A three-dimensional water vapor tomography inversion method and system fusing beidou and fy data

Abstract

The application relates to a three-dimensional water vapor tomography inversion method and system fusing Beidou and Fengyun data, and comprises the following steps: constructing a multi-source data joint observation system integrating Beidou observation signals, Fengyun remote sensing observation signals and Fengyun occultation observation signals, wherein the Beidou observation signals are provided by Beidou satellites, and the Fengyun remote sensing observation signals and the Fengyun occultation observation signals are provided by Fengyun satellites; and establishing and solving a joint observation equation set to obtain a three-dimensional water vapor field. Firstly, the multi-source data joint observation system is constructed, the Beidou observation signals guarantee a basic geometric framework and a high-precision benchmark required by tomography; the Fengyun remote sensing observation signals provide fine structure information in the horizontal direction of the atmosphere; and the Fengyun occultation observation signals provide a high-vertical-resolution water vapor profile as a strong constraint in the vertical direction of tomography. Then, the three-dimensional water vapor field is obtained by establishing and solving the joint observation equation set, and the geometric configuration of the observation signals is fundamentally optimized, so that the inversion precision is effectively improved.

Description

Technical Field

[0001] This invention relates to the field of three-dimensional water vapor tomography inversion technology, specifically to a three-dimensional water vapor tomography inversion method and system that integrates BeiDou and wind and cloud data. Background Technology

[0002] Since its initial realization, three-dimensional water vapor tomography has become an important research branch of GNSS meteorology. Its basic principle is to use ground-based GNSS observation networks to obtain inclined path wet delay observations, construct a set of tomographic equations through three-dimensional spatial discretization, and then invert the atmospheric wet refractive index field.

[0003] Compared to single GPS observations, related technologies improve the accuracy of water vapor tomography by fusing multi-source data, such as fusing multi-constellation GNSS observation data, or fusing other satellite source data such as radio occultation data from meteorological ionospheric and climate constellation observation systems. However, due to the natural differences in the observation geometry of heterogeneous data in multi-source data, the fusion effect of existing methods is limited, affecting the accuracy of tomography. Summary of the Invention

[0004] This application provides a three-dimensional water vapor tomography inversion method and system that integrates BeiDou and wind cloud data, aiming to solve the technical problem that the fusion effect is limited and affects the tomography accuracy when multiple source data are fused in related technologies.

[0005] This application provides a three-dimensional water vapor tomography inversion method that integrates BeiDou and wind and cloud data, which includes the following steps:

[0006] Construct a multi-source data joint observation system integrating BeiDou observation signals, Fengyun remote sensing observation signals, and Fengyun occultation observation signals, wherein the BeiDou observation signals are provided by BeiDou satellites, and the Fengyun remote sensing observation signals and Fengyun occultation observation signals are provided by Fengyun satellites;

[0007] Establish and solve the joint observation equations to obtain the three-dimensional water vapor field;

[0008] The joint observation equations are expressed as follows:

[0009] ;

[0010] Among them, SWV BDS and SWV RS PWV represents the slant path delay of the BeiDou observation signal and the virtual slant path delay of the Fengyun remote sensing observation signal, respectively. OM A represents the observed water vapor value formed by the signal from the observation of occultation by wind and clouds. BDS A RS A OMThese represent the intercept data of the corresponding BeiDou observation signal, Fengyun remote sensing observation signal, and Fengyun occultation observation signal in the tomographic voxel block, respectively. A H and A V These represent the coefficient matrices corresponding to the horizontal and vertical constraints, respectively, and X represents the three-dimensional water vapor field.

[0011] In one embodiment, the virtual slant path delay (SWV) of the wind and cloud remote sensing observation signal RS The calculation process includes:

[0012] Each pixel in the wind and cloud remote sensing observation signal is regarded as a virtual station, and the water vapor content corresponding to the virtual station is used as the water vapor observation value of the pixel, forming a joint observation with the Beidou observation signal.

[0013] Calculate the virtual slant path delay (SWV) of wind and cloud remote sensing observation signals. RS ;

[0014] The calculation formula is: ;

[0015] Wherein, PWV represents the water vapor observation value formed from wind and cloud remote sensing signals. and Let these represent the wet mapping function and the gradient mapping function, respectively. Represents the horizontal gradient delay term. and These represent the azimuth and elevation angles, respectively. This represents the conversion factor.

[0016] In one implementation, the horizontal gradient delay term The calculations were performed using the tropospheric parameter difference estimation method.

[0017] In one embodiment, the water vapor observation value PWV generated by the wind occultation observation signal is... OM The calculation process includes:

[0018] Calculate the water vapor content value at the occultation observation point based on the aforementioned wind and cloud occultation observation signal;

[0019] The calculation formula is: ;

[0020] in, This represents the density of liquid water. Indicates the first i The specific humidity of the layer, g Represents gravitational acceleration. Indicates the first i Layer and first i+1 Pressure difference between layers n This indicates the number of air pressure layers.

[0021] In one embodiment, the water vapor observation value PWV generated by the wind occultation observation signal is... OM The calculation process includes:

[0022] Based on the principle of water vapor tomography using voxel block discretization, the vertical integral value of the water vapor density of the atmosphere above the occultation observation point is used as the water vapor observation value.

[0023] Discretized expression is: ;

[0024] in, and They represent the first k Water vapor density and thickness values ​​of the chromatographic voxel block.

[0025] In one implementation, before solving the joint observation equations, the grid resolution is adaptively adjusted based on the observation signal density, which includes:

[0026] For all voxels in the tomographic voxel block, a three-dimensional signal density field of the tomographic region is constructed based on the radial basis kernel function;

[0027] Sensitive layers are defined by identifying vertical layer sensitivity based on horizontal density changes.

[0028] The grid resolution of the sensitive layer is adaptively adjusted.

[0029] In one embodiment, constructing a three-dimensional signal density field of the tomographic region based on a radial basis kernel function for all voxels in the tomographic voxel block includes:

[0030] An adaptive non-uniform exponential layering method is used for discretization in the vertical direction, and a fixed resolution method is used for uniform discretization in the horizontal direction to establish an initial grid.

[0031] Based on the kernel density estimation method, the signal density per unit voxel in the chromatographic voxel block is statistically analyzed.

[0032] Constructing a three-dimensional signal density field based on radial basis kernel functions;

[0033] Expressed as: ;

[0034] in, Voxels ( x , y , h The signal density value of ) x , y , h These represent the horizontal coordinates and elevation of the voxel center, respectively. Indicates the first kThe normalized distance of the ray from the current voxel center This represents the radial basis kernel function.

[0035] In one implementation, the identification of vertical layer sensitivity based on horizontal density changes defines the sensitive layer as follows:

[0036] Calculate the relative density change index in the horizontal direction ;

[0037] The calculation formula is: ;

[0038] in, This represents the average density of all voxels in that horizontal layer;

[0039] Calculate local dispersion To determine the anisotropy of the observed signal density distribution;

[0040] The calculation formula is: ;

[0041] Set an experience threshold When the local dispersion of a certain horizontal layer When this occurs, the layer is defined as a sensitive layer.

[0042] In one implementation, adaptively adjusting the grid resolution of the sensitive layer includes:

[0043] Calculate the density gradient of each voxel in the sensitive layer to represent the change in observed signal density;

[0044] The calculation formula is: ;

[0045] Set density threshold and gradient threshold Determine whether the voxel meets the following conditions;

[0046] ;

[0047] For voxels that meet the conditions, a four-splitting method is used for encryption.

[0048] This application embodiment also provides a three-dimensional water vapor tomography inversion system that integrates BeiDou and wind cloud data, applying the three-dimensional water vapor tomography inversion method for integrating BeiDou and wind cloud data as described in any of the above claims, the system comprising:

[0049] The observation system construction module is configured to: construct a multi-source data joint observation system integrating BeiDou observation signals, Fengyun remote sensing observation signals, and Fengyun occultation observation signals, wherein the BeiDou observation signals are provided by BeiDou satellites, and the Fengyun remote sensing observation signals and Fengyun occultation observation signals are provided by Fengyun satellites;

[0050] The equation solving module is configured to establish and solve the joint observation equations to obtain the three-dimensional water vapor field.

[0051] The beneficial effects of the technical solutions provided in this application include:

[0052] This application provides a three-dimensional water vapor tomography inversion method and system that integrates BeiDou and Fengyun data. First, a multi-source data joint observation system is constructed, integrating data sources from different satellite observation technologies. BeiDou observation signals provide high-precision, high-temporal-resolution absolute water vapor content values ​​through slant path delay, ensuring the basic geometric framework and high-precision benchmark required for tomography. Fengyun remote sensing observation signals provide fine-grained atmospheric horizontal structure information through virtual slant path delay, compensating for the sparsity of GNSS ray distribution. Fengyun occultation observation signals provide high-vertical-resolution water vapor profiles through water vapor observation values, serving as a strong constraint for tomography in the vertical direction. Then, by establishing and solving the joint observation equation set—that is, by solving the three-dimensional water vapor field through the three-dimensional water vapor tomography equation—the three observation signals with different times, spatial resolutions, and advantages are unified into a single spatiotemporal framework (voxel grid), fundamentally optimizing the geometric configuration of the observation signals and effectively improving inversion accuracy. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Figure 1 This is a flowchart illustrating the steps of a three-dimensional water vapor tomography inversion method that integrates BeiDou and wind and cloud data in one embodiment of the present invention.

[0055] Figure 2 This is a schematic diagram of joint observation by Beidou satellites and Fengyun satellites in one embodiment of the present invention.

[0056] Figure 3 This is a flowchart illustrating the adaptive adjustment of grid resolution based on observed signal density in one embodiment of the present invention. Detailed Implementation

[0057] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0058] This application provides a three-dimensional water vapor tomography inversion method that integrates BeiDou and wind cloud data, aiming to solve the technical problem in related technologies where the fusion effect of multi-source data fusion is limited, affecting the tomography accuracy.

[0059] First, the meanings of some proper nouns will be explained.

[0060] Fengyun remote sensing refers to the technology of acquiring global meteorological and environmental information by using various remote sensors carried on Fengyun satellites to receive electromagnetic waves (from visible light and infrared to microwaves) reflected or radiated by the Earth and atmosphere. The observation principle is passive reception of electromagnetic wave signals. Polar-orbiting satellites can achieve global coverage, while geostationary satellites can continuously observe fixed areas for minutes, thus providing rich planar information, especially the fine structure of the atmosphere in the horizontal direction, such as cloud images, surface temperature, vegetation, aerosols, and sea ice.

[0061] Fengyun occultation is an active detection technology carried by Fengyun satellites, primarily used for precise detection of the vertical structure of the atmosphere. The observation principle is based on the refraction, delay, and bending of signals emitted by GNSS navigation satellites (such as GPS, BeiDou, and Galileo) as they pass through the Earth's atmosphere. The GNSS occultation receiver on the Fengyun satellites specifically receives these signals passing through the edge of the atmosphere (i.e., "occultation"). It features high vertical resolution, high accuracy, and global coverage, achieving vertical resolution down to hundreds of meters, clearly revealing the fine layered structure of the atmosphere. Because it is based on absolute time measurements and requires no calibration, the data accuracy is extremely high, often used as a "benchmark" for other observations. It is particularly suitable for covering regions lacking traditional observation methods, such as oceans, deserts, and polar regions.

[0062] In tomography, a tomographic voxel is a collection of regular grid cells (i.e., voxels) into which the entire three-dimensional space being studied is divided.

[0063] like Figure 1 As shown, Figure 1 This is a flowchart illustrating the steps of a three-dimensional water vapor tomography inversion method that integrates BeiDou and wind and cloud data in one embodiment of the present invention.

[0064] This embodiment provides a three-dimensional water vapor tomography inversion method that integrates BeiDou and wind cloud data, which includes the following steps:

[0065] Step S1: Construct a multi-source data joint observation system integrating BeiDou observation signals, Fengyun remote sensing observation signals, and Fengyun occultation observation signals. The BeiDou observation signals are provided by BeiDou satellites, while the Fengyun remote sensing observation signals and Fengyun occultation observation signals are provided by Fengyun satellites.

[0066] Step S2: Establish and solve the joint observation equations to obtain the three-dimensional water vapor field;

[0067] The joint observation equations are expressed as follows:

[0068] ;

[0069] Among them, SWV BDS and SWV RS PWV represents the slant path delay of the BeiDou observation signal and the virtual slant path delay of the Fengyun remote sensing observation signal, respectively. OM A represents the observed water vapor value formed by the signal from the observation of occultation by wind and clouds. BDS A RS A OM These represent the intercept data of the corresponding BeiDou observation signal, Fengyun remote sensing observation signal, and Fengyun occultation observation signal in the tomographic voxel block, respectively. A H and A V These represent the coefficient matrices corresponding to the horizontal and vertical constraints, respectively, and X represents the three-dimensional water vapor field.

[0070] This embodiment provides a three-dimensional water vapor tomography inversion method that integrates BeiDou and Fengyun data. First, a multi-source data joint observation system is constructed, integrating data sources from different satellite observation technologies. BeiDou observation signals provide high-precision, high-temporal-resolution absolute water vapor content values ​​through slant path delay, ensuring the basic geometric framework and high-precision benchmark required for tomography. Fengyun remote sensing observation signals provide fine-grained atmospheric horizontal structure information through virtual slant path delay, compensating for the sparsity of GNSS ray distribution. Fengyun occultation observation signals provide high-vertical-resolution water vapor profiles through water vapor observation values, serving as a strong constraint for tomography in the vertical direction. Then, by establishing and solving the joint observation equation set—that is, by solving the three-dimensional water vapor field through the three-dimensional water vapor tomography equation—the three observation signals with different times, spatial resolutions, and advantages are unified into a single spatiotemporal framework (voxel grid), fundamentally optimizing the geometric configuration of the observation signals and effectively improving inversion accuracy.

[0071] In one embodiment, the virtual slant path delay (SWV) of the wind and cloud remote sensing observation signal RS The calculation process includes:

[0072] Each pixel in the Fengyun remote sensing observation signal is regarded as a virtual station, and the water vapor content corresponding to the virtual station is used as the water vapor observation value of that pixel, forming a joint observation with the Beidou observation signal.

[0073] Specifically, such as Figure 2 As shown, Figure 2 This is a schematic diagram of joint observation by BeiDou and Fengyun satellites in one embodiment of the present invention. Four BeiDou satellites are located at the top, and one Fengyun satellite is located at the bottom. Both BeiDou and Fengyun satellites transmit signals to the Earth's surface. The Fengyun remote sensing observation signal is considered a "pseudo" signal transmitted by the Fengyun satellite to the Earth's surface. Each pixel is simulated as a virtual station, and the water vapor content corresponding to each virtual station is approximately equivalent to the observed water vapor value of that pixel. Since the observation format of the Fengyun remote sensing signal differs from that of GNSS, joint observation is formed by simulating a virtual station and combining it with the BeiDou observation signal to increase the number of observations and ensure the stability and accuracy of the tomographic equations, i.e., the joint observation equation set.

[0074] Calculate the virtual slant path delay (SWV) of wind and cloud remote sensing observation signals. RS ;

[0075] The calculation formula is: ;

[0076] Wherein, PWV represents the water vapor observation value formed from wind and cloud remote sensing signals. and Let these represent the wet mapping function and the gradient mapping function, respectively. Represents the horizontal gradient delay term. and These represent the azimuth and elevation angles, respectively. This represents the conversion factor.

[0077] Specifically, the virtual slant path delay (SWV) of the Fengyun remote sensing observation signal. RS This also represents the oblique path water vapor content value corresponding to each virtual station. Fengyun remote sensing signals include the elevation and azimuth angle information from the sensors on the Fengyun satellite to each surface pixel.

[0078] The formula for calculating the conversion factor is:

[0079] ;

[0080] in, This represents the weighted average temperature of the atmosphere. This represents the density of liquid water. This represents the contribution factor of dry air to the total refractive index of the atmosphere. and These represent two contribution factors of atmospheric water vapor to the total refractive index of the atmosphere. and Let represent the dry and wet gas constants, respectively, with values ​​of 287 J·(K·kg). -1 And 461.5 J·(K·kg) -1 .

[0081] In one embodiment, the horizontal gradient delay term The calculations were performed using the tropospheric parameter difference estimation method.

[0082] The core idea is to treat the horizontal gradient as an unknown parameter, solving for it along with parameters such as receiver coordinates and zenith delay in the data processing adjustment model. Specifically, the calculation process is as follows:

[0083] (1) Establish a three-dimensional wet refractive index field

[0084] Based on the global atmospheric reanalysis grid product released by the National Meteorological Information Center, the wet refractive index Nw of each grid point was calculated.

[0085] (2) Convert the grid point potential to normal height

[0086] The height information provided by the reanalysis grid product is the geopotential value corresponding to different pressure layers. It is converted into normal height to obtain the elevation information of the corresponding area, and then the meteorological data in the reanalysis grid product is used.

[0087] Convert potential value to potential height: ;

[0088] in, and These represent potential and high potential, respectively. This represents the gravitational acceleration at the grid point.

[0089] Convert high potential to positive potential: ;

[0090] in, This represents the latitude value at a grid point. and These represent the effective radius of the Earth and the gravitational acceleration on the surface of the Earth's ellipsoid at the grid point, respectively. This represents the gravitational acceleration at a grid point when the latitude is 45 degrees.

[0091] Combining geoid height N and elevation anomaly values ​​provided by a global ultra-high-order model Convert the positive height to the normal height: .

[0092] (3) Constructing virtual signal lines

[0093] For each pixel in the wind and cloud remote sensing signal, several inclined signal lines originating from that pixel are constructed, with different elevation and azimuth angles set. Based on the previous three-dimensional wet refractive index field, the inclined path wet delay of each signal is calculated:

[0094] The calculation formula is: ;

[0095] Where SWD represents the slant path wet delay for each signal. and They represent the first i The atmospheric wet refractive index value of the layer and the length information of the observation signal.

[0096] (4) Solve the atmospheric horizontal gradient value pixel by pixel

[0097] According to the higher-order atmospheric horizontal gradient model, the slant path delay of each signal is represented as follows:

[0098] ;

[0099] in, and This represents the first-order horizontal gradient terms in the north-south and east-west directions of the station. and This represents the corresponding second-order horizontal gradient term.

[0100] By subtracting the delays of two slanted paths with the same elevation angle signal, we obtain the difference equation shown below:

[0101] ;

[0102] By pairwise combining observation signals with the same elevation angle in the virtual signal line of each pixel to form multiple difference equations, and then solving them pixel by pixel using the least squares method, the atmospheric horizontal gradient value at each pixel can be obtained, i.e., the horizontal gradient delay term. .

[0103] The above scheme calculates the horizontal gradient delay term using the tropospheric parameter difference estimation method, effectively correcting the influence of atmospheric horizontal inhomogeneity on signal path delay. This results in more accurate virtual oblique path delay of wind and cloud remote sensing signals, eliminating most of the systematic errors caused by horizontal inhomogeneity, improving the quality of the original data, enhancing the tomographic inversion effect, and thus obtaining a more realistic three-dimensional water vapor field.

[0104] In one embodiment, the water vapor observation value PWV formed by the wind occultation observation signal is... OM The calculation process includes:

[0105] Calculate the water vapor content at the occultation observation point based on the occultation observation signal from the wind and clouds;

[0106] The calculation formula is: ;

[0107] in, This represents the density of liquid water. Indicates the first i The specific humidity of the layer, g Represents gravitational acceleration. Indicates the first i Layer and first i+1 Pressure difference between layers n This indicates the number of air pressure layers.

[0108] Through the above scheme, the wet atmospheric profile product of the Fengyun satellite occultation detector provides the wet atmospheric profile and related auxiliary data for a single occultation event, including the vertical profile data of air pressure, temperature and water vapor pressure at the occultation observation point, thereby calculating high-precision water vapor observation values, which serve as additional constraints to improve the accuracy of three-dimensional water vapor tomography.

[0109] In one embodiment, the water vapor observation value PWV formed by the wind occultation observation signal is... OM The calculation process includes:

[0110] Based on the principle of water vapor tomography using voxel block discretization, the vertical integral value of the water vapor density of the atmosphere above the occultation observation point is used as the water vapor observation value.

[0111] Discretized expression is: ;

[0112] in, and They represent the first k Water vapor density and thickness values ​​of the chromatographic voxel block.

[0113] In related technologies, tropospheric three-dimensional tomography generally employs a voxel-based method, and the regional grid structure directly affects the accuracy of parameter calculation. However, most water vapor tomography grid division methods are based on pre-defined layers using single prior features of vertical water vapor variation or external meteorological models. When the grid density does not match the density distribution of the observed signal, it cannot fundamentally improve the ill-conditioned equations caused by signal sparsity, thus limiting further improvement in overall inversion accuracy. Therefore, in this embodiment, the grid resolution is adaptively adjusted based on the observed signal density before solving the joint observation equations.

[0114] like Figure 3 As shown, Figure 3 This is a flowchart illustrating the adaptive adjustment of grid resolution based on observed signal density in one embodiment of the present invention.

[0115] In one embodiment, before solving the joint observation equations, the grid resolution is adaptively adjusted based on the observation signal density, which includes:

[0116] Step S201: For all voxels in the tomographic voxel block, construct a three-dimensional signal density field of the tomographic region based on the radial basis kernel function;

[0117] Step S202: Identify the sensitivity of the vertical layer based on the density change in the horizontal direction, and define the sensitive layer;

[0118] Step S203: Adaptively adjust the grid resolution of the sensitive layer.

[0119] The above scheme achieves dynamic resolution by dynamically adjusting the discretization degree of the tomographic region. In areas with dense signals and sufficient information, the grid is densified to improve resolution and finely depict water vapor details. At the same time, in areas with sparse signals and insufficient information, a coarser grid is used to reduce resolution, so that the density structure of the grid and the density distribution of the observed signal are optimally matched. This improves the accuracy of tomography and effectively avoids the ill-conditioned problem of the equation.

[0120] In one embodiment, step S201, constructing a three-dimensional signal density field of the tomographic region based on the radial basis kernel function for all voxels in the tomographic voxel block, includes:

[0121] Step S2011: Discretize using adaptive non-uniform exponential layering in the vertical direction and uniform discretize using fixed resolution in the horizontal direction to establish an initial grid.

[0122] Specifically, discretization is performed in both the vertical and horizontal directions to facilitate the layered calculation of differences in the observed signal density.

[0123] Step S2012: Based on the kernel density estimation method, the signal density within a unit voxel in the chromatographic voxel block is statistically analyzed.

[0124] Specifically, after statistically analyzing the signal density using the kernel density estimation method, the signal penetration paths of BeiDou satellites and Fengyun satellites are transformed into quantifiable three-dimensional signal density distribution data.

[0125] Step S2013: Construct a three-dimensional signal density field based on the radial basis kernel function;

[0126] Expressed as: ;

[0127] in, Voxels ( x , y , h The signal density value of ) x , y , hThese represent the horizontal coordinates and elevation of the voxel center, respectively. Indicates the first k The normalized distance of the ray from the current voxel center This represents the radial basis kernel function.

[0128] Using the above scheme, a three-dimensional signal density field is constructed in the tomographic region. The kernel density estimation method smoothly diffuses the influence range of each discrete satellite ray path into the surrounding three-dimensional space, forming a continuous and differentiable density field, which can effectively reduce the spurious structure caused by mesh discretization. The radial basis kernel function naturally expresses the physical intuition that the closer the distance, the greater the influence, ensuring a smooth transition of the density field and laying a good foundation for subsequent gradient calculations.

[0129] In one embodiment, step S202, identifying the sensitivity of the vertical layer based on the density change in the horizontal direction, and defining the sensitive layer includes:

[0130] Step S2021: Calculate the relative density change index in the horizontal direction. ;

[0131] The calculation formula is: ;

[0132] in, This represents the average density of all voxels in that horizontal layer;

[0133] Step S2022: Calculate the local dispersion To determine the anisotropy of the observed signal density distribution;

[0134] The calculation formula is: ;

[0135] Step S2023: Set the experience threshold When the local dispersion of a certain horizontal layer When this occurs, the layer is defined as a sensitive layer.

[0136] Specifically, an empirical threshold of 1.8 can be set to determine whether the layer has significant density differences; when the local dispersion of a certain horizontal layer... When the signal density distribution within the layer is significantly uneven, this indicates that the layer is defined as a sensitive layer.

[0137] The above scheme quantifies the density variation in the horizontal direction, further optimizing the grid resolution. The relative density variation index eliminates the influence of absolute density values, making density variations between different height layers comparable. Local dispersion cleverly combines absolute and relative differences. Finally, the relationship between an empirical threshold and local dispersion is used to perform a coarse screening in the vertical direction, identifying the horizontal layers with greater non-uniformity.

[0138] In one embodiment, step S203, adaptively adjusting the grid resolution of the sensitive layer, includes:

[0139] Step S2031: Calculate the density gradient of each voxel in the sensitive layer to represent the change in observed signal density;

[0140] The calculation formula is: ;

[0141] Step S2032: Set density threshold and gradient threshold Determine whether the voxel meets the following conditions;

[0142] ;

[0143] Step S2033: For voxels that meet the conditions, use the four-splitting method for encryption.

[0144] It should be noted that encryption will not be performed if the conditions are not met. For each sub-voxel, if it still meets the density threshold or gradient threshold, the four-splitting operation continues until the minimum resolution is reached.

[0145] The above scheme quantifies the changes in signal density in the identified sensitive layers. For areas with high signal density or drastic density changes, the grid resolution is finely adjusted. Specifically, a density threshold is set to capture dense signal areas, which have rich observational information, and increasing the resolution may reveal more detailed structures. A gradient threshold is set to capture boundary areas with drastic changes in signal density, which are often the boundaries of different air masses and weather systems. Increasing the resolution in these areas is crucial for characterizing fronts and convection triggering. The four-splitting method is the standard operation for structured adaptive grid refinement, which is simple and efficient. Recursive refinement ensures that the final grid can meticulously depict all key structures.

[0146] This embodiment also provides a three-dimensional water vapor tomography inversion system that integrates BeiDou and wind cloud data. Applying the above-mentioned three-dimensional water vapor tomography inversion method, the system includes:

[0147] The observation system construction module is configured to: construct a multi-source data joint observation system integrating BeiDou observation signals, Fengyun remote sensing observation signals, and Fengyun occultation observation signals, wherein the BeiDou observation signals are provided by BeiDou satellites, and the Fengyun remote sensing observation signals and Fengyun occultation observation signals are provided by Fengyun satellites;

[0148] The equation solving module is configured to establish and solve the joint observation equations to obtain the three-dimensional water vapor field.

[0149] The functions of the relevant modules have been described above and will not be repeated here.

[0150] In the description of this application, it should be noted that the terms "upper," "lower," etc., indicating the orientation or positional relationship are based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application. Unless otherwise expressly specified and limited, the terms "installed," "connected," and "linked" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication between two elements. For those skilled in the art, the specific meaning of the above terms in this application can be understood according to the specific circumstances.

[0151] It should be noted that in this application, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0152] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0153] The above are merely specific embodiments of this application, enabling those skilled in the art to understand or implement this application. 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 this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A three-dimensional water vapor tomography inversion method integrating BeiDou and wind and cloud data, characterized in that, It includes the following steps: Construct a multi-source data joint observation system integrating BeiDou observation signals, Fengyun remote sensing observation signals, and Fengyun occultation observation signals, wherein the BeiDou observation signals are provided by BeiDou satellites, and the Fengyun remote sensing observation signals and Fengyun occultation observation signals are provided by Fengyun satellites; Establish and solve the joint observation equations to obtain the three-dimensional water vapor field; The joint observation equations are expressed as follows: ； Among them, SWV BDS and SWV RS PWV represents the slant path delay of the BeiDou observation signal and the virtual slant path delay of the Fengyun remote sensing observation signal, respectively. OM A represents the observed water vapor value formed by the signal from the observation of occultation by wind and clouds. BDS A RS A OM These represent the intercept data of the corresponding BeiDou observation signal, Fengyun remote sensing observation signal, and Fengyun occultation observation signal in the tomographic voxel block, respectively. A H and A V These represent the coefficient matrices corresponding to the horizontal and vertical constraints, respectively, and X represents the three-dimensional water vapor field. Among them, the virtual slant path delay (SWV) of the wind and cloud remote sensing observation signal RS The calculation process includes: Each pixel in the wind and cloud remote sensing observation signal is regarded as a virtual station, and the virtual slant path delay (SWV) of the wind and cloud remote sensing observation signal is... RS The water vapor content along the oblique path corresponding to each virtual station is used as the water vapor observation value of that pixel, and is used in conjunction with the BeiDou observation signal to form a joint observation. Calculate the virtual slant path delay (SWV) of wind and cloud remote sensing observation signals. RS ; The calculation formula is: ; Wherein, PWV represents the water vapor observation value formed from wind and cloud remote sensing signals. and Let these represent the wet mapping function and the gradient mapping function, respectively. Represents the horizontal gradient delay term. and These represent the azimuth and elevation angles, respectively. This represents the conversion factor.

2. The three-dimensional water vapor tomography inversion method integrating BeiDou and wind and cloud data as described in claim 1, characterized in that, The horizontal gradient delay term The calculations were performed using the tropospheric parameter difference estimation method.

3. The three-dimensional water vapor tomography inversion method fusing BeiDou and wind and cloud data as described in claim 1, characterized in that, The water vapor observation value PWV formed by the occultation observation signal of the wind cloud OM The calculation process includes: Calculate the water vapor content value at the occultation observation point based on the aforementioned wind and cloud occultation observation signal; The calculation formula is: ; in, This represents the density of liquid water. Indicates the first i The specific humidity of the layer, g Represents gravitational acceleration. Indicates the first i Layer and first i+1 Pressure difference between layers n Indicates the number of air pressure layers.

4. The three-dimensional water vapor tomography inversion method integrating BeiDou and wind and cloud data as described in claim 1, characterized in that, The water vapor observation value PWV formed by the occultation observation signal of the wind cloud OM The calculation process includes: Based on the principle of water vapor tomography using voxel block discretization, the vertical integral value of the water vapor density of the atmosphere above the occultation observation point is used as the water vapor observation value. Discretized expression: ; in, and They represent the first k Water vapor density and thickness values ​​of the chromatographic voxel block.

5. The three-dimensional water vapor tomography inversion method fusing BeiDou and wind and cloud data as described in claim 1, characterized in that, Before solving the joint observation equations, the grid resolution is adaptively adjusted based on the observed signal density, which includes: For all voxels in the tomographic voxel block, a three-dimensional signal density field of the tomographic region is constructed based on the radial basis kernel function; Sensitive layers are defined by identifying vertical layer sensitivity based on horizontal density changes. The grid resolution of the sensitive layer is adaptively adjusted.

6. The three-dimensional water vapor tomography inversion method fusing BeiDou and wind and cloud data as described in claim 5, characterized in that, The construction of the three-dimensional signal density field of the tomographic region based on the radial basis kernel function for all voxels in the tomographic voxel block includes: An adaptive non-uniform exponential layering method is used for discretization in the vertical direction, and a fixed resolution method is used for uniform discretization in the horizontal direction to establish an initial grid. Based on the kernel density estimation method, the signal density per unit voxel in the chromatographic voxel block is statistically analyzed. Constructing a three-dimensional signal density field based on radial basis kernel functions; Expressed as: ; in, Voxel ( x , y , h The signal density value of ) x , y , h These represent the horizontal coordinates and elevation of the voxel center, respectively. Indicates the first k The normalized distance of the ray from the current voxel center This represents the radial basis kernel function.

7. The three-dimensional water vapor tomography inversion method fusing BeiDou and wind and cloud data as described in claim 6, characterized in that, The method for identifying vertical layer sensitivity based on horizontal density changes defines the sensitive layer as follows: Calculate the relative density change index in the horizontal direction ; The calculation formula is: ; in, This represents the average density of all voxels in that horizontal layer; Calculate local dispersion To determine the anisotropy of the observed signal density distribution; The calculation formula is: ; Set an experience threshold When the local dispersion of a certain horizontal layer When this occurs, the layer is defined as a sensitive layer.

8. The three-dimensional water vapor tomography inversion method fusing BeiDou and wind and cloud data as described in claim 7, characterized in that, The adaptive adjustment of the grid resolution for the sensitive layer includes: Calculate the density gradient of each voxel in the sensitive layer to represent the change in observed signal density; The calculation formula is: ; Set density threshold and gradient threshold Determine whether the voxel meets the following conditions; ； For voxels that meet the conditions, a four-splitting method is used for encryption.

9. A three-dimensional water vapor tomography inversion system integrating BeiDou and wind and cloud data, characterized in that, The system, employing the three-dimensional water vapor tomography inversion method fusing BeiDou and wind and cloud data as described in any one of claims 1 to 8, comprises: The observation system construction module is configured to: construct a multi-source data joint observation system integrating BeiDou observation signals, Fengyun remote sensing observation signals, and Fengyun occultation observation signals, wherein the BeiDou observation signals are provided by BeiDou satellites, and the Fengyun remote sensing observation signals and Fengyun occultation observation signals are provided by Fengyun satellites; The equation solving module is configured to establish and solve the joint observation equations to obtain the three-dimensional water vapor field.

Images