Meteorological data acquisition method and device, computer device and storage medium

Abstract

This application relates to a meteorological data acquisition method, apparatus, computer equipment, and storage medium. The method includes: acquiring satellite meteorological data containing default values ​​for a target area; gridding the target area to determine the satellite meteorological data of each grid point in the target area grid, obtaining a first data sample set; acquiring numerical model meteorological data, and determining the numerical model meteorological data of each grid point in the target area grid according to the gridding resolution of the target area, obtaining a second data sample set; updating the numerical model meteorological data of the target grid points in the second data sample set to the default values ​​according to the correspondence between the first and second data sample sets and the default values ​​of the satellite meteorological data, obtaining a third data sample set; training a deep learning model based on the first and third data sample sets; and inputting the satellite meteorological data into the trained deep learning model to obtain updated satellite meteorological data. This method rapidly obtains complete satellite meteorological data.

Description

Technical Field

[0001] This application relates to the field of satellite observation technology, and in particular to a meteorological data acquisition method, apparatus, computer equipment, and storage medium. Background Technology

[0002] With the development of satellite observation technology, the technology of collecting meteorological data using polar-orbiting meteorological satellites has emerged. Polar-orbiting meteorological satellites operate in orbits between 650 and 1500 kilometers above the Earth, orbiting the North and South Poles. Their orbital period is approximately 115 minutes. Due to their low orbits, polar-orbiting meteorological satellites can accurately acquire various meteorological data of the Earth's atmosphere and surface, which can be used for precise monitoring and forecasting of the weather in target areas.

[0003] However, due to the Earth's rotation, meteorological data acquired by polar-orbiting meteorological satellites is discontinuous. Unlike geostationary satellites, they cannot acquire complete meteorological observation data of the target monitoring area within a single time period. Therefore, they cannot meet the meteorological monitoring needs of the target monitoring area. Summary of the Invention

[0004] Therefore, it is necessary to provide a meteorological data acquisition method, device, computer equipment, and storage medium to address the aforementioned technical problems.

[0005] A method for acquiring meteorological data, the method comprising:

[0006] Acquire satellite meteorological data for a target area within a target time frame;

[0007] The target area is divided into grids, and the satellite meteorological data corresponding to each grid point in the target area grid is determined to obtain a first data sample set, which contains default values ​​of satellite meteorological data.

[0008] Acquire numerical model meteorological data, and determine the numerical model meteorological data corresponding to each grid point in the grid of the target area according to the same grid resolution as the grid division of the target area, to obtain a second data sample set;

[0009] Based on the correspondence between the first data sample set and the second data sample set and the default value of the satellite meteorological data, the numerical model meteorological data corresponding to the target grid point in the second data sample set is updated to the preset default value to obtain the third data sample set;

[0010] The second data sample set and the third data sample set are input into a preset deep learning model to train the deep learning model;

[0011] The satellite meteorological data of the target area is input into the trained deep learning model to obtain the updated satellite meteorological data of the target area.

[0012] In one embodiment, the target area is divided into grids, and the satellite meteorological data corresponding to each grid point in the target area grid is determined to obtain a first data sample set. The first data sample set contains default values ​​for satellite meteorological data, including:

[0013] The target area is divided into grids to obtain the statistical range of the satellite meteorological data corresponding to each grid point;

[0014] When there is no satellite meteorological data within the statistical range of the satellite meteorological data, the data corresponding to the grid points included in the statistical range will be determined as the default value of the satellite meteorological data.

[0015] When satellite meteorological data exists within the statistical range of the satellite meteorological data, the data corresponding to the grid points included in the statistical range are determined as the average value of the satellite meteorological data within the statistical range;

[0016] The satellite meteorological data corresponding to each grid point in the target area grid is used as the first data sample set.

[0017] In one embodiment, the step of acquiring numerical model meteorological data involves determining the numerical model meteorological data corresponding to each grid point in the grid of the target area based on the same grid resolution as the grid division of the target area, thereby obtaining a second data sample set, including:

[0018] Based on the preset resolution corresponding to the numerical model gridding, the numerical model meteorological data corresponding to each grid point in the numerical model grid are calculated.

[0019] Based on the same grid resolution as the target area, the numerical model meteorological data is mapped to grid points with the same grid resolution as the target area using spatial interpolation, thus obtaining a second data sample set of numerical model meteorological data with the same grid resolution as the target area.

[0020] In one embodiment, the step of updating the numerical model meteorological data corresponding to the target grid points in the second data sample set to preset default values ​​based on the correspondence between the first data sample set and the second data sample set and the default values ​​of the satellite meteorological data, to obtain the third data sample set, includes:

[0021] Based on the correspondence between satellite meteorological data at each grid point in the first data sample set and numerical model meteorological data at each grid point in the second data sample set, the grid points in the second data sample set that correspond to the default values ​​of the satellite meteorological data in the first data sample set are determined as target grid points.

[0022] Using the preset default value as a mask, in the second data sample set, the numerical model meteorological data corresponding to the target grid point is assigned the preset default value to obtain the third data sample set.

[0023] In one embodiment, the step of mapping the numerical model meteorological data to grid points with the same grid resolution as the target area using spatial interpolation to obtain a second data sample set of numerical model meteorological data with the same grid resolution as the target area includes:

[0024] Based on the same gridding resolution as the target region grid, the target region grid is projected with equal latitude and longitude to obtain the projected grid.

[0025] Based on the latitude and longitude information of each grid point in the projection grid, the numerical model meteorological data of the corresponding grid points in the numerical model grid with the same latitude and longitude are spatially interpolated onto the grid points of the projection grid to obtain a second data sample set of numerical model meteorological data based on the grid resolution of the target area.

[0026] A meteorological data acquisition device, the device comprising:

[0027] The acquisition module is used to acquire satellite meteorological data for a target area within a target time range;

[0028] The first determining module is used to divide the target area into grids, determine the satellite meteorological data corresponding to each grid point in the target area grid, and obtain a first data sample set, wherein the first data sample set contains default values ​​of satellite meteorological data.

[0029] The second determining module is used to acquire numerical model meteorological data and determine the numerical model meteorological data corresponding to each grid point in the grid of the target area according to the resolution of the gridding of the target area, so as to obtain a second data sample set.

[0030] The third determining module is used to update the numerical model meteorological data corresponding to the target grid point in the second data sample set to the preset default value according to the correspondence between the first data sample set and the second data sample set and the default value of the satellite meteorological data, so as to obtain the third data sample set.

[0031] The training module is used to input the second data sample set and the third data sample set into a preset deep learning model to train the deep learning model;

[0032] The update module is used to input the satellite meteorological data of the target area into the trained deep learning model to obtain the updated satellite meteorological data of the target area.

[0033] In one embodiment, the second determining module is specifically used to calculate the numerical model meteorological data corresponding to each grid point in the numerical model grid according to the preset resolution corresponding to the numerical model grid.

[0034] Based on the grid resolution of the target area, the numerical model meteorological data is mapped to grid points with the same grid resolution as the target area using spatial interpolation, resulting in a second data sample set of numerical model meteorological data with the same grid resolution as the target area.

[0035] In one embodiment, the third determining module is specifically used to determine the grid points in the second data sample set that correspond to the default values ​​of the satellite meteorological data in the first data sample set as target grid points based on the correspondence between the satellite meteorological data at each grid point in the first data sample set and the numerical model meteorological data at each grid point in the second data sample set.

[0036] Using the preset default value as a mask, in the second data sample set, the numerical model meteorological data corresponding to the target grid point is assigned the preset default value to obtain the third data sample set.

[0037] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program performing the following steps:

[0038] Acquire satellite meteorological data for a target area within a target time frame;

[0039] The target area is divided into grids, and the satellite meteorological data corresponding to each grid point in the target area grid is determined to obtain a first data sample set, which contains default values ​​of satellite meteorological data.

[0040] Acquire numerical model meteorological data, and determine the numerical model meteorological data corresponding to each grid point in the grid of the target area according to the same grid resolution as the grid division of the target area, to obtain a second data sample set;

[0041] Based on the correspondence between the first data sample set and the second data sample set and the default value of the satellite meteorological data, the numerical model meteorological data corresponding to the target grid point in the second data sample set is updated to the preset default value to obtain the third data sample set;

[0042] The second data sample set and the third data sample set are input into a preset deep learning model to train the deep learning model;

[0043] The satellite meteorological data of the target area is input into the trained deep learning model to obtain the updated satellite meteorological data of the target area.

[0044] A computer-readable storage medium having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0045] Acquire satellite meteorological data for a target area within a target time frame;

[0046] The target area is divided into grids, and the satellite meteorological data corresponding to each grid point in the target area grid is determined to obtain a first data sample set, which contains default values ​​of satellite meteorological data.

[0047] Acquire numerical model meteorological data, and determine the numerical model meteorological data corresponding to each grid point in the grid of the target area according to the same grid resolution as the grid division of the target area, to obtain a second data sample set;

[0048] Based on the correspondence between the first data sample set and the second data sample set and the default value of the satellite meteorological data, the numerical model meteorological data corresponding to the target grid point in the second data sample set is updated to the preset default value to obtain the third data sample set;

[0049] The second data sample set and the third data sample set are input into a preset deep learning model to train the deep learning model;

[0050] The satellite meteorological data of the target area is input into the trained deep learning model to obtain the updated satellite meteorological data of the target area.

[0051] The aforementioned meteorological data acquisition method, apparatus, computer equipment, and storage medium acquire satellite meteorological data for a target area within a target time range; divide the target area into grids, determine the satellite meteorological data corresponding to each grid point in the target area grid, and obtain a first data sample set, which includes default values ​​for satellite meteorological data; acquire numerical model meteorological data, and determine the numerical model meteorological data corresponding to each grid point in the target area grid according to the same grid resolution as the grid division of the target area, and obtain a second data sample set; update the numerical model meteorological data corresponding to the target grid points in the second data sample set to preset default values ​​according to the correspondence between the first data sample set and the second data sample set and the default values ​​of the satellite meteorological data, and obtain a third data sample set; input the second data sample set and the third data sample set into a preset deep learning model to train the deep learning model; input the satellite meteorological data of the target area into the trained deep learning model to obtain updated satellite meteorological data of the target area. Using this method, the acquisition time of satellite meteorological data can be reduced, and spatiotemporally continuous satellite meteorological data can be obtained. Attached Figure Description

[0052] Figure 1 This is a flowchart illustrating a meteorological data acquisition method in one embodiment;

[0053] Figure 2 This is a flowchart illustrating a target region meshing method in one embodiment;

[0054] Figure 3 This is a flowchart illustrating a method for obtaining a second data sample set in one embodiment;

[0055] Figure 4 This is a flowchart illustrating the spatial interpolation step in the acquisition of the second data sample set in one embodiment;

[0056] Figure 5 This is a flowchart illustrating a method for obtaining a third data sample set in one embodiment;

[0057] Figure 6 This is a block diagram showing the results of a meteorological data acquisition device in one embodiment;

[0058] Figure 7 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0060] In one embodiment, such as Figure 1 As shown, a method for acquiring meteorological data is provided. This embodiment illustrates the application of this method to a terminal device. It is understood that this method can also be applied to a server, or to a system including both a terminal and a server, and is implemented through interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0061] Step 101: Obtain satellite meteorological data for the target area within the target time range.

[0062] In practice, polar-orbiting meteorological satellites can clearly acquire various meteorological data of the Earth's atmosphere and surface, including latitude and longitude information and specific meteorological data values. Computer equipment, based on a preset latitude and longitude range, acquires satellite meteorological data (e.g., AOD (Aerosol Optical Depth) data) collected by polar-orbiting meteorological satellites within a target time range for a specific area. Specifically, the computer equipment can acquire AOD data for the target area collected by the MODIS (Moderate Resolution Imaging Spectroradiometer) sensor on the polar-orbiting satellite within a week. AOD data is an important physical quantity characterizing the degree of atmospheric turbidity.

[0063] Optionally, in addition to AOD data, satellite meteorological data can also be other atmospheric or meteorological data such as air pollutant data. Therefore, this application does not limit the specific satellite meteorological data used in its embodiments.

[0064] Step 102: Divide the target area into grids, determine the satellite meteorological data corresponding to each grid point in the target area grid, and obtain the first data sample set.

[0065] The first data sample set contains default values ​​for satellite meteorological data.

[0066] In implementation, the computer equipment divides the latitude and longitude range of the target area into grids of equal latitude and longitude (for example, the grid size can be set to 10-100 km), resulting in a grid of the target area. Within this grid, the intersecting grid points are used as centers to determine the corresponding satellite meteorological data within the range of each grid point. Because the collected satellite meteorological data cannot perfectly correspond to every grid point, when there is no satellite meteorological data in the vicinity of a grid point, the default value of the satellite meteorological data corresponding to that grid point is used. Thus, the first data sample set of satellite meteorological data corresponding to each grid point is obtained.

[0067] Step 103: Obtain numerical model meteorological data. Based on the same grid resolution as the target area, determine the numerical model meteorological data corresponding to each grid point in the target area grid to obtain the second data sample set.

[0068] In implementation, numerical model meteorological data can be calculated using a pre-set numerical model model. For example, a numerical weather prediction model (CMM) is a method that, under certain initial and boundary conditions, uses a large computer to perform numerical calculations to solve the fluid dynamics and thermodynamic equations describing weather evolution, in order to predict atmospheric motion and weather phenomena within a target time range. The meteorological data obtained through this numerical model method is continuous. Therefore, in this embodiment, the computer equipment uses the meteorological element data of the target area for the past three years as input values ​​to predict continuous numerical model meteorological data within the target time range (the meteorological data is also AOD data to maintain consistency with satellite meteorological data). The computer equipment acquires the numerical model meteorological data of the target area, and then, according to the same grid resolution as the grid division of the target area (i.e., the grid resolution used for dividing satellite meteorological data), the numerical model meteorological data is assigned to the grid corresponding to the grid resolution of the target area. That is, each grid point in the grid of the target area's grid resolution corresponds to one numerical model meteorological data point, thus obtaining a second data sample set of numerical model meteorological data for each grid point.

[0069] Optionally, in addition to numerical weather prediction models, numerical model models can also be CMAQ (Community Multiscale Air Quality) models, CAMx (Comprehensive Air Quality) models, etc. Therefore, the embodiments of this application do not limit the means of acquiring numerical model meteorological data using numerical model models.

[0070] Step 104: Based on the correspondence between the first data sample set and the second data sample set and the default values ​​of satellite meteorological data, update the numerical model meteorological data corresponding to the target grid points in the second data sample set to the preset default values ​​to obtain the third data sample set.

[0071] In implementation, the computer equipment updates the numerical model meteorological data corresponding to the target grid points in the second data sample set to preset default values ​​(represented by NA) based on the correspondence between the latitude and longitude grid points between the first data sample set and the second data sample set, as well as the default values ​​of the satellite meteorological data contained in the first data sample set. The updated numerical model meteorological data containing the preset default values ​​is then determined as the third data sample set.

[0072] Step 105: Input the second data sample set and the third data sample set into the preset deep learning model to train the deep learning model.

[0073] In practice, the computer equipment inputs the second and third data sample sets into the preset deep learning model to train the deep learning model.

[0074] Optionally, since the data sample set obtained after gridding is equivalent to image data in a two-dimensional array, the deep learning model can be a partialconv model, or other image inpainting models such as DCGAN (Deep Convolution Generative Adversarial Networks), and this application does not limit the specific model.

[0075] Specifically, taking the DCGAN model as an example, a brief introduction to model training is as follows: First, the computer inputs the second and third data sample sets into the DCGAN model to be trained, training the model generator. The images (two-dimensional arrays) generated by the model generator are then fed into the discriminator, and the mean squared error of its output is calculated with the label array, which is all 1s. The resulting loss is used as the adversarial loss. Then, the completed data corresponding to the missing values ​​in the output and the numerical model meteorological data covered by the missing value mask are used to calculate the pixel-wise L1 (pixel-wise is the basic unit of image at the pixel level) loss. The coefficients of the two are preset to 0.001 and 0.999, respectively. The two losses are added together and used in reverse to update the parameters of the model generator until the training of the model generator is completed. Then, the computer equipment trains the model discriminator of the deep learning model. The computer equipment inputs the missing value-completed data and the numerical model meteorological data covered by the missing value mask into the model discriminator, respectively, calculates the loss with the label arrays of all 1s and all 0s, and then multiplies it by a preset coefficient of 0.5 as the adversarial loss, which is used to update the model discriminator until the training of the model discriminator is completed. Based on this, the DCGAN model training is completed.

[0076] Step 106: Input the satellite meteorological data of the target area into the trained deep learning model to obtain the updated satellite meteorological data of the target area.

[0077] In practice, satellite meteorological data (discontinuous satellite meteorological data) of the target area is input into the deep learning model based on the trained deep learning model. The model generator in the model processes the input satellite meteorological data, and finally, complete spatiotemporally continuous satellite meteorological data can be obtained.

[0078] For example, if the trained deep learning model is a DCGAN model, and the computer equipment applies the trained DCGAN model to complete the AOD data of polar-orbiting satellites, then the model generator of the DCGAN model is used to process the input grid image (grid point data sample set), and then the Poisson mixture method is used to improve the boundary effect. Finally, the updated (completed) spatiotemporally continuous satellite AOD data can be obtained.

[0079] In the above meteorological data acquisition method, satellite meteorological data of the target area within the target time range is acquired; the target area is divided into grids, and the satellite meteorological data corresponding to each grid point in the target area grid is determined to obtain a first data sample set, which includes default values ​​for satellite meteorological data; numerical model meteorological data is acquired, and the numerical model meteorological data corresponding to each grid point in the target area grid is determined according to the same grid resolution as the grid division of the target area to obtain a second data sample set; according to the correspondence between the first data sample set and the second data sample set and the default values ​​of satellite meteorological data, the numerical model meteorological data corresponding to the target grid points in the second data sample set is updated to preset default values ​​to obtain a third data sample set; the second data sample set and the third data sample set are input into a preset deep learning model to train the deep learning model; the satellite meteorological data of the target area is input into the trained deep learning model to obtain updated satellite meteorological data of the target area. This method involves inputting a continuous second set of data samples and a discontinuous third set of data samples containing default values ​​from numerical model meteorological data into a deep learning model for training. This yields the relationship between continuous and discontinuous data. Furthermore, based on the principle of relational transfer learning, discontinuous satellite meteorological data is then input into the trained deep learning model to obtain complete spatiotemporal continuous satellite meteorological data. This reduces the difficulty of acquiring continuous satellite meteorological data and improves the efficiency of acquiring continuous satellite meteorological data.

[0080] In one embodiment, such as Figure 2 As shown, the specific processing procedure for step 102 is as follows:

[0081] Step 1021: Divide the target area into grids to obtain the statistical range of satellite meteorological data corresponding to each grid point.

[0082] In implementation, the computer equipment divides the target area into equal latitude and longitude sections, resulting in a target area grid composed of grid lines and intersection points (grid points). Within this target area grid, a circular area with a preset distance as its radius, centered on a grid point, serves as the statistical range for satellite meteorological data. For example, a circular area within 50km centered on a grid point is used as the statistical range for satellite AOD data, allowing for the statistical analysis of satellite AOD data for each statistical range area.

[0083] Optionally, the size of the statistical range (i.e., the preset distance length) and the shape of the statistical range can be adjusted according to the different types of satellite meteorological data to be statistically analyzed. Therefore, the specific method of dividing the statistical range is not limited in this application embodiment.

[0084] Step 1022: When there is no satellite meteorological data within the statistical range of satellite meteorological data, the data corresponding to the grid points included in the statistical range shall be determined as the default value of satellite meteorological data.

[0085] In practice, when there is no corresponding satellite meteorological data within the statistical range of satellite meteorological data, the computer equipment determines the data corresponding to the grid points included in the statistical range as the default value of satellite meteorological data, denoted as NA.

[0086] Step 1023: When satellite meteorological data exists within the statistical range of satellite meteorological data, the data corresponding to the grid points included in the statistical range is determined as the average value of the satellite meteorological data within the statistical range.

[0087] In practice, when there are multiple corresponding satellite meteorological data within a statistical range of satellite meteorological data, the computer equipment calculates the average value of the satellite meteorological data within the statistical range and uses this average value as the satellite meteorological data corresponding to the grid points included in the statistical range.

[0088] Step 1024: Use the satellite meteorological data corresponding to each grid point in the target area grid as the first data sample set.

[0089] In implementation, the computer equipment uses the satellite meteorological data corresponding to each grid point in the target area grid as the first data sample set, which includes missing values ​​of the satellite meteorological data.

[0090] In this embodiment, by performing gridding on the target area and further obtaining the satellite meteorological data corresponding to each grid point, the discontinuous satellite meteorological data scattered in the target area is transformed into image data in the form of a two-dimensional array. Then, the image data containing missing values ​​can be completed using the image completion method to obtain complete image information, that is, to obtain continuous satellite meteorological data.

[0091] In one embodiment, such as Figure 3 As shown, the specific processing procedure for step 103 is as follows:

[0092] Step 1031: Calculate the numerical model meteorological data corresponding to each grid point in the numerical model grid according to the preset resolution of the numerical model grid.

[0093] In practice, the numerical model meteorological data calculated by the computer equipment based on the numerical model can also have a corresponding grid resolution. Therefore, the computer equipment obtains the numerical model meteorological data corresponding to each grid point in the numerical model grid according to the preset resolution of the numerical model grid.

[0094] Specifically, the numerical model meteorological data calculated by the numerical model is continuous data, which means that the numerical model meteorological data is denser in the target area, and the corresponding grid that needs to be divided is also denser. The computer equipment calculates the numerical model meteorological data corresponding to each grid point in this dense grid.

[0095] Step 1032: Based on the same grid resolution as the target area, the numerical model meteorological data is mapped to each grid point of the target area grid resolution using spatial interpolation, thus obtaining a second data sample set of numerical model meteorological data based on the target area grid resolution.

[0096] In practice, the computer equipment uses spatial interpolation to map the meteorological data of the numerical model to each grid point of the target area grid resolution based on the resolution of the target area grid set for the satellite meteorological data. This results in numerical model meteorological data with the same resolution as the target area grid of the satellite meteorological data, which is then used as the second data sample set.

[0097] In one embodiment, such as Figure 4 As shown, the specific processing procedure for step 1032 is as follows:

[0098] Step 401: Based on the same grid resolution as the target area grid, perform equal latitude and longitude projection on the target area grid to obtain the projected grid.

[0099] In practice, the computer equipment projects the grid of the target area with the same grid resolution as the grid division of the target area, using the latitude and longitude of the grid points as a reference, to obtain the projected grid.

[0100] Step 402: Based on the latitude and longitude information of each grid point in the projection grid, spatially interpolate the numerical model meteorological data of the corresponding grid points with the same latitude and longitude in the numerical model grid to the grid points of the projection grid to obtain the second data sample set of numerical model meteorological data based on the grid resolution of the target area.

[0101] In practice, the computer equipment uses the latitude and longitude information of each grid point in the obtained projection grid to query the corresponding grid points of the same latitude and longitude in the high-resolution grid obtained by the numerical model calculation. The numerical model meteorological data at the grid points of the same latitude and longitude are mapped to the grid points of the same latitude and longitude in the projection grid through spatial interpolation (e.g., ESMF model coupling and interpolation). In this way, the numerical model meteorological data corresponding to each grid point in the projection grid is obtained. The computer equipment uses the numerical model meteorological data as the second data sample set.

[0102] In this embodiment, high-resolution numerical model meteorological data is converted into a low-resolution projection grid based on the resolution of satellite meteorological data by spatial interpolation, thereby unifying the resolution standard of the data samples and facilitating the transfer of relationship results after learning the relationship between continuous and discontinuous numerical model meteorological data.

[0103] In one embodiment, such as Figure 5 As shown, the specific processing procedure for step 104 is as follows:

[0104] Step 1041: Based on the correspondence between the satellite meteorological data at each grid point in the first data sample set and the numerical model meteorological data at each grid point in the second data sample set, determine the grid points in the second data sample set that correspond to the default values ​​of the satellite meteorological data in the first data sample set as target grid points.

[0105] In practice, the grid corresponding to the second data sample set is projected onto the grid corresponding to the first data sample set. Therefore, there is a one-to-one correspondence between the first and second data sample sets. The first data sample set contains preset default values. Therefore, the computer equipment determines the grid points in the grid corresponding to the second data sample set that correspond to the default values ​​of the satellite meteorological data in the first data sample set as the target grid points of the second data sample.

[0106] Step 1042: Using the preset default value as a mask, assign the preset default value to the numerical model meteorological data corresponding to the target grid point in the second data sample set to obtain the third data sample set.

[0107] In practice, the computer equipment uses the default value NA as a mask. In the continuous numerical model meteorological data (i.e., the second data sample set), the numerical model meteorological data corresponding to the determined target grid point is assigned the preset default value. In other words, the target grid point meteorological data in the second data sample set is covered by the default value mask, thereby obtaining the third data sample set containing the default value of the numerical model meteorological data.

[0108] In this embodiment, by using preset default values ​​as a mask, the numerical model meteorological data corresponding to the determined target grid points in the continuous numerical model meteorological data sample set are covered with the default values ​​to obtain a third data sample set (i.e., a discontinuous numerical model meteorological dataset). This is used to construct training samples for deep learning models, which facilitates the learning of the relationship between continuous and discontinuous data.

[0109] It should be understood that, although Figure 1-5 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1-5 At least some of the steps in the process may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but may be executed at different times. The execution order of these steps or stages is not necessarily sequential, but may be executed in turn or alternately with other steps or at least some of the steps or stages in other steps.

[0110] In one embodiment, such as Figure 6 As shown, a meteorological data acquisition device 600 is provided, including: an acquisition module 610, a first determination module 620, a second determination module 630, a third determination module 640, a training module 650, and an update module 660, wherein:

[0111] The acquisition module 610 is used to acquire satellite meteorological data of the target area within the target time range.

[0112] The first determining module 620 is used to divide the target area into grids, determine the satellite meteorological data corresponding to each grid point in the target area grid, and obtain a first data sample set, which contains default values ​​of satellite meteorological data.

[0113] The second determining module 630 is used to acquire numerical model meteorological data and, based on the same grid resolution as the grid division of the target area, determine the numerical model meteorological data corresponding to each grid point in the grid of the target area, thereby obtaining the second data sample set.

[0114] The third determining module 640 is used to update the numerical model meteorological data corresponding to the target grid points in the second data sample set to preset default values ​​based on the correspondence between the first data sample set and the second data sample set and the default values ​​of satellite meteorological data, so as to obtain the third data sample set.

[0115] The training module 650 is used to input the second and third data sample sets into a preset deep learning model to train the deep learning model.

[0116] The update module 660 is used to input satellite meteorological data of the target area into the trained deep learning model to obtain updated satellite meteorological data of the target area.

[0117] In one embodiment, the first determining module 620 is specifically used to divide the target area into grids to obtain the statistical range of satellite meteorological data corresponding to each grid point after gridding.

[0118] When there is no satellite meteorological data within the statistical range of satellite meteorological data, the data corresponding to the grid points included in the statistical range will be determined as the default value of satellite meteorological data.

[0119] When satellite meteorological data exists within the statistical range of satellite meteorological data, the data corresponding to the grid points included in the statistical range are determined as the average value of the satellite meteorological data within the statistical range.

[0120] The satellite meteorological data corresponding to each grid point in the target area grid is used as the first data sample set.

[0121] In one embodiment, the second determining module 630 is specifically used to calculate the numerical model meteorological data corresponding to each grid point in the numerical model grid according to the preset resolution of the numerical model grid.

[0122] Based on the same grid resolution as the target area, the numerical model meteorological data are mapped to each grid point of the target area's grid resolution using spatial interpolation, resulting in a second data sample set of numerical model meteorological data based on the target area's grid resolution.

[0123] In one embodiment, the third determining module 640 is specifically used to determine the grid points in the second data sample set that correspond to the default values ​​of the satellite meteorological data in the first data sample set as target grid points based on the correspondence between the satellite meteorological data at each grid point in the first data sample set and the numerical model meteorological data at each grid point in the second data sample set.

[0124] Using the preset default value as a mask, the numerical model meteorological data corresponding to the target grid point in the second data sample set are assigned the preset default value to obtain the third data sample set.

[0125] In one embodiment, the second determining module 630 is further configured to perform equal latitude and longitude projection on the target area grid according to the same gridding resolution as the grid division of the target area, to obtain a projected grid.

[0126] Based on the latitude and longitude information of each grid point in the projection grid, the numerical model meteorological data of the corresponding grid points with the same latitude and longitude in the numerical model grid are spatially interpolated to the grid points of the projection grid to obtain the second data sample set of numerical model meteorological data based on the grid resolution of the target area.

[0127] The aforementioned meteorological data acquisition device includes an acquisition module 610 for acquiring satellite meteorological data for a target area within a target time range; a first determination module 620 for dividing the target area into grids and determining the satellite meteorological data corresponding to each grid point in the target area grid, resulting in a first data sample set containing default values ​​for satellite meteorological data; a second determination module 630 for acquiring numerical model meteorological data and determining the numerical model meteorological data corresponding to each grid point in the target area grid based on the same grid resolution as the target area grid, resulting in a second data sample set; a third determination module 640 for updating the numerical model meteorological data corresponding to the target grid points in the second data sample set to preset default values ​​based on the correspondence between the first and second data sample sets and the default values ​​for satellite meteorological data, resulting in a third data sample set; a training module 650 for inputting the second and third data sample sets into a preset deep learning model for training the deep learning model; and an update module 660 for inputting the satellite meteorological data of the target area into the trained deep learning model to obtain updated satellite meteorological data for the target area. Using this device, by inputting a continuous second data sample set and a discontinuous third data sample set containing default values ​​from numerical model meteorological data into a deep learning model for training, the relationship between continuous and discontinuous data is obtained. Furthermore, based on the principle of relational transfer learning, discontinuous satellite meteorological data can be input into the trained deep learning model to obtain complete spatiotemporal continuous satellite meteorological data, reducing the difficulty of acquiring continuous satellite meteorological data and improving the efficiency of acquiring continuous satellite meteorological data.

[0128] Specific limitations regarding the meteorological data acquisition device can be found in the limitations on meteorological data acquisition methods described above, and will not be repeated here. Each module in the aforementioned meteorological data acquisition device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0129] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 7 As shown, the computer device includes a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a meteorological data acquisition method. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0130] Those skilled in the art will understand that Figure 7 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0131] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0132] Acquire satellite meteorological data for a target area within a target time frame;

[0133] The target area is divided into grids, and the satellite meteorological data corresponding to each grid point in the target area grid is determined to obtain the first data sample set, which contains the default values ​​of satellite meteorological data.

[0134] Acquire numerical model meteorological data, and determine the numerical model meteorological data corresponding to each grid point in the grid of the target area according to the same grid resolution as the grid division of the target area, to obtain the second data sample set;

[0135] Based on the correspondence between the first data sample set and the second data sample set and the default values ​​of satellite meteorological data, the numerical model meteorological data corresponding to the target grid points in the second data sample set are updated to the preset default values ​​to obtain the third data sample set;

[0136] The second and third data sample sets are input into a pre-set deep learning model to train the deep learning model.

[0137] The satellite meteorological data of the target area is input into the trained deep learning model to obtain the updated satellite meteorological data of the target area.

[0138] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0139] The target area is divided into grids to obtain the statistical range of satellite meteorological data corresponding to each grid point;

[0140] When there is no satellite meteorological data within the statistical range of satellite meteorological data, the data corresponding to the grid points included in the statistical range will be determined as the default value of satellite meteorological data.

[0141] When satellite meteorological data exists within the statistical range of satellite meteorological data, the data corresponding to the grid points included in the statistical range are determined as the average value of the satellite meteorological data within the statistical range;

[0142] The satellite meteorological data corresponding to each grid point in the target area grid is used as the first data sample set.

[0143] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0144] Based on the preset resolution corresponding to the numerical model gridding, the numerical model meteorological data corresponding to each grid point in the numerical model grid are calculated.

[0145] Based on the same grid resolution as the target area, the numerical model meteorological data are mapped to each grid point of the target area's grid resolution using spatial interpolation, resulting in a second data sample set of numerical model meteorological data based on the target area's grid resolution.

[0146] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0147] Based on the correspondence between satellite meteorological data at each grid point in the first data sample set and numerical model meteorological data at each grid point in the second data sample set, the grid points in the second data sample set whose numerical model meteorological data correspond to the default values ​​of satellite meteorological data in the first data sample set are determined as target grid points.

[0148] Using the preset default value as a mask, the numerical model meteorological data corresponding to the target grid point in the second data sample set are assigned the preset default value to obtain the third data sample set.

[0149] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0150] Based on the same gridding resolution as the target area's grid, the target area's grid is projected onto the same latitude and longitude to obtain the projected grid.

[0151] Based on the latitude and longitude information of each grid point in the projection grid, the numerical model meteorological data of the corresponding grid points with the same latitude and longitude in the numerical model grid are spatially interpolated to the grid points of the projection grid to obtain the second data sample set of numerical model meteorological data based on the grid resolution of the target area.

[0152] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0153] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0154] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0155] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for acquiring meteorological data, characterized in that, The method includes: Acquire satellite meteorological data for a target area within a target time range; the satellite meteorological data is aerosol optical thickness data. The target area is divided into grids, and the satellite meteorological data corresponding to each grid point in the grid of the target area is determined to obtain a first data sample set, which contains default values ​​of satellite meteorological data. Numerical model meteorological data is acquired. Based on the same grid resolution as the grid division of the target area, the numerical model meteorological data corresponding to each grid point in the grid of the target area is determined to obtain a second data sample set. The numerical model meteorological data is calculated by a preset numerical model model. The numerical model meteorological data is continuous meteorological data. Based on the correspondence between the first data sample set and the second data sample set and the default value of the satellite meteorological data, the numerical model meteorological data corresponding to the target grid point in the second data sample set is updated to the preset default value to obtain the third data sample set; The second data sample set and the third data sample set are input into a preset deep learning model to train the deep learning model; The satellite meteorological data of the target area is input into the trained deep learning model to obtain the updated satellite meteorological data of the target area.

2. The method according to claim 1, characterized in that, The process involves dividing the target area into grids, determining the satellite meteorological data corresponding to each grid point in the target area grid, and obtaining a first data sample set. This first data sample set includes default values ​​for satellite meteorological data, including: The target area is divided into grids to obtain the statistical range of the satellite meteorological data corresponding to each grid point; When there is no satellite meteorological data within the statistical range of the satellite meteorological data, the data corresponding to the grid points included in the statistical range will be determined as the default value of the satellite meteorological data. When satellite meteorological data exists within the statistical range of the satellite meteorological data, the data corresponding to the grid points included in the statistical range are determined as the average value of the satellite meteorological data within the statistical range; The satellite meteorological data corresponding to each grid point in the target area grid is used as the first data sample set.

3. The method according to claim 1, characterized in that, The acquisition of numerical model meteorological data involves determining the numerical model meteorological data corresponding to each grid point in the grid of the target area, based on the same grid resolution as the grid division of the target area, to obtain a second data sample set, including: Based on the preset resolution corresponding to the numerical model gridding, the numerical model meteorological data corresponding to each grid point in the numerical model grid are calculated. Based on the same grid resolution as the target area, the numerical model meteorological data is mapped to grid points with the same grid resolution as the target area using spatial interpolation, thus obtaining a second data sample set of numerical model meteorological data with the same grid resolution as the target area.

4. The method according to claim 1, characterized in that, The step involves updating the numerical model meteorological data corresponding to the target grid points in the second data sample set to preset default values ​​based on the correspondence between the first data sample set and the second data sample set, and the default values ​​of the satellite meteorological data, to obtain the third data sample set, which includes: Based on the correspondence between satellite meteorological data at each grid point in the first data sample set and numerical model meteorological data at each grid point in the second data sample set, the grid points in the second data sample set that correspond to the default values ​​of the satellite meteorological data in the first data sample set are determined as target grid points. Using the preset default value as a mask, in the second data sample set, the numerical model meteorological data corresponding to the target grid point is assigned the preset default value to obtain the third data sample set.

5. The method according to claim 3, characterized in that, The step involves mapping the numerical model meteorological data to grid points with the same grid resolution as the target area using spatial interpolation, thereby obtaining a second data sample set of numerical model meteorological data with the same grid resolution as the target area. This second data sample set includes: Based on the same gridding resolution as the target region grid, the target region grid is projected with equal latitude and longitude to obtain a projected grid. Based on the latitude and longitude information of each grid point in the projection grid, the numerical model meteorological data of the corresponding grid points in the numerical model grid with the same latitude and longitude are spatially interpolated onto the grid points of the projection grid to obtain a second data sample set of numerical model meteorological data based on the grid resolution of the target area.

6. A meteorological data acquisition device, characterized in that, The device includes: The acquisition module is used to acquire satellite meteorological data of a target area within a target time range; the satellite meteorological data is aerosol optical thickness data. The first determining module is used to divide the target area into grids, determine the satellite meteorological data corresponding to each grid point in the target area grid, and obtain a first data sample set, wherein the first data sample set contains default values ​​of satellite meteorological data. The second determining module is used to acquire numerical model meteorological data, and determine the numerical model meteorological data corresponding to each grid point in the grid of the target area according to the same grid resolution as the grid division of the target area, so as to obtain a second data sample set; the numerical model meteorological data is calculated by a preset numerical model model; the numerical model meteorological data is continuous meteorological data. The third determining module is used to update the numerical model meteorological data corresponding to the target grid point in the second data sample set to the preset default value according to the correspondence between the first data sample set and the second data sample set and the default value of the satellite meteorological data, so as to obtain the third data sample set. The training module is used to input the second data sample set and the third data sample set into a preset deep learning model to train the deep learning model; The update module is used to input the satellite meteorological data of the target area into the trained deep learning model to obtain the updated satellite meteorological data of the target area.

7. The apparatus according to claim 6, characterized in that, The second determining module is specifically used to calculate the numerical model meteorological data corresponding to each grid point in the numerical model grid according to the preset resolution of the numerical model grid. Based on the same grid resolution as the target area, the numerical model meteorological data is mapped to grid points with the same grid resolution as the target area using spatial interpolation, thus obtaining a second data sample set of numerical model meteorological data with the same grid resolution as the target area.

8. The apparatus according to claim 6, characterized in that, The third determining module is specifically used to determine the grid points in the second data sample set that correspond to the default values ​​of the satellite meteorological data in the first data sample set as target grid points based on the correspondence between the satellite meteorological data at each grid point in the first data sample set and the numerical model meteorological data at each grid point in the second data sample set. Using the preset default value as a mask, in the second data sample set, the numerical model meteorological data corresponding to the target grid point is assigned the preset default value to obtain the third data sample set.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images