis (e.g., 0.6–0.8 times) to reflect the asymmetry of the propagation.

[0043] Finally, using the previously obtained spatial center coordinates, temporal center coordinates, temporal radius, spatial radius, main propagation direction angle, and geometric shape parameters (ellipse or circle), a three-dimensional spatiotemporal correlation window is generated. This window describes the impact range of the disturbance within a specific time and space range. This window will serve as the analysis region in subsequent steps to filter various observation data points falling within this range and calculate their relative deviation correlations, thereby identifying the atmospheric state in which the disturbance is located (such as frontal, convective development, stratus advection, etc.).

[0044] See Figure 3 , the figure shows in detail the technical solution for establishing a spatiotemporal correlation window based on the propagation speed of disturbances. The figure uses a two-dimensional coordinate system, where the X-axis represents the spatial distance (unit: km), the Y-axis represents the time (unit: minute), and the coordinate origin is located in the lower left corner. The center of the key observation point pair is marked at the center of the coordinate system (coordinate point 230,300). The center point is determined by calculating the geometric midpoint of the key observation point pair as the spatial center coordinate and the average observation time of the two observation points as the time center coordinate. The figure shows two different spatiotemporal correlation window forms: The first is a diffusion propagation window, represented by a solid circle. This circular window has a radius of 2 km and is applicable when the disturbance propagation velocity is less than or equal to 30 m / s. When a diffusion propagation phenomenon is determined, the system sets the circular shape parameters to indicate that the disturbance has a non-directional expansion characteristic.

[0045] The second type is a directional propagation window, represented by a dashed ellipse. This ellipse has a major axis of 4 km and a minor axis of 2 km, with the major axis pointing northeast (shown at a 45-degree angle in the figure). It is suitable for situations where the disturbance propagation speed is greater than 30 m / s. When a directional propagation phenomenon is determined, the system sets the ellipse's shape parameters so that the major axis of the ellipse aligns with the main propagation direction.

[0046] The dashed line in the figure shows the relationship between the spatial radius and the temporal radius. The spatial radius is calculated by multiplying the disturbance propagation speed by the temporal radius. The temporal radius is calculated by dividing a preset reference time constant (e.g., 300 seconds) by the disturbance propagation speed. In this example, the spatial radius is 2 km, corresponding to a temporal radius of 2 minutes.

[0047] The main propagation direction is determined by calculating the azimuth of the line connecting the key observation points, which is represented by a straight line with an arrow in the figure and points to the northeast. This direction angle is used to determine the direction of the major axis of the elliptical window.

[0048] By establishing this spatiotemporal correlation window, the system can extract various observational data within the window and perform matching analysis with numerical forecast background data, providing spatial and temporal constraints for subsequent calculation of deviation correlation coefficients and determination of atmospheric state types. This window design fully considers the physical propagation characteristics of meteorological disturbances, ensuring the scientific and accurate spatiotemporal consistency analysis of multi-source observational data.

[0049] It should be noted that when the disturbance propagation speed is low, even if the calculated spatial radius is small, the system can set a minimum spatial radius threshold (such as 1km) to avoid the window range being insufficient to cover the effective observation area due to too low a speed; in addition, for nonlinear trajectory disturbances (such as squall lines and vortices), multiple overlapping windows can be constructed through multiple sets of key point pairs to improve the stability and coverage of state recognition.

[0050] Furthermore, in step S3, calculating the deviation correlation coefficient between each observation source within the spatiotemporal correlation window includes: Extract various types of observation data within the spatiotemporal correlation window and obtain the numerical forecast background field data corresponding to the window area; Match various observation data with the background field grid points in the numerical forecast background field data, and use the nearest neighbor method to determine the background field value corresponding to each observation point to form an observation-background field matching data set; Based on the observation-background field matching data set, the deviation value sequence of each observation source and the background field is calculated, where the deviation value sequence includes ground deviation sequence, radar deviation sequence and satellite deviation sequence; The deviation correlation coefficients between observation sources are calculated through the deviation value sequence, and the deviation correlation coefficients between the ground deviation sequence and the radar deviation sequence, the deviation correlation coefficients between the ground deviation sequence and the satellite deviation sequence, and the deviation correlation coefficients between the radar deviation sequence and the satellite deviation sequence are calculated respectively.

[0051] In some embodiments, during a spring severe convective weather monitoring event, the system collected temperature and humidity observations (at 10-minute intervals) from ground-based automatic weather stations, reflectivity products from weather radar (updated every 2 minutes), and cloud top brightness temperature data from geostationary satellites (updated every 15 minutes) within an established 50 km × 50 km spatiotemporal correlation window. Simultaneously, numerical forecast background field data (with a horizontal resolution of 5 km) for the corresponding time period was retrieved from the WRF model. To achieve point-by-point matching between the observations and the background field, the system used the nearest neighbor method. For example, if a weather station is located at 120.25° east longitude and 31.55° north latitude, the nearest grid point in the background field grid is (120.20°, 31.60°). The system directly uses the forecast temperature at this grid point as the background reference value for that station. Thus, each record contains the observation location, time, observation value, and corresponding background value. The system then calculates the deviation: deviation = observation value – background value. For example, if the actual temperature measured at the weather station is 22.8°C and the background value is 21.4°C, the temperature deviation is +1.4°C. Similarly, the deviations of radar reflectivity, satellite brightness temperature and other factors are calculated one by one.

[0052] Finally, the Pearson correlation coefficient is used to calculate the linear correlation between the three types of deviation series: Pearson correlation coefficient between the ground bias series and the radar bias series; Pearson correlation coefficient between the ground bias series and the satellite bias series; Pearson correlation coefficient between the radar bias series and the satellite bias series.

[0053] These bias correlation coefficients can be used to determine whether the observation sources are observing the same disturbance in the region, whether the disturbance is consistent with the background error, and whether the disturbance has systematic characteristics. Furthermore, the bias correlation coefficients can be used to classify the atmospheric state type in the current region (such as front, convective activity, convergence zone, background steady state, etc.), and accordingly map different multi-source coordination parameters in subsequent steps.

[0054] It should be noted that: in order to avoid the impact of insufficient sample number on the stability of the Pearson correlation coefficient results, the system can set a minimum sample number threshold (such as more than 10 valid matching points) as a prerequisite for calculating the deviation correlation coefficient; in addition, the "missing points" or "obscured values" in the radar or satellite need to be eliminated or interpolated before calculation to prevent distortion of the correlation results.

[0055] Furthermore, determining the atmospheric state type by using the deviation correlation coefficient in step S3 includes: Check the positive and negative correlation coefficients of the ground-radar bias, the ground-satellite bias, and the radar-satellite bias, and record the sign combination pattern of the three correlation coefficients; When the sign combination mode shows that all three correlation coefficients are positive, it is determined as a stratiform stable atmospheric state; when the sign combination mode shows that the bias correlation coefficient between the ground and the radar is negative and the other two are positive, it is determined as a convective development atmospheric state; when the sign combination mode shows two positive and one negative or one positive and two negative, it is determined as a frontal passage atmospheric state, to obtain a basic state type; The bias correlation coefficient is divided into three levels of strong correlation, medium correlation and weak correlation according to a preset threshold rule; when the intensity level of the bias correlation coefficient is strong correlation, it is determined as a stable state; when the intensity level is weak correlation, it is determined as a transition state; and when the intensity level difference is large, it is determined as a composite state, to obtain a stability feature; The basic state type and the stability feature are combined, and an atmospheric state type is output according to a preset corresponding rule.

[0056] In some embodiments, the bias correlation coefficient of the multi-modal observation source is used to identify the atmospheric state, which is a key link to realize fine classification of complex atmospheric processes and reasonable allocation of fusion weights. Based on the sign (positivity or negativity) and numerical intensity of the bias correlation coefficient between the three main observation sources, different basic atmospheric state types and their stability features are distinguished by a preset discrimination rule, and finally an accurate and physically meaningful atmospheric state type is output.

[0057] Referring to Figure 4 The figure details the complete decision-making process of determining the atmospheric state type by the bias correlation coefficient. The tree structure layout from top to bottom is adopted in the figure, which clearly presents all the judgment logic from the sign check to the final state type output.

[0058] Figure 4The top node is the sign check node for the bias correlation coefficients. This node checks the positive and negative values ​​of three key parameters: the ground-radar bias correlation coefficient, the ground-satellite bias correlation coefficient, and the radar-satellite bias correlation coefficient. It also records the sign combination pattern of the three correlation coefficients. This step is fundamental to determining the atmospheric state type. In the second-level judgment branch, the system divides the judgment path into three main branches based on the sign combination pattern. The left branch corresponds to the sign combination pattern of "all three are positive" (+, +, +), represented by the solid white circles in the figure. When the ground-radar, ground-satellite, and radar-satellite bias correlation coefficients are all positive, it indicates that the three types of observation data biases are changing in the same direction, the disturbance is layered and relatively stable, and the system determines that the atmospheric state is layered and stable. The middle branch corresponds to the sign combination pattern of "ground-radar is negative and other values ​​are positive" (-, +, +), represented by the dashed circles in the figure. When the ground-radar correlation coefficient is negative, while the ground-satellite and radar-satellite correlation coefficients are both positive, it indicates that there is a significant reverse deviation change between the ground and radar observations. This phenomenon usually corresponds to strong local convective activity, and the system determines it to be a convective development type atmospheric state. The right branch corresponds to a mixed sign combination pattern of "two positive and one negative or one positive and two negative", including (+, +, -), (+, -, -) and other patterns, which are represented by a circle filled with slashes in the figure. When the correlation coefficient shows a complex positive and negative mixed pattern, it reflects the existence of a frontal transit process at the disturbance boundary, resulting in a complex and staggered distribution of deviations between observations, and the system determines it to be a frontal transit type atmospheric state. The third layer is the basic state type output layer, which corresponds to three different atmospheric state types: Layered stable atmospheric state: characterized by uniform atmospheric structure, weak and continuous disturbances, and highly consistent deviation distributions between the three types of observation sources and the background field.

[0059] Convective development atmospheric state: characterized by the presence of significant local convective activity, with ground observations and radar observations showing opposite deviations, while satellite observations maintain a positive correlation with other observation sources.

[0060] Frontal transit type atmospheric state: characterized by unstable and drastic changes in atmospheric structure, and the deviation relationship between different observation sources is complex and changeable, reflecting the influence of the frontal system.

[0061] Figure 4The final state type output node is located at the bottom, combining the basic state type with the stability characteristics. The system categorizes the deviation correlation coefficient into three levels based on preset thresholds: strong correlation (|r| ≥ 0.7), moderate correlation (0.4 ≤ |r| < 0.7), and weak correlation (|r| < 0.4). The stability characteristics are further determined based on the strength distribution of the three correlation coefficients: a stable state is determined when the deviation correlation coefficient strength levels are all strongly correlated; a transitional state is determined when the strength levels show weak correlation; and a complex state is determined when the strength levels differ significantly.

[0062] The final atmospheric state type is output by combining the basic state type and the stability characteristics according to preset corresponding rules. For example, when the basic state type is stratified stable and the stability characteristic is stable, the output is stratified stable; when the basic state type is convective development and the stability characteristic is transitional, the output is local convective active; when the basic state type is frontal transit and the stability characteristic is composite, the output is frontal instability, etc.

[0063] Combine the basic state type with the stability characteristics and output the final atmospheric state type according to the preset mapping rules. When the basic state type is layered stable type and the stability characteristic is stable state, the corresponding atmospheric state type is "stratified stable state", which is characterized by uniform atmospheric structure, weak disturbance and continuous stability.

[0064] When the basic state type is layered stable and the stability characteristic is transitional state, the corresponding atmospheric state type is "stratification transition state", which is characterized by fluctuations in atmospheric stratification and slight disturbances.

[0065] When the basic state type is layered stable type and the stability characteristic is composite state, the corresponding atmospheric state type is “stratified composite disturbance state”, which is characterized by complex disturbances in the stratified structure and various changes.

[0066] When the basic state type is convective development type and the stability characteristic is stable state, the corresponding atmospheric state type is "local convective stable state", which is characterized by convective activity that has formed but is relatively stable.

[0067] When the basic state type is convective development type and the stability characteristics are transitional state, the corresponding atmospheric state type is "local convective active state", which is characterized by enhanced convection, rapid development and enhanced disturbance.

[0068] When the basic state type is convective development type and the stability characteristics are complex state, the corresponding atmospheric state type is "convective complex interlaced state", which is characterized by the superposition of multi-level convective disturbances and complex structure.

[0069] When the basic state type is the frontal passage type and the stability feature is the stable state, the corresponding atmospheric state type is a "frontal stable passage state", which is characterized by clear frontal influence but overall stability.

[0070] When the basic state type is the frontal passage type and the stability feature is the transition state, the corresponding atmospheric state type is a "frontal active passage state", which is characterized by obvious frontal disturbance and frequent meteorological phenomena.

[0071] When the basic state type is the frontal passage type and the stability feature is the composite state, the corresponding atmospheric state type is a "frontal interlaced unstable state", which is characterized by superimposed frontal and other disturbances, and extreme instability.

[0072] It should be noted that in actual application, the above category names and determination thresholds can be appropriately adjusted or expanded according to regional climate characteristics, observation quality and business requirements, to better adapt to specific meteorological scenarios. In addition, when there is insufficient observation data or the correlation coefficient cannot be clearly determined, a default or fuzzy determination strategy can be designed to ensure stable system operation It should be noted that: the determination threshold can be adjusted according to the specific regional climate characteristics and observation quality to improve the discrimination sensitivity; in addition, the symbol combination mode is only used as a basis for basic discrimination, and for special complex weather processes, the system can be reviewed in combination with historical data and artificial intelligence auxiliary methods to avoid misjudgment.

[0073] Further, the mapping of the atmospheric state type to the multi-source coordinated parameters in step S3 includes: mapping the stratiform stable atmospheric state in the atmospheric state type to a satellite observation dominant mode, the flow development type atmospheric state to a radar observation dominant mode, and the frontal passage type atmospheric state to a ground observation dominant mode; allocating a reference weight to the identified dominant observation source and a secondary weight to the non-dominant observation source to obtain a basic weight distribution strategy, wherein the reference weight is greater than the secondary weight, and the sum of the weights of the three observation sources meets a preset total weight requirement; when the stability feature is the stable state, the basic weight distribution remains unchanged, when the stability feature is the transition state, the weight of the dominant observation source is adjusted by a first level according to a preset adjustment coefficient, and when the stability feature is the composite state, the weight of the dominant observation source is adjusted by a second level according to a preset adjustment coefficient, to obtain the multi-source coordinated parameters combined with the stability feature.

[0074] In some embodiments, the atmospheric state type obtained according to the above determination is mapped to the corresponding dominant observation mode to guide the weight distribution of different meteorological observation data. Specifically, if the atmospheric state type is layered stable type, it is determined to be the satellite observation dominant mode; if it is convective development type, it is determined to be the radar observation dominant mode; if it is frontal transit type, it is determined to be the ground observation dominant mode. Based on this, the identified dominant observation source is given a higher baseline weight, and the other two non-dominant observation sources are assigned lower secondary weights to ensure that the sum of the weights of the three types of observation sources meets the preset total weight requirements, usually a total of 1 or 100%. For example, if the total weight is 1, the weight of the dominant observation source can be set to 0.6, and the weights of the two non-dominant observation sources are each 0.2 to ensure overall weighted balance. The weights are dynamically adjusted in combination with the stability characteristics. If the stability characteristics are determined to be stable, the above basic weight distribution remains unchanged, reflecting the relatively balanced quality of observation data under the current atmospheric state. If it is a transitional state, the weight of the dominant observation source is increased by a preset adjustment coefficient, such as 10%-20%, to strengthen the data influence of the dominant observation source and assist in the fusion process during the smooth transition phase. If it is a composite state, a higher-level second-level adjustment is performed, which may further increase the weight of the dominant observation source to 70% or higher, while correspondingly reducing the weight of non-dominant observation sources to highlight the responsiveness of the main observation source to complex atmospheric disturbances. For example, in a convective developing atmospheric state with a composite stability characteristic, the weight of radar observations as the dominant observation source may be increased from the basic 0.6 to 0.75, while the weights of ground station and satellite observations may be reduced from 0.2 to 0.125, respectively, to ensure that the fusion result better reflects the dominant information of radar observations.

[0075] It should be noted that the specific magnitude and proportion of weight adjustments must be pre-determined based on the quality statistics of historical observation data and business needs to avoid excessive adjustments that may lead to information loss or bias amplification. Furthermore, under certain extreme meteorological conditions, when data from non-dominant observation sources is highly anomaly, the weight adjustment strategy can appropriately incorporate the results of abnormal area detection and locally adjust the weights of the observation sources corresponding to the abnormal areas to improve the accuracy and stability of the overall fusion.

[0076] Step S4: Calculate the spatial gradient consistency index of different observation sources in the same area, and mark the observation abnormal area when the gradient direction or amplitude diverges; In some embodiments, spatial gradient calculations are performed on meteorological elements (such as temperature, humidity, wind speed, etc.) observed by different observation sources in the same area based on multimodal meteorological observation data. Specifically, for each observation source data in the area, the numerical difference of adjacent spatial points is used to calculate the gradient vector, which contains the direction and amplitude information of the gradient. Then, the gradient vectors corresponding to different observation sources are compared within the same spatial grid or target area, focusing on the consistency of the gradient direction and the similarity of the amplitude. During the comparison process, by calculating the relative difference in the angle and amplitude of the gradient direction, if the angle exceeds a preset threshold (for example, greater than 30 degrees) or the amplitude difference exceeds a set ratio (such as more than 20%), the spatial point is judged to be a gradient inconsistency point, and the spatial range where the point is located is marked as an observation anomaly area. This anomaly mark reflects the incoordination between different observation sources in the changing trends of meteorological elements in the area, which may be due to observation errors, data delays or local complex meteorological phenomena.

[0077] It should be noted that determining the consistency of spatial gradients not only relies on thresholds for direction and amplitude, but also considers the temporal synchronization and spatial resolution differences of the observed data to avoid misjudgments caused by time delays or differences in spatial scale. Furthermore, the marking of abnormal regions can be combined with a neighborhood expansion strategy to avoid local misjudgments caused by single-point noise, thereby improving the robustness and accuracy of anomaly detection.

[0078] Step S5: Optimize the geometric complementary weights of the multimodal meteorological observation data using multi-source coordination parameters and observation anomaly areas, generate a three-dimensional meteorological field, and realize visualization output.

[0079] It is particularly important that step S5 includes: The multi-source coordination parameters are adjusted according to the observation abnormal area, the weight of the corresponding observation source is reduced in the abnormal area, and the original weight is maintained in the normal area to obtain the adjusted fusion weight; The adjusted fusion weights are used to perform weighted fusion processing on the multimodal meteorological observation data to generate fused meteorological data; A three-dimensional grid structure is established based on the spatial range of the spatiotemporal correlation window, and the fused meteorological data is interpolated onto the grid nodes to form regularized three-dimensional meteorological field data. Perform volume rendering on 3D meteorological field data, set the color mapping and transparency parameters of meteorological elements, generate 3D visualization images, and realize dynamic display output of 3D meteorological fields.

[0080] In some embodiments, according to the preceding steps, an observation anomaly range located in a certain area of the coastal region of East China is identified, and the radar observation echo intensity in the area is obviously higher than the variation trend of the ground station wind speed and the satellite cloud top temperature. In order to weaken the influence of abnormal data on the fusion result, the system adjusts the fusion weight of the radar observation in the area from the original 0.5 to 0.3, while the ground station weight is increased from 0.25 to 0.35, and the satellite observation weight remains unchanged at 0.25. For other normal areas, the original weight distribution is continued to be used. In the fusion stage, the system calculates the temperature, humidity and wind speed and other elements at each 5km×5km grid point by weighting. For example, for the temperature of a certain grid point, the ground station observation value is 18.2℃, the radar inversion value is 19.7℃, and the satellite inversion value is 18.8℃, and the weighted result is: T fuse = 18.2×0.35 + 19.7×0.3 + 18.8×0.25≈18.8℃. This way ensures that a single abnormal observation source will not excessively affect the result. A three-dimensional grid with a vertical height of 0-12km and a horizontal resolution of 5km is established within the entire spatiotemporal correlation window, and the fused meteorological data is inserted into each grid node. For example, when the inverse distance weighted (IDW) method is used, the value of each grid node is obtained by weighting and summing the surrounding 8 observation points according to the inverse of the square of the distance, thereby forming a continuous and smooth data distribution in space. Finally, the system loads the three-dimensional grid data into the volume rendering engine, sets the gray scale mapping of temperature (the color of the area with higher temperature is lighter), the density mapping of the dashed line of humidity (the dashed line of the area with higher humidity is denser), and sets the transparency of the low humidity area to be higher, so that the high humidity area is more prominent in the three-dimensional view. Users can watch the dynamic change process of the meteorological field from the sea to the inland by playing the time axis function within the past 6 hours, so as to intuitively master the moving trend of the front.

[0081] It should be noted that the weight adjustment range and the selection of the interpolation method should be determined flexibly according to the characteristics of the actual observation data and the business requirements, so as to balance the calculation efficiency and the fusion effect. At the same time, the color and transparency mapping parameters of the volume rendering need to be reasonably set to avoid visual misdirection and ensure the scientific expression of meteorological elements.

[0082] The application also provides a three-dimensional dynamic simulation system 100 based on multi-modal weather data, which is used to execute the three-dimensional dynamic simulation method based on multi-modal weather data as described above, and the three-dimensional dynamic simulation system based on multi-modal weather data comprises: A multi-source data acquisition module 101 is configured to acquire multi-modal meteorological observation data including ground observation station information, radar observation information and satellite observation information, and calculate time difference values between observation times of each data source; A propagation speed calculation module 102 is configured to take the ratio of the time difference value to the spatial distance between the key observation points as the disturbance propagation speed; The state identification weight module 103 is used to establish a spatiotemporal correlation window based on the disturbance propagation speed, calculate the deviation correlation coefficient between each observation source within the spatiotemporal correlation window, determine the atmospheric state type based on the deviation correlation coefficient, and map the atmospheric state type to a multi-source coordination parameter; The abnormal region detection module 104 is used to calculate the spatial gradient consistency index of different observation sources in the same area, and mark the observation abnormal region when the gradient direction or amplitude diverges; The three-dimensional dynamic reconstruction module 105 is used to optimize the geometric complementary weights of multi-modal meteorological observation data using multi-source coordination parameters and observation anomaly areas, and generate a three-dimensional meteorological field and realize visual output.

[0083] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced within the present invention.

[0084] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A three-dimensional dynamic simulation method based on multimodal weather data, characterized in that: The following steps are involved: Step S1: Acquire multimodal meteorological observation data including ground observation station information, radar observation information, and satellite observation information, and calculate the time difference between the observation times of each data source; Step S2: The ratio of the time difference to the spatial distance between the key observation points is used as the disturbance propagation speed; Step S3: Establish a spatiotemporal correlation window based on the disturbance propagation speed, calculate the deviation correlation coefficient between each observation source within the spatiotemporal correlation window, and determine the atmospheric state type based on the deviation correlation coefficient; map the atmospheric state type to a multi-source coordination parameter; Step S4: Calculate the spatial gradient consistency index of different observation sources in the same area, and mark the observation abnormal area when the gradient direction or amplitude diverges; Step S5: Optimize the geometric complementary weights of the multimodal meteorological observation data using multi-source coordination parameters and observation anomaly areas, generate a three-dimensional meteorological field, and realize visualization output.

2. The three-dimensional dynamic simulation method based on multimodal weather data according to claim 1, characterized in that: Step S1 includes the following steps: At the same time, it receives point-like meteorological data from ground automatic weather stations, spatial scanning data from weather radars, and spatial scanning data from meteorological satellites as original observation data sets; Convert the original observation data set into a standardized format to obtain multimodal meteorological observation data; Calculate the time difference between different data sources in the multimodal meteorological observation data within the same observation period, where the time difference includes the time difference between ground station and satellite observation, the time difference between ground station and radar observation, and the time difference between radar and satellite observation.

3. The three-dimensional dynamic simulation method based on multimodal weather data according to claim 2, characterized in that: Calculating the time difference between different data sources in the multimodal meteorological observation data within the same observation period includes: Extracting the valid scanning point closest to the spatial position of the target ground station from the spatial scanning data of the weather radar; Compare the actual observation timestamp of the target ground station with the scanning timestamp of the valid scanning point, and calculate the absolute difference, which is recorded as the ground station-radar time difference of the target ground station; According to the pixel unit information corresponding to the geographical location of the target ground station in the spatial scanning data of the meteorological satellite, the precise observation timestamp of the pixel unit is read to obtain the pixel timestamp; The pixel timestamp is compared with the actual observation timestamp of the target ground station, and the absolute time interval between the two timestamps is taken as the ground station-satellite time difference of the target ground station; Calculate the time difference between the meteorological radar and meteorological satellite observation data based on the target area range to obtain the radar-satellite time difference of the target area; The ground station-radar time difference, ground station-satellite time difference and radar-satellite time difference are combined into the time difference value.

4. The three-dimensional dynamic simulation method based on multimodal weather data according to claim 3, characterized in that: The time difference between the meteorological radar and meteorological satellite observation data is calculated based on the target area range, including: Determine a target area, wherein the target area is specifically a spatial grid unit or a weather characteristic area; The average time of all scanning points constituting the target area observation is extracted from the spatial scanning data of the weather radar and recorded as the regional radar average time; The average time of all satellite pixels covering the target area is extracted from the spatial scanning data of the meteorological satellite and recorded as the regional satellite average time; The absolute difference between the regional radar average time and the regional satellite average time is calculated to obtain the radar-satellite time difference of the target area.

5. The three-dimensional dynamic simulation method based on multimodal weather data according to claim 4, characterized in that: Step S2 includes: Select two different types of observation points with the smallest observation time difference from the multimodal meteorological observation data as key observation point pairs, including the combination of ground station and radar station, the combination of ground station and satellite pixel, or the combination of radar station and satellite pixel; Calculate the actual spatial distance between key observation point pairs; Determine the propagation time based on the key observation point pairs and time differences. When there are multiple time differences, the time difference with the smallest value is selected as the propagation time. The actual spatial distance between the key observation point pairs is divided by the propagation time to obtain the dynamic propagation velocity.

6. The three-dimensional dynamic simulation method based on multimodal weather data according to claim 5, characterized in that: Establishing a spatiotemporal correlation window based on the disturbance propagation speed in step S3 includes: The geometric midpoint of the key observation point pair is used as the spatial center coordinate, and the average observation time of the two observation points is used as the temporal center coordinate; The preset reference time constant is divided by the disturbance propagation speed to obtain the time radius value, and the disturbance propagation speed is multiplied by the time radius value to obtain the space radius value; Calculate the azimuth angle of the line connecting the key observation point pairs as the main propagation direction angle; When the disturbance propagation speed is greater than a preset speed threshold, it is determined to be a directional propagation phenomenon and an elliptical shape parameter is set; when the disturbance propagation speed is less than or equal to the preset speed threshold, it is determined to be a diffusion propagation phenomenon and a circular shape parameter is set to obtain geometric shape parameters; The spatial center coordinates and the time center coordinates are used as the positioning center of the space-time correlation window. The spatial radius value and the time radius value are used to determine the spatial boundary range and the time extension range of the window respectively. The circular or elliptical structure of the window is set according to the geometric shape parameters. When it is an ellipse, the direction of the major axis of the ellipse is determined by the main propagation direction angle to obtain the space-time correlation window.

7. The three-dimensional dynamic simulation method based on multimodal weather data according to claim 6, characterized in that: Calculating the deviation correlation coefficient between each observation source in the spatiotemporal correlation window in step S3 includes: Extract various types of observation data within the spatiotemporal correlation window and obtain the numerical forecast background field data corresponding to the window area; Match various observation data with the background field grid points in the numerical forecast background field data, and use the nearest neighbor method to determine the background field value corresponding to each observation point to form an observation-background field matching data set; Based on the observation-background field matching data set, the deviation value sequence of each observation source and the background field is calculated, where the deviation value sequence includes ground deviation sequence, radar deviation sequence and satellite deviation sequence; The deviation correlation coefficients between observation sources are calculated through the deviation value sequence, and the deviation correlation coefficients between the ground deviation sequence and the radar deviation sequence, the deviation correlation coefficients between the ground deviation sequence and the satellite deviation sequence, and the deviation correlation coefficients between the radar deviation sequence and the satellite deviation sequence are calculated respectively.

8. The three-dimensional dynamic simulation method based on multimodal weather data according to claim 7, characterized in that: In step S3, judging the atmospheric state type by the deviation correlation coefficient includes: Check the positive and negative correlation coefficients of the ground-radar bias, the ground-satellite bias, and the radar-satellite bias, and record the sign combination pattern of the three correlation coefficients; When the symbol combination pattern shows that all three correlation coefficients are positive, it is determined to be a layered stable atmospheric state; when the symbol combination pattern shows that the ground-radar deviation correlation coefficient is negative and the other two are positive, it is determined to be a convective development atmospheric state; when the symbol combination pattern shows two positive and one negative or one positive and two negative patterns, it is determined to be a frontal transit atmospheric state, and the basic state type is obtained; The deviation correlation coefficient is divided into three levels: strong correlation, medium correlation and weak correlation according to the preset threshold rule. When the deviation correlation coefficient strength levels are all strongly correlated, it is determined to be a stable state; when the strength levels are weakly correlated, it is determined to be a transitional state; when the strength levels are greatly different, it is determined to be a composite state, and the stability characteristics are obtained; Combine the basic state type with the stability characteristics and output the atmospheric state type according to the preset corresponding rules.

9. The three-dimensional dynamic simulation method based on multimodal weather data according to claim 8, characterized in that: Mapping the atmospheric state type to the multi-source coordination parameter in step S3 includes: The layered stable atmospheric state is mapped to the satellite observation-dominated mode, the flow development atmospheric state is mapped to the radar observation-dominated mode, and the frontal transit atmospheric state is mapped to the ground observation-dominated mode. Assign a baseline weight to the identified dominant observation source and a secondary weight to the non-dominant observation source, and obtain a basic weight allocation strategy, where the baseline weight is greater than the secondary weight, and the sum of the weights of the three observation sources meets the preset total weight requirement; When the stability characteristic is in a stable state, the basic weight distribution remains unchanged. When the stability characteristic is in a transitional state, the weight of the dominant observation source is adjusted at the first level according to the preset adjustment coefficient. When the stability characteristic is in a composite state, the weight of the dominant observation source is adjusted at the second level according to the preset adjustment coefficient to obtain the multi-source coordination parameters combined with the stability characteristic.

10. A three-dimensional dynamic simulation system based on multimodal weather data, characterized in that: For executing the three-dimensional dynamic simulation method based on multimodal weather data according to claim 1, the three-dimensional dynamic simulation system based on multimodal weather data comprises: Multi-source data acquisition module, used to obtain multi-modal meteorological observation data including ground observation station information, radar observation information and satellite observation information, and calculate the time difference between the observation time of each data source; The propagation speed calculation module is used to take the ratio of the time difference to the spatial distance between the key observation points as the disturbance propagation speed; The state recognition weight module is used to establish a spatiotemporal correlation window based on the disturbance propagation speed, calculate the deviation correlation coefficient between each observation source within the spatiotemporal correlation window, and determine the atmospheric state type based on the deviation correlation coefficient; the atmospheric state type is mapped to a multi-source coordination parameter; The abnormal region detection module is used to calculate the spatial gradient consistency index of different observation sources in the same area, and mark the observation abnormal region when the gradient direction or amplitude diverges; The three-dimensional dynamic reconstruction module is used to optimize the geometric complementary weights of multi-modal meteorological observation data using multi-source coordination parameters and observation anomaly areas, generate a three-dimensional meteorological field and realize visual output.

Images