Fusion surface rainfall prediction method

Abstract

The invention provides a fused area rainfall prediction method. The fused area rainfall prediction method comprises the following steps: acquiring environmental parameters around an observation station and satellite signals on each satellite signal link, calculating a rainfall attenuation value, calculating rainfall information of each satellite signal link, and establishing a high-temporal-spatial-resolution two-dimensional rainfall distribution field in a regional range; wet delay information of each satellite signal path is extracted based on GNSS resolving software, a three-dimensional water vapor distribution field constraint model is constructed, three-dimensional water vapor distribution information is obtained through inversion, and the three-dimensional water vapor distribution information is converted into a two-dimensional water vapor distribution field; and inputting the two-dimensional rainfall distribution field, the two-dimensional water vapor distribution field and the meteorological data into the areal rainfall prediction model, and outputting an areal rainfall prediction result. According to the invention, through the innovative combination of calculating the rainfall attenuation amount by using the satellite link signal, inverting the rainfall amount according to the rainfall attenuation amount and resolving the water vapor field by GNSS chromatography, the accuracy and coverage range of rainfall amount monitoring and early warning are greatly improved, and meanwhile, the method has the advantages of low cost and efficient operation.

Description

Technical Field

[0001] The present invention relates to the field of rainfall prediction, and more specifically, to a fusion surface rainfall prediction method. Background Art

[0002] Rainfall is an important driving factor of the basin and regional hydrological cycle, and is also one of the key basic data for water resources management and scheduling. Accurate rainfall monitoring and early warning are of great significance to agricultural production, water conservancy development, human life and travel, and disaster prevention and mitigation. However, traditional rainfall monitoring methods have limitations in spatial coverage and timeliness, and it is difficult to meet the needs of rapid and refined early warning.

[0003] In the context of the rapid development of the information society, radio technology has been widely used as a core means of information acquisition and transmission. However, the complexity and variability of the natural environment, especially severe meteorological conditions (such as sandstorms, clouds, rain, snow, fog and other hydrometeors), pose a severe challenge to radio communications. In Ka / Ku band satellite communications, the meteorological environment in the troposphere causes significant attenuation of the communication link signal, among which the rain attenuation caused by rainfall is particularly prominent and significantly increases with the increase of frequency. In response to this problem, domestic and foreign scholars have conducted in-depth research and proposed a variety of anti-rain attenuation technologies and prediction models. Among them, the ITU-R model is widely used, and the satellite beacon method compares the satellite signal frequency under rainy conditions and sunny conditions, measures the rain attenuation value, and combines the rain gauge to record the rainfall rate, further establishing the relationship between rain attenuation and rainfall rate, which provides an important theoretical basis for the research of anti-rain attenuation technology. However, research based on rain attenuation measurement for surface rainfall monitoring and early warning is still relatively limited.

[0004] Traditional surface rainfall monitoring and early warning methods have made some progress. For example, meteorological stations are equipped with rain gauges and rainfall intensity meters to provide real-time precipitation data; satellite remote sensing, airborne sensors, etc. can provide large-scale precipitation information; ground radar systems can monitor high-temporal and spatial resolution changes in local precipitation. For water vapor detection, equipment such as radiosonde balloons and microwave radiometers are mainly used to obtain high-precision data in a small area. However, these methods still have room for improvement in terms of spatial coverage, timeliness, and cost-effectiveness. Summary of the invention

[0005] Aiming at the research gap of the prior art, the present invention provides a fusion surface rainfall prediction method, which fills the research gap of the prior art.

[0006] The present invention provides a fusion surface rainfall prediction method, comprising: Obtain the environmental parameters around the observation site and the satellite signals of each satellite signal link, and calculate the rainfall attenuation value of each satellite signal link; According to the rainfall attenuation value of each satellite signal link, the rainfall information of each satellite signal link is inverted; Based on the rainfall information of each satellite signal link, a two-dimensional rainfall distribution field with high temporal and spatial resolution is established within the regional scope; Based on GNSS solution software, the wet delay information of each satellite signal path is extracted to construct a three-dimensional water vapor distribution field constraint model; Inverting the three-dimensional water vapor distribution field constraint model to obtain three-dimensional water vapor distribution information, and converting the three-dimensional water vapor distribution information into a two-dimensional water vapor distribution field; The two-dimensional rainfall distribution field, the two-dimensional water vapor distribution field and meteorological data are input into a surface rainfall prediction model, and a surface rainfall prediction result is output.

[0007] The present invention provides a fusion surface rainfall prediction method, which obtains environmental parameters around the measuring station and satellite signals on each satellite signal link, calculates rainfall attenuation values, infers rainfall information for each satellite signal link, and establishes a two-dimensional rainfall distribution field with high temporal and spatial resolution within the regional range; extracts wet delay information of each satellite signal path based on GNSS solution software, constructs a three-dimensional water vapor distribution field constraint model, inverts three-dimensional water vapor distribution information, and converts it into a two-dimensional water vapor distribution field; inputs the two-dimensional rainfall distribution field, the two-dimensional water vapor distribution field and meteorological data into the surface rainfall prediction model, and outputs the surface rainfall prediction result. The present invention uses satellite link signals to calculate rain attenuation, and innovatively combines rainfall inversion based on rain attenuation with GNSS tomography to solve the water vapor field, which not only greatly improves the accuracy and coverage of rainfall monitoring and early warning, but also has the advantages of low cost and high efficiency operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 A flow chart of a fusion surface rainfall prediction method provided by the present invention; Figure 2 It is the overall flow chart of the fusion surface rainfall prediction method; Figure 3 A block diagram of a fusion surface rainfall prediction system provided by the present invention. DETAILED DESCRIPTION

[0009] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. In addition, the technical features in the various embodiments or single embodiments provided by the present invention can be arbitrarily combined with each other to form a feasible technical solution. This combination is not subject to the constraints of the sequence of steps and / or the structural composition mode, but must be based on the ability of ordinary technicians in this field to achieve. When the combination of technical solutions is contradictory or cannot be achieved, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0010] In view of the research gap in the background technology, the present invention is based on the strong correlation between rain attenuation and rainfall rate, and measures the rain attenuation on the satellite signal link to invert the surface rainfall information, establish a two-dimensional rainfall field with high temporal and spatial resolution, and provide large-scale and high-precision data support for surface rainfall warning. In addition, water vapor, as an important greenhouse gas in the atmosphere, is the material basis of weather events such as rainfall, and its temporal and spatial changes and transportation directly affect the occurrence of extreme weather (such as drought and rainstorms). Compared with traditional water vapor detection methods (such as radiosonde balloons, microwave radiometers and satellite remote sensing), GNSS water vapor detection technology has outstanding advantages such as high precision, high temporal and spatial resolution, no weather influence and low cost. The present invention obtains the water vapor delay information of the GNSS site and the satellite signal link by processing the satellite navigation signal message information, and combines the tomography technology to reconstruct a large-scale water vapor field with high precision and high temporal and spatial resolution, and provides key water vapor temporal and spatial change characteristic information for surface rainfall warning.

[0011] The rainfall prediction provided by the present invention mainly includes three parts: (1) In terms of rain attenuation and rainfall inversion, the signal attenuation information caused by rainfall is extracted based on the rain attenuation principle, and the linear average rainfall intensity is obtained by combining the ITU-R model inversion, thereby realizing the accurate calculation of rainfall. On this basis, a two-dimensional rainfall distribution field is further constructed to provide reliable rainfall data support for rainfall monitoring and early warning. (2) In terms of GNSS water vapor tomography, the GNSS satellite navigation message solution data is used, combined with the tomography algorithm to reconstruct a two-dimensional water vapor field with high temporal and spatial resolution, providing global water vapor distribution characteristics for rainfall early warning. (3) In terms of multi-source meteorological parameter fusion, the two-dimensional rainfall field and the two-dimensional water vapor field are linked and applied to surface rainfall monitoring and early warning in combination with multi-source meteorological data such as temperature, humidity, and pressure.

[0012] Figure 1 A flow chart of a fusion surface rainfall prediction method provided by the present invention is as follows: Figure 1 and Figure 2 As shown, the method includes: Step 1: Obtain the environmental parameters around the measuring station and the satellite signals on each satellite signal link, and calculate the rainfall attenuation value on each satellite signal link.

[0013] In a possible implementation manner of the present invention, the step of acquiring environmental parameters around the observation site and the satellite signal of each satellite signal link, and calculating the rainfall attenuation value of each satellite signal link includes: Step 11: Calculate the atmospheric attenuation rate γ under clear sky conditions based on the environmental parameters around the observation site. R, .

[0014] Among them, the atmospheric attenuation rate γ under clear sky conditions is calculated based on the environmental parameters around the observation site. R ,include: Obtain the ambient temperature information and relative humidity information around the observation site; According to the ambient temperature information, the saturated water vapor pressure is calculated using the calculation formula of saturated water vapor pressure: ; in, E Indicates the saturated water vapor pressure at the current temperature, in hPa; t It is the current ambient temperature information in °C.

[0015] , Then, the water content density r0 is calculated by combining the ambient temperature information and the ambient relative humidity information. The calculation formula is: ; Among them, r0 is the water content density, the unit is g / m3; h is the relative humidity information of the environment, the unit is %; T is the Kelvin temperature, T=t+273.16, the unit is K.

[0016] Next, the atmospheric attenuation model calculation code ITU-R P.676 provided by ITU-R is used. Its calling function form is itu676a (f, ph, T, r0) to calculate the atmospheric attenuation rate γ under clear sky conditions. R , in dB / km. Input parameters include: satellite signal link operating frequency f (in GHz), atmospheric pressure p ℎ (in hPa), Kelvin temperature T and the water content density r0 calculated above.

[0017] Step 12: According to the atmospheric attenuation rate γ under clear sky conditions R , calculate the atmospheric attenuation value under clear sky conditions .

[0018] It can be understood that according to the atmospheric attenuation rate γ R , further calculate the total atmospheric attenuation value under clear sky conditions , and its calculation formula is: ; in, L is the distance of the satellite signal link in km.

[0019] Step 13, calculate the transmission power Pr of each satellite signal link, and calculate the total attenuation value under rainfall conditions based on the transmission power Pr .

[0020] In a possible implementation manner of the present invention, the transmit power Pr of each satellite signal link is calculated, and the total attenuation value under rainfall conditions is calculated based on the transmit power Pr. ,include: Step 131, under continuous sunny conditions before rainfall, receiving the satellite signal power of each satellite signal link at different times, averaging the satellite signal power of continuous time periods under sunny conditions to obtain a reference level value under clear sky conditions .

[0021] It can be understood that, taking the observation path of a satellite signal link as an example, under continuous sunny conditions before rainfall (such as one day), the satellite signal is received by the system antenna module and the signal is transmitted to the digital receiver module to obtain the level output data of the continuous time period under sunny conditions. All level data are averaged to obtain the average level value under sunny conditions. , thereby determining the reference level value.

[0022] Step 132, based on the atmospheric attenuation value under clear sky conditions and the reference level under clear sky conditions , calculate the transmission power Pr of each satellite signal link: ; in, L is the distance of the satellite signal link in km.

[0023] Step 133, under rainfall conditions, by receiving the satellite signal power of each satellite signal link, the level value under rainfall conditions is obtained .

[0024] It can be understood that under rainfall conditions, the satellite signal is received by the system antenna module and the signal is transmitted to the digital receiver module to obtain the satellite signal power level value under rainfall conditions. .

[0025] Step 134, based on the level value under rainfall conditions on each satellite signal link and transmit power Pr, calculate the total attenuation value of each satellite signal link under rainfall conditions : .

[0026] Step 135, the atmospheric attenuation value under clear sky conditions and the total attenuation value under rainfall conditions , calculate the rainfall attenuation value of each satellite signal link ,include: .

[0027] Step 14, according to the atmospheric attenuation value under clear sky conditions and the total attenuation value under rainfall conditions , calculate the rainfall attenuation value of each satellite signal link : .

[0028] Step 2: According to the rainfall attenuation value on each satellite signal link, the rainfall information of each satellite signal link is inverted.

[0029] In a possible implementation manner of the present invention, the inverting rainfall information of each satellite signal link according to the rainfall attenuation value of each satellite signal link includes: Step 21: Based on the relationship between rain attenuation and rainfall intensity in the ITU-R rain attenuation prediction model, the rainfall attenuation value associated with each satellite signal link is inverted. Corresponding linear average rainfall intensity information.

[0030] It is understandable that, assuming that the rainfall rate is evenly distributed over the entire satellite link, the rainfall intensity inversion formula is established based on the ITU-R rain attenuation model: ,in, is the rain attenuation value on the i-th link, i=1, 2, 3, ..., N. a i and b i is the ITU-R model parameter value corresponding to the i-th link, which is related to the link frequency, polarization mode and rainfall particle morphology and can be found in the ITU-R Recommendation. i is the attenuation coefficient, b i is the decay index; R i is the linear average rainfall intensity on the ith link, in mm / h; L i is the link length of the ith link, in km. Based on the length information, attenuation information of multiple links and the rain attenuation value obtained in step 1 The relationship between rain attenuation and rainfall intensity in the above ITU-R rain attenuation model can be used to invert the linear average rainfall intensity information R on multiple satellite links at the observation point.

[0031] Step 22, based on the linear average rainfall intensity information of each satellite signal link, the rainfall information of each satellite signal link is calculated.

[0032] It can be understood that, based on the linear average rainfall intensity information of each satellite path, the rainfall non-uniformity distribution information within a radius within a certain time window is calculated.

[0033] Step 3: Based on the rainfall information of each satellite signal link, a two-dimensional rainfall distribution field with high temporal and spatial resolution is established within the regional scope.

[0034] Step 4: Extract the wet delay information of each satellite signal path based on the GNSS solution software and construct a three-dimensional water vapor distribution field constraint model.

[0035] In a possible implementation manner of the present invention, the method of extracting wet delay information of each satellite signal path based on GNSS solution software and constructing a three-dimensional water vapor distribution field constraint model includes: Step 41 : extracting the zenith direction wet delay information ZWD on each satellite signal path based on the GNSS solution software, and mapping the zenith direction wet delay information ZWD on each satellite signal path to the wet delay amount SWD on each satellite signal slant path.

[0036] It is understandable that mature GNSS solution software (such as RTKLIB) is used to obtain the longitude and latitude coordinates, elevation angle, azimuth angle, zenith wet delay information ZWD and post-verification residual information of each GNSS measuring station.

[0037] The zenith wet delay on each satellite path is converted through a mapping function, and the zenith wet delay ZWD is mapped to the wet delay SWD on the slant path of each satellite signal. The mapping formula is: ; in, is the mapping function from ZWD to SWD. This model uses the Global Mapping Function (GMF). and Respectively represent the azimuth and elevation angles between a GNSS station and a certain satellite; To fit the residuals, the effects of the unmodeled errors are comprehensively represented; and They are the horizontal gradient terms of wet delay in the north-south and east-west directions respectively. These parameters can be obtained through the processing results of GNSS solution software.

[0038] Step 42, dividing the entire three-dimensional study area into grid voxel points, and dividing the entire three-dimensional study area into multiple vertical layers, each vertical layer containing multiple grid voxel points in the vertical direction.

[0039] Step 43, calculating the intercept distance between two adjacent layers of grid voxel points when each satellite signal oblique path passes through different vertical layers; Step 44, establishing a constraint equation, wherein the constraint equation includes a horizontal constraint, a vertical constraint and a priori constraints.

[0040] It is understandable that the entire three-dimensional study area is divided into grid voxel points, and the study area is divided into multiple vertical layers, each containing a number of grid voxel points. On this basis, for each ray path of the satellite signal, the intercept distance between the voxel points of two adjacent layers is calculated when it passes through different vertical layers. At the same time, the constraint equations are established: 1) horizontal constraints, using Gaussian weighting functions; 2) vertical constraints, the model is established based on the characteristic that water vapor decreases exponentially with vertical height; 3) prior constraints, combining numerical weather forecast information, standard atmospheric parameters and radiosonde data.

[0041] Step 45, constructing a three-dimensional water vapor distribution field constraint model according to the divided three-dimensional grid voxel points, the intercept distance of each grid voxel point and the constraint condition equation.

[0042] Step 5: Invert the three-dimensional water vapor distribution field constraint model to obtain three-dimensional water vapor distribution information, and convert the three-dimensional water vapor distribution information into a two-dimensional water vapor distribution field.

[0043] In a possible implementation manner of the present invention, inverting the three-dimensional water vapor distribution field constraint model to obtain three-dimensional water vapor distribution information includes: Step 51 : discretize the wet delay information SWD on each satellite signal slant path to obtain the wet delay information SWD of each grid voxel point.

[0044] It is understandable that the wet delay information SWD on each satellite ray link obtained in the above steps is discretized to obtain the wet delay amount information of each grid voxel point SWD.

[0045] Step 52: Establish a tomographic model matrix equation according to the wet delay information SWD of each grid voxel point and the intercept distance of each grid voxel point. ,in is the observation vector of SWD, is a large sparse matrix of ray path intercepts, is the wet refractive index distribution to be solved.

[0046] Step 53, the least square method and singular value decomposition method are used to solve the matrix equation of the tomographic model, and the ill-conditioned problem is optimized by combining the constraint condition equation to invert and obtain a more accurate three-dimensional water vapor distribution information, that is, the wet refractive index of the entire area obtained by the tomographic model above. Three-dimensional distribution.

[0047] The three-dimensional water vapor distribution information is converted into a two-dimensional water vapor distribution field, including: integrating the three-dimensional water vapor grid information obtained by water vapor tomography inversion in the zenith direction, and the formula is: , Nw is the wet refractive index information obtained by the solution. The zenith wet delay information ZWD of the grid points in the two-dimensional large-scale study area is obtained, and then converted into atmospheric precipitable water volume (PWV) water vapor information using an empirical formula, the formula is: , ;in, is the conversion factor; is the density of liquid water; is the water vapor gas constant; , is the atmospheric physical parameter. Thus, a two-dimensional water vapor distribution field with high temporal and spatial resolution is constructed.

[0048] Step 6: input the two-dimensional rainfall distribution field, the two-dimensional water vapor distribution field and meteorological data into a surface rainfall prediction model, and output a surface rainfall prediction result.

[0049] It is understandable that the two-dimensional rainfall data obtained in step 3, the two-dimensional water vapor distribution data obtained in step 5, and the meteorological data (including temperature, humidity, pressure, etc.) obtained by the system meteorological sensor module are imported as multi-source meteorological data into the trained improved Transformer surface rainfall prediction model. The surface rainfall prediction model consists of an encoder, a decoder, and an output layer.

[0050] Among them, 1) the encoder extracts the features of multi-source meteorological data; 2) the decoder uses the extracted features to predict the surface rainfall distribution in a certain time period in the future; 3) the output layer outputs the final surface rainfall prediction result. Through the above method, the accurate prediction of surface rainfall in a specific time period in the future is achieved, thereby achieving the effect of regional rainfall monitoring and early warning.

[0051] See also Figure 3 , is a fusion surface rainfall prediction system provided by the present invention, which is mainly composed of the following modules: Satellite antenna: used to receive satellite link signals in the Ku and Ka bands, providing a basis for subsequent signal analysis and processing.

[0052] GNSS antenna: responsible for receiving satellite signals in multiple frequency bands of the GNSS system and providing necessary data for water vapor tomography and wet delay calculation.

[0053] Digital receiver: Amplifies (LNA) the signal received by the satellite antenna, observes the changes in the satellite beacon signal, and records the power information of each satellite link in real time. These data provide full support for the subsequent calculation of rain attenuation values ​​and the inversion of two-dimensional rainfall fields.

[0054] GNSS receiver: Processes the multi-band signals received by the GNSS antenna to generate positioning satellite message data. These data provide basic support for the wet delay information required for water vapor tomography calculation.

[0055] Meteorological sensor: responsible for collecting meteorological parameters such as ambient temperature, humidity, and air pressure as input for atmospheric attenuation calculations, and also provides the required meteorological data for the surface rainfall prediction model.

[0056] Computer data processing module: connects digital receivers, GNSS receivers and meteorological sensors through serial ports or network interfaces, and performs the following core functions: 1) Calculation of rain attenuation and inversion of two-dimensional rainfall field, receives rain attenuation measurement data and environmental temperature and humidity information, obtains accurate rain attenuation values ​​after data processing, and then obtains linear average rainfall intensity information by combining with ITU-R model inversion, and finally generates a two-dimensional rainfall field with high temporal and spatial resolution; 2) Wet delay calculation and water vapor tomography, receives satellite message data generated by GNSS receiver, calculates wet delay information using GNSS solution software (such as RTKLIB), and obtains a two-dimensional water vapor distribution field with high temporal and spatial resolution based on the water vapor tomography model; 3) Precise prediction of surface rainfall, uses the inverted rainfall data, water vapor distribution data and other multi-source meteorological information as input, imports them into the surface rainfall prediction model, and accurately predicts the surface rainfall in a specific time period in the future. This structural design combines satellite signal inversion, water vapor tomography and multi-source meteorological data fusion technology, providing comprehensive support for high temporal and spatial resolution surface rainfall monitoring and early warning.

[0057] The present invention provides a fusion surface rainfall prediction method, which utilizes the innovative combination of rain attenuation inversion of rainfall and GNSS tomography to solve the water vapor field. It not only greatly improves the accuracy and coverage of rainfall monitoring and early warning, but also has the advantages of low cost and high efficiency. At the application level, the research results can provide more comprehensive and reliable data support for meteorological forecasting, water resources management, flood control and disaster prevention and other fields, and promote the development of surface rainfall monitoring and early warning technology. In the future, the present invention is expected to play a more important role in extreme weather prevention and control and natural disaster response, opening up new possibilities for scientific research and practical applications.

[0058] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0059] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0060] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0061] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0062] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0063] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0064] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A fusion surface rainfall prediction method, characterized in that: include: Obtain the environmental parameters around the observation site and the satellite signals of each satellite signal link, and calculate the rainfall attenuation value of each satellite signal link; According to the rainfall attenuation value of each satellite signal link, the rainfall information of each satellite signal link is inverted; Based on the rainfall information of each satellite signal link, a two-dimensional rainfall distribution field with high temporal and spatial resolution is established within the regional scope; Based on GNSS solution software, the wet delay information of each satellite signal path is extracted to construct a three-dimensional water vapor distribution field constraint model; Inverting the three-dimensional water vapor distribution field constraint model to obtain three-dimensional water vapor distribution information, and converting the three-dimensional water vapor distribution information into a two-dimensional water vapor distribution field; The two-dimensional rainfall distribution field, the two-dimensional water vapor distribution field and meteorological data are input into a surface rainfall prediction model, and a surface rainfall prediction result is output.

2. The fusion surface rainfall prediction method according to claim 1, characterized in that: The step of obtaining environmental parameters around the observation site and satellite signals of each satellite signal link, and calculating the rainfall attenuation value of each satellite signal link, includes: Based on the environmental parameters around the observation site, the atmospheric attenuation rate γ under clear sky conditions is calculated. R ; According to the atmospheric attenuation rate γ under clear sky conditions R , calculate the atmospheric attenuation value under clear sky conditions ; Calculate the transmission power Pr of each satellite signal link, and calculate the total attenuation value under rainfall conditions based on the transmission power Pr ; According to the atmospheric attenuation value under clear sky conditions and the total attenuation value under rainfall conditions , calculate the rainfall attenuation value of each satellite signal link .

3. The fusion surface rainfall prediction method according to claim 2 is characterized in that: The atmospheric attenuation rate γ under clear sky conditions is calculated based on the environmental parameters around the observation site. R ,include: Obtain the ambient temperature information and relative humidity information around the observation site; According to the ambient temperature information, the saturated water vapor pressure is calculated: ； in, E Indicates the saturated water vapor pressure at the current temperature, in hPa; t It is the current ambient temperature information, in ℃; According to the ambient temperature information and the ambient relative humidity information, the water content density is calculated: ； Among them, r0 is the water content density, the unit is g / m3; h is the relative humidity information of the environment, the unit is %; T is the Kelvin temperature, T=t+273.16, the unit is K; According to the operating frequency of each satellite signal link f , atmospheric pressure ph , Kelvin temperature T and water content density r0, call the function itu676a (f, ph, T, r0) in the ITU-R atmospheric attenuation model to obtain the atmospheric attenuation rate γ under clear sky conditions R ; The atmospheric attenuation rate γ under clear sky conditions R , calculate the atmospheric attenuation value under clear sky conditions ,include: ； in, L is the distance of the satellite signal link in km.

4. The fusion surface rainfall prediction method according to claim 2 is characterized in that: The transmit power Pr of each satellite signal link is calculated, and the total attenuation value under rainfall conditions is calculated based on the transmit power Pr. ,include: Under continuous sunny conditions before rainfall, the satellite signal power of each satellite signal link at different times is received, and the average value of the satellite signal power in continuous time periods under sunny conditions is calculated to obtain the reference level value under clear sky conditions. ; According to the atmospheric attenuation value under clear sky conditions and the reference level under clear sky conditions , calculate the transmission power Pr of each satellite signal link: ； in, L is the distance of the satellite signal link, in km; Under rainfall conditions, the level value under rainfall conditions is obtained by receiving the satellite signal power of each satellite signal link. ; According to the level value of rainfall conditions on each satellite signal link and transmit power Pr, calculate the total attenuation value of each satellite signal link under rainfall conditions : ; The atmospheric attenuation value according to clear sky conditions and the total attenuation value under rainfall conditions , calculate the rainfall attenuation value of each satellite signal link ,include: 。 5. The fusion surface rainfall prediction method according to claim 1, characterized in that: The step of inverting the rainfall information of each satellite signal link according to the rainfall attenuation value of each satellite signal link comprises: According to the relationship between rain attenuation and rainfall intensity in the ITU-R rain attenuation prediction model, the rainfall attenuation value of each satellite signal link is inverted. Corresponding linear average rainfall intensity information; Based on the linear average rainfall intensity information of each satellite signal link, the rainfall information of each satellite signal link is calculated.

6. The fusion surface rainfall prediction method according to claim 5, characterized in that: The relationship between rain attenuation and rainfall intensity in the ITU-R rain attenuation prediction model is: ； in, is the rainfall attenuation value on the i-th satellite signal link, i=1, 2, 3, ..., N; a i and b i is the ITU-R predicted rain attenuation model parameter value corresponding to the i-th satellite signal link, which is related to the frequency, polarization mode and rainfall particle morphology of the satellite signal link. i is the attenuation coefficient, b i is the decay index; R i is the linear average rainfall intensity information on the ith satellite signal link, in mm / h; L i is the link length of the ith satellite signal link, in km.

7. The fusion surface rainfall prediction method according to claim 1, characterized in that: The method of extracting wet delay information of each satellite signal path based on GNSS solution software and constructing a three-dimensional water vapor distribution field constraint model includes: Extract the zenith direction wet delay information ZWD on each satellite signal path based on GNSS solution software, and map the zenith direction wet delay information ZWD on each satellite signal path to the wet delay amount SWD on each satellite signal slant path; The entire three-dimensional study area is divided into grid voxel points, and the entire three-dimensional study area is divided into a plurality of vertical layers, each vertical layer comprising a plurality of grid voxel points in a vertical direction; Calculate the intercept distance between two adjacent grid voxel points when each satellite signal oblique path passes through different vertical layers; Establishing a constraint condition equation, wherein the constraint condition equation includes a horizontal constraint, a vertical constraint and a priori constraints; A three-dimensional water vapor distribution field constraint model is constructed based on the divided three-dimensional grid voxel points, the intercept distance of each grid voxel point and the constraint condition equation.

8. The fusion surface rainfall prediction method according to claim 7, characterized in that: The mapping of the zenith direction wet delay information ZWD on each satellite signal path into the wet delay amount SWD on each satellite signal slant path includes: ； in, is the mapping function from ZWD to SWD; parameter and Respectively represent the azimuth and elevation angles between the GNSS station and the specified satellite; is the fitting residual; and are the horizontal gradient terms of wet delay in the north-south and east-west directions, respectively.

9. The fusion surface rainfall prediction method according to claim 7, characterized in that: The three-dimensional water vapor distribution field constraint model is inverted to obtain three-dimensional water vapor distribution information, including: Discretize the wet delay information SWD on each satellite signal slant path to obtain the wet delay information SWD of each grid voxel point; According to the wet delay information SWD of each grid voxel point and the intercept distance of each grid voxel point, the tomographic model matrix equation is established. ,in is the observation vector of SWD, is a large sparse matrix of ray path intercepts, Then is the wet refractive index distribution to be solved; The least squares method and singular value decomposition method are used to solve the matrix equation of the tomographic model, and the constraint equation is combined to optimize the ill-conditioned problem, and the wet refractive index of the entire area is inverted. The three-dimensional distribution of water vapor is the three-dimensional distribution information.

10. The fusion surface rainfall prediction method according to claim 9, characterized in that: The converting the three-dimensional water vapor distribution information into a two-dimensional water vapor distribution field comprises: The three-dimensional wet refractive index Integrate in the vertical direction to obtain the zenith wet delay information ZWD of the grid voxel points in the two-dimensional study area; The zenith wet delay information ZWD of the grid voxel points in the two-dimensional study area is converted into atmospheric precipitable water vapor information PWV to construct a two-dimensional water vapor distribution field with high temporal and spatial resolution, where: ； ； in, is the conversion factor, is the density of liquid water, , , is the specific gas constant related to water vapor.

Images