A three-dimensional water vapor distribution fusion method based on GNSS-PWV

Abstract

The present invention relates to a GNSS-PWV-based three-dimensional water vapor distribution fusion method, comprising the following steps: S1, obtaining the GNSS-PWV distribution of atmospheric precipitable water over a ground-based GNSS-MET weather station; S2, obtaining nearby radiosonde data, ground-based microwave radiometer water vapor profiles or numerical meteorological models as three-dimensional water vapor initial field profiles; S3, by developing a three-dimensional water vapor fusion technology based on GNSS-PWV, obtaining the three-dimensional distribution of water vapor observations at each observation time of each ground-based GNSS station; the GNSS-PWV-based three-dimensional water vapor distribution fusion method effectively utilizes the characteristics of high accuracy and high observation frequency of the entire layer of GNSS-PWV water vapor, and can achieve complementary advantages with sounding data, ground-based microwave radiometers and other profile layers, which is beneficial to front-line forecasters and numerical meteorological models to monitor and analyze weather systems in real time, and provides observation support for improving the accuracy of daily severe convection and typhoon forecasts and warnings.

Description

Technical Field

[0001] The invention belongs to the technical field of meteorological observation data fusion, and relates to a three-dimensional water vapor distribution fusion method based on GNSS-PWV. Background Art

[0002] Water vapor observation at ground-based GNSS-MET meteorological stations (Global Navigation Satellite System, GNSS) uses ground receiving equipment to measure the characteristics of the bending and delay of the transmission path caused by refraction of navigation satellite signals when passing through the atmosphere, and inverts the atmospheric water vapor information. Traditional water vapor detection methods have specific disadvantages such as insufficient spatial or temporal resolution, limited accuracy or small scope of application. Askne and Nordius (1987) first proposed the relationship between the zenith wet delay of navigation satellite signals and the atmospheric water vapor value, and Bevis et al. (1992) demonstrated the technical feasibility of GPS (Global Position System) inverting water vapor content. In theory, GNSS-MET inversion of atmospheric precipitable water vapor (PWV) has the advantages of low cost, high accuracy, and no need for calibration, which is a powerful supplement to traditional atmospheric water vapor observation methods. At present, my country's ground-based GNSS-MET observations have been included in the basic business of the China Meteorological Administration, and GNSS-MET stations have basically achieved full coverage of cities and counties across the country, and the subsequent system is being accelerated. However, since the precipitable water content of the entire atmosphere observed by GNSS-MET cannot directly reflect the three-dimensional water vapor distribution in the atmosphere, the ability to analyze changes in atmospheric structure and stratification stability is insufficient, especially for water vapor field data, whose concentration changes dramatically with time and spatial distribution. Monitoring of three-dimensional water vapor distribution and changes is a very important physical quantity for predicting rainfall, small and medium-scale severe weather, and global climate change.

[0003] GNSS water vapor detection technology is one of the important means to reconstruct the three-dimensional distribution of atmospheric water vapor with high temporal and spatial resolution. It is generally used in the form of tomographic grid division to perform water vapor three-dimensional tomography (Flores et al., 2000; Seko et al., 2000; Troller et al., 2006). GNSS water vapor three-dimensional tomography discretizes the tropospheric water vapor above the local area of ​​the GNSS-MET meteorological network, and evenly divides many three-dimensional grids in the three-dimensional space. Each layer of grids has a relative volume. The total water vapor content on the slant path of the GNSS signal (Slant-Path Water Vapor, SWV) is calculated through the model function; then, the integral equation is listed based on the slant path of the GNSS signal. When there is enough observation information, the equation group can be solved to obtain the water vapor density of each three-dimensional grid. However, the tomography technology is currently limited by many factors, including the geometric configuration of the tomography area, the addition of constraints, the density of the GNSS-MET station network, and the high-precision SWV inversion technology, which significantly affect the reliability and stability of the water vapor three-dimensional tomography results. NOAA proposed a new form of mesoscale analysis and developed a three-dimensional data fusion system called LAPS (Local Analysis and Prediction System), which can analyze and fuse a whole set of observation data (meteorological ground observation network, radar, satellite, vertical detector, aircraft, etc.) to obtain three-dimensional, high-resolution three-dimensional data. ARPS (The Advanced Regional Prediction System) and ADAS (ARPS Data Analysis System) developed on this basis have powerful capabilities for multivariate data fusion, but the calculation is complex and depends on model background field data. Kuo et al. (1993) used a three-dimensional variational method to assimilate GPS-PWV to improve the initial field of the numerical weather model. Du Mingbin et al. (2022) used GNSS-PWV to improve the water vapor profile product of my country's Fengyun meteorological satellite FY-4A / GIIRS, thereby improving the accuracy of inverting atmospheric water vapor profiles using satellite data.

[0004] In view of the importance of three-dimensional water vapor distribution information in daily weather monitoring and forecasting, we currently have a variety of observation methods such as ground observation, sounding, GNSS-MET, radar, satellite, and ground-based microwave radiometers. These observation methods can detect the distribution of water vapor in the atmosphere from different angles, but the various data obtained from them have obvious limitations. How to develop three-dimensional water vapor fusion technology to complement the advantages of various water vapor observation data and effectively improve the quantitative application level of remote sensing water vapor inversion information is the key to the current multi-element water vapor application, while also taking into account the data source stability and algorithm execution efficiency of technical applications. Summary of the invention

[0005] In view of this, in order to solve the problems existing in the above-mentioned prior art in how to obtain accurate three-dimensional water vapor distribution information, the present invention provides a three-dimensional water vapor distribution fusion method based on GNSS-PWV, which fully utilizes the characteristics of GNSS-PWV's high accuracy of whole-layer water vapor, all-weather, and high temporal and spatial resolution, and introduces an initial guess field of three-dimensional water vapor profile to achieve complementary advantages of the two types of water vapor information and then fuse them to generate a three-dimensional water vapor distribution product, and the three-dimensional water vapor fusion product based on this technical method is shared in real time on a meteorological information platform.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] A 3D water vapor distribution fusion method based on GNSS-PWV effectively utilizes the characteristics of high accuracy and high observation frequency of the whole layer water vapor of GNSS-PWV, which specifically includes the following steps:

[0008] S1. Obtain the GNSS-PWV distribution of atmospheric precipitable water at each observation time of the ground-based GNSS-MET meteorological station;

[0009] S2. Obtain the initial three-dimensional water vapor field profile, and directly obtain the initial three-dimensional atmospheric water vapor field, mainly including the initial water vapor profile field, by using radiosonde equipment, ground-based microwave radiometers or numerical model forecast products;

[0010] S3. Develop 3D water vapor fusion technology based on GNSS-PWV to obtain the 3D distribution of water vapor observations at each ground-based GNSS station at each observation time.

[0011] The water vapor distribution of each pressure layer above GNSS-MET station i is obtained by formula (1):

[0012]

[0013] Where, PWV i PWV is the precipitable water volume of the whole atmosphere observed by ground-based GNSS meteorological observations GNSS-PWV. c represents the atmospheric precipitable water calculated from the initial field profile of the station, q c (k) is the initial field profile; where,

[0014]

[0015] In formula (2), ρ is the water density, g is the gravitational acceleration, p0 is the ground pressure at the station, and k represents the kth layer of the atmosphere.

[0016] Furthermore, in step S1, the GNSS-PWV distribution of the entire atmospheric precipitable water is obtained by relying on ground-based GNSS meteorological observation equipment. The atmospheric precipitable water GNSS-PWV distribution observation equipment is a ground-based GNSS-MET meteorological station, which can obtain GNSS-PWV once every hour or half an hour. In the future, the site density and GNSS-PWV time frequency can reach once every 5 minutes.

[0017] Furthermore, step S3 takes into account that directly adjusting the relative humidity will cause the atmospheric analysis field to be oversaturated. Since the GNSS-MET station contains ground temperature, air pressure and humidity, the relationship between temperature and height is used as the constraint of relative humidity. The atmospheric saturated relative humidity profile is calculated according to the temperature profile as q si (k); if q i (k)>q si (k), then q i (k) = q si (k); According to the Magnus saturated water vapor pressure empirical formula, the saturated water vapor pressure E si (k), and then calculate q si (k) is used as the saturation constraint value. For the specific method, refer to formula (3) and formula (4).

[0018]

[0019] Where t(k) is the atmospheric temperature of the kth layer (in degrees Celsius), P(k) is the atmospheric pressure of the kth layer; the atmospheric precipitable water of the updated water vapor profile is calculated as PWV i (j) (j is the jth iteration), set |PWV i (j)-PWV i (j+1)|<ε; where ε is the convergence index, and the ε value is set according to the specific conditions of each ground-based GNSS-MET meteorological station.

[0020] Furthermore, in order to more accurately obtain the three-dimensional profile structure of water vapor above the ground-based GNSS-MET weather station, step S3 introduces the Cressman objective analysis method and performs grid analysis on the sounding stations that are not far apart;

[0021]

[0022] Where, the subscript i is the number of m radiosonde stations within the radius R, and d i is the distance between the ith sounding station and the GNSS-MET station, W i is the distance weight factor, and the specific calculation formula is:

[0023]

[0024] The water vapor profile with radiosonde data uses the radiosonde data as the initial field, while the water vapor profile with no radiosonde data uses the ground-based microwave radiometer data or numerical prediction field as the initial field.

[0025] Furthermore, in step S3, the value of the profile convergence index ε is set to 0.1 mm.

[0026] The beneficial effects of the present invention are:

[0027] 1. The three-dimensional water vapor distribution fusion method based on GNSS-PWV disclosed in the present invention utilizes the characteristics of high accuracy, all-weather, and high temporal and spatial resolution of the whole layer of water vapor of GNSS-PWV, and introduces the initial guess field of the three-dimensional water vapor profile to achieve the complementary advantages of the two water vapor information and then fuse to generate a three-dimensional water vapor distribution product. The three-dimensional water vapor fusion product based on this technical method is shared in real time on the meteorological information platform. After verification, the relative humidity error of the three-dimensional water vapor profile after fusion is less than 20% at the altitude layer below 800hPa, and the error of high-level water vapor is relatively large. Through the application of important weather process cases, it can be seen that GNSS observation of water vapor distribution and its three-dimensional fusion products have important reference value in the weather diagnosis and analysis of typhoons and severe convection processes, and can effectively reflect the precipitation change trend of offshore typhoons and the severe convection triggering area of ​​hail weather. GNSS water vapor detection is one of the important means to reconstruct the three-dimensional distribution of atmospheric water vapor with high temporal and spatial resolution. The research and development of multivariate data fusion based on GNSS water vapor products has broad application prospects in comprehensive meteorological observation applications.

[0028] 2. The three-dimensional water vapor distribution fusion method based on GNSS-PWV disclosed in the present invention is based on GNSS-PWV business observation data. The technical application data source is highly stable. Through the research and development of three-dimensional water vapor fusion technology, three-dimensional water vapor products can be generated according to the time resolution of GNSS-PWV products. The time resolution of GNSS-PWV is usually higher than once every half an hour, and it can reach once every 5 minutes in the future. The spatial resolution is determined by the construction of the station network. Due to the low operating cost and high stability of GNSS-MET meteorological stations, there are currently more than 1,000 GNSS-MET meteorological stations in operation across the country. The three-dimensional water vapor distribution fusion product based on GNSS-PWV is conducive to front-line forecasters and numerical meteorological models to monitor and analyze weather systems in real time, and provides observation support for improving the accuracy of daily severe convection and typhoon forecasts and warnings.

[0029] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below in conjunction with the accompanying drawings, wherein:

[0031] Figure 1 This is the relative humidity error map of the three-dimensional water vapor profile based on GNSS-PWV fusion of the present invention;

[0032] Figure 2 The spatial distribution of the sounding profile of the three-dimensional water vapor fusion product based on GNSS-PWV of the present invention, where the blue station numbers are conventional radiosonde sites, and the red station numbers are the distribution of the sounding sites developed by the present invention;

[0033] Figure 3 This is a business operation diagram of the three-dimensional water vapor fusion product based on GNSS-PWV on the pilot platform of the present invention;

[0034] Figure 4 This is a three-dimensional water vapor analysis map of the Shanghai area, which includes Figure 4 (a) is the horizontal distribution diagram of 925hPa; Figure 4 (b) is the horizontal distribution diagram at 850hPa; Figure 4 (c) is the vertical section in the north-south direction at longitude 121.4E; Figure 4 (d) is the vertical section in the east-west direction at 31.2N north latitude;

[0035] Figure 5 It is a multi-layer horizontal distribution diagram of the multi-element water vapor three-dimensional fusion product of the present invention;

[0036] Figure 6 This is the distribution map of water vapor impact of Typhoon Lekima 201909 from August 9 to August 10, including Figure 6 (a) is the distribution map of water vapor impact on August 9; Figure 6 (b) is the distribution map of water vapor impact on August 10;

[0037] Figure 7 This is the east-west profile of humidity at 31°N from August 9 to August 10, 2019, where Figure 7 (a) is the east-west profile of humidity at 31°N on August 9; Figure 7 (b) is the east-west profile of humidity at 31°N on August 10;

[0038] Figure 8 This is the north-south profile of humidity at 121.6°E from August 9 to August 10, 2019, where Figure 8 (a) is the north-south profile of humidity at 121.6°E on August 9; Figure 8 (b) is the north-south profile of humidity at 121.6°E on August 10;

[0039] Fig. 9 is the three-dimensional water vapor vertical profile and radar reflectivity on June 10, 2023, where Fig. 9 (a) is the vertical section in the north-south direction at 124.2°E; Fig. 9 (b) is the radar reflectivity factor diagram of 0.5 degrees. DETAILED DESCRIPTION

[0040] The following describes the embodiments of the present invention through specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention.

[0041] At present, the observation frequency of conventional radiosonde twice a day and the station distribution cannot fully meet the requirements of water vapor temporal and spatial variability observation. Ground-based detection instruments have high temporal resolution and vertical resolution, but are limited by the geographical distribution range or detection altitude and cannot provide stable and long-term atmospheric parameter data. GNSS-MET (GNSS meteorology) observation of atmospheric precipitable water (GNSS-PWV) has the advantages of low cost, high accuracy, and no need for calibration, and is a powerful supplement to traditional atmospheric water vapor observation methods.

[0042] Example

[0043] 1. Three-dimensional water vapor dynamic monitoring technology based on GNSS-PWV

[0044] 1.1 Data and Methods

[0045] A 3D water vapor distribution fusion method based on GNSS-PWV effectively utilizes the characteristics of high accuracy and high observation frequency of the whole layer water vapor of GNSS-PWV, which specifically includes the following steps:

[0046] S1. Obtain the GNSS-PWV distribution of atmospheric precipitable water above the ground-based GNSS-MET meteorological station at each observation time; the existing technology relies on ground-based GNSS meteorological observation equipment to obtain the GNSS-PWV distribution of atmospheric precipitable water in the entire layer. The atmospheric precipitable water GNSS-PWV distribution observation equipment is a ground-based GNSS-MET meteorological station, which can obtain GNSS-PWV once every hour or half an hour. In the future, the station density and GNSS-PWV time frequency can reach once every 5 minutes.

[0047] S2. Obtain the initial three-dimensional water vapor field profile, and directly obtain the initial three-dimensional atmospheric water vapor field, mainly including the initial water vapor profile field, by using radiosonde equipment, ground-based microwave radiometers or numerical model forecast products;

[0048] S3. Develop 3D water vapor fusion technology based on GNSS-PWV to obtain the 3D distribution of water vapor observations at each ground-based GNSS station at each observation time.

[0049] The water vapor distribution of each pressure layer above GNSS-MET station i is obtained by formula (1):

[0050]

[0051] Where, PWV i PWV is the precipitable water volume of the whole atmosphere observed by ground-based GNSS meteorological observations GNSS-PWV. c represents the atmospheric precipitable water calculated from the initial field profile of the station, q c (k) is the initial field profile; where,

[0052]

[0053] In formula (2), ρ is the water density, g is the gravitational acceleration, p0 is the ground pressure at the station, and k represents the kth layer of the atmosphere.

[0054] Considering that direct adjustment of specific humidity will cause oversaturation of the atmospheric analysis field, since the GNSS-MET station contains ground temperature, air pressure and humidity, the relationship between temperature and height is used as the constraint of relative humidity, and the atmospheric saturated specific humidity profile is calculated according to the temperature profile as q si (k); if q i (k)>q si (k), then q i (k) = q si (k); According to the Magnus saturated water vapor pressure empirical formula, the saturated water vapor pressure E si (k), and then calculate q si (k) is used as the saturation constraint value. For the specific method, refer to formula (3) and formula (4).

[0055]

[0056] Where t(k) is the atmospheric temperature of the kth layer (in degrees Celsius), P(k) is the atmospheric pressure of the kth layer; the atmospheric precipitable water of the updated water vapor profile is calculated as PWV i (j) (j is the jth iteration), set |PWV i (j)-PWV i (j+1)|<ε; where ε is the convergence index, and the ε value is set according to the specific conditions of each ground-based GNSS-MET meteorological station.

[0057] In order to obtain the three-dimensional water vapor profile structure above the ground-based GNSS-MET meteorological station more accurately, the Cressman objective analysis method is introduced to perform grid analysis on the nearby sounding stations.

[0058]

[0059] Where, the subscript i is the number of m radiosonde stations within the radius R, and d i is the distance between the ith sounding station and the GNSS-MET station, W i is the distance weight factor, and the specific calculation formula is:

[0060]

[0061] The water vapor profile with radiosonde data is based on the radiosonde data as the initial field, and the water vapor profile without radiosonde data is based on the ground-based microwave radiometer data or numerical prediction field as the initial field. The profile convergence index ε is set to 0.1 mm.

[0062] Three-dimensional water vapor fusion based on GNSS-PWV data is carried out. Currently, three-dimensional fused humidity products of atmospheric water vapor can be obtained every hour. In the future, with the increase in the frequency of GNSS-PWV observations, the temporal resolution of the obtained three-dimensional fused humidity products of water vapor can be further improved.

[0063] 1.2. Testing the accuracy of the three-dimensional fusion humidity product obtained in the embodiment

[0064] Conventional radiosonde is currently the world's unified standard method for detecting high-altitude atmospheric elements. It has a stable data source and is widely distributed. Under ideal conditions and after careful correction, radiosonde is very accurate. The general accuracy of radiosonde is ±0.5K for air temperature, ±1hPa for air pressure, and ±5% for relative humidity. The ascent height of the sounding instrument can generally reach more than 30 kilometers, but in the troposphere and stratosphere, the accuracy of air pressure and temperature decreases. Conventional radiosonde data can be used to test the GNSS-PWV three-dimensional fusion effect. Figure 1 As shown in the figure, the water vapor profiles for the past three months from June to August 2020 were checked. The relative humidity errors at altitudes below 800hPa were all less than 20%, and the water vapor errors at high altitudes were relatively large. Through analysis, it was found that the relative error was large because the water vapor content at high altitudes was low and the relative humidity of water vapor was low. Figure 2 The figure shows the spatial distribution of the sounding profile of the three-dimensional water vapor fusion product based on GNSS-PWV. The blue station numbers are conventional radiosonde sites, and the red station numbers are the distribution of the sounding sites developed by the present invention.

[0065] 1.3. Introduction to the 3D Humidity Product Integration Platform

[0066] By utilizing the ability of the sounding data to detect atmospheric humidity profiles and the high precision and high spatiotemporal resolution of GNSS-PWV, effective fusion of water vapor information can be achieved, and a three-dimensional water vapor field can be obtained every hour. In theory, each observation at each GNSS-MET station can generate a water vapor profile through data fusion methods, and can provide three-dimensional water vapor products for diagnostic analysis of precipitation forecasts. Figure 3 As shown, based on the GNSS-PWV product, using multivariate data fusion technology, since March 2019, a set of three-dimensional element humidity observation and analysis fields in the Shanghai area has been generated every hour, and the real-time business of the fusion product has been uploaded to the urban meteorological service pilot platform. At present, the product system runs stably and can be shared globally through the intranet system.

[0067] The diagnostic analysis products of precipitation forecasts provided by this platform can further improve the level of short-term precipitation forecasts, which is of great significance for small and medium-scale severe weather forecasts, and is also an inherent requirement for improving the quantitative level of multi-element water vapor inversion information. The three-dimensional element humidity observation and analysis field in Shanghai includes three products: different levels, longitudinal profiles, and latitudinal profiles. The three-dimensional information comparison of the products is shown in Table 1. Figure 4 As shown in the figure, the three-dimensional water vapor fusion product includes the water vapor horizontal distribution map at the 1025-500hPa height level, the north-south vertical profile map along the longitude of 120.8-122.2E, and the east-west vertical profile map along the latitude of 30.6-32N. Figure 4 (a) is the horizontal distribution diagram of 925hPa; Figure 4 (b) is the horizontal distribution diagram at 850hPa; Figure 4 (c) is the vertical section in the north-south direction at longitude 121.4E; Figure 4 (d) is the vertical section in the east-west direction at 31.2N north latitude; Figure 5 As shown, according to the water vapor distribution at different altitudes, a multi-layer horizontal distribution map of a multi-element water vapor three-dimensional fusion product can be produced.

[0068] Table 1. 3D water vapor product information comparison table

[0069]

[0070] 2. Application of water vapor 3D fusion products based on GNSS-PWV

[0071] 2.1 Analysis of typhoon impact process

[0072] Typhoon Lekima was classified as a typhoon by the Central Meteorological Observatory on August 7, and further strengthened into a super typhoon at 23:00 on August 7. It continued to move northwestward, approaching the coast of Zhejiang Province, and landed at the coast of Chengnan Town, Wenling City, Zhejiang Province at about 1:45 on August 10, causing great impact on East China. As Typhoon Lekima moved westward and northward, the water vapor in Shanghai area increased significantly from August 9. The hourly GNSS-PWV monitoring showed that the typhoon system affected the water vapor distribution significantly, and the water vapor in the Yangtze River Delta region increased significantly. Figure 6 As shown, Figure 6 (a) is the distribution map of water vapor impact on August 9; Figure 6 (b) is the water vapor impact distribution map on August 10.

[0073] From the humidity profile at 31°N in Shanghai ( Figure 7 ) Figure 7 (a) is the east-west profile of humidity at 31°N on August 9; Figure 7 (b) is the east-west profile of humidity at 31°N on August 10; before the impact of "Lekima" on August 9, the overall relative humidity in the east-west direction was below 80%. As the typhoon moved westward and northward, the relative humidity in the entire layer of Shanghai area increased rapidly in the afternoon of August 9, especially in the eastern areas that were directly affected by the typhoon, with relative humidity reaching 95%, and the height of the high humidity area also reached above 700hPa, and later reached above 500hPa.

[0074] From the humidity profile at 121.6°E in Shanghai ( Figure 8 ) Figure 8 (a) is the north-south profile of humidity at 121.6°E on August 9; Figure 8 (b) is the north-south profile of humidity at 121.6°E on August 10; before the impact of Typhoon Lekima on August 9, the overall relative humidity in the north-south direction was below 80%. As the typhoon moved westward and northward, the relative humidity in the entire layer of Shanghai area increased significantly in the afternoon of August 9, especially in the south, where the relative humidity reached 95%. As the typhoon moved northward, the high humidity in the north-south profile changed from north-south to north-south, reflecting the process distribution of the typhoon's water vapor impact.

[0075] 2.2 Analysis of severe convective weather processes

[0076] Affected by strong thunderstorm clouds, the city will experience showers or thunderstorms from the afternoon of June 10, accompanied by strong convective weather such as strong winds, hail and short-term heavy rainfall. From 16:00 on June 10 to 06:00 on June 11: 4.7% of the towns and villages exceeded 25.0 mm, and 27.0% of the towns and villages exceeded 10.0 mm, among which 37.5 mm was the largest in Xincheng Road, Jiading, and the maximum rainfall in 1 hour occurred in Xincheng Road, Jiading, 37.0 mm (18:20-19:20 on the 10th). In addition, Chongming, Baoshan, Jiading, Songjiang, Minhang, Fengxian, Pudong and the central urban area experienced thunderstorms and strong winds of 7 to 9 levels, among which Sheshan Town, Songjiang had the largest level of 9 (21.2 m / s). Hail appeared in Chongming, the urban area and other places.

[0077] like Fig. 9 As shown, the vertical profile of water vapor along 124.2°E at 19:00 on June 10 and the 0.5 degree reflectivity factor at 19:04 are shown. Fig. 9 (a) is the vertical section in the north-south direction at 124.2°E; Fig. 9 (b) is the 0.5 degree radar reflectivity factor map; at this time, the convective echo developed the strongest, and hail appeared in many places. The vertical profile of water vapor shows that the boundary layer (below 950hPa) is nearly saturated, the water vapor decreases rapidly upward, and it is dry above 600hPa. There is a clear contrast between dry and wet in the middle and high altitudes and low altitudes. Combined with the sounding at 20:00 on the same day (figure omitted), the humidity in the lower layer is relatively high, and there is an obvious dry area in the middle layer. This type of atmospheric stratification is prone to produce weather such as thunderstorms, gales, and hail. The radar map shows that the reflectivity factor exceeds 65dBz, and the profile shows that the 50dBz echo exceeds 10km (figure omitted). It has a typical echo overhang structure. Considering that the -20℃ wet bulb height of the sounding at 08:00 on the same day was about 6.7km, the probability of hail is relatively high.

[0078] In summary, accurate water vapor observation data is the basic guarantee for the effective forecast of small- and medium-scale severe weather. Traditional water vapor detection methods, such as radiosondes, ground meteorological stations, and microwave radiometers, have a series of drawbacks such as insufficient spatial or temporal resolution, limited accuracy, or a small scope of application. Based on ground-based GNSS-MET observations, combined with radiosonde and ground-based microwave radiometer sounding profiles, the differences between multivariate water vapor observation data are studied, and data fusion is performed to obtain a high-temporal and high-precision three-dimensional water vapor fusion data set, which provides data support for short-term weather forecasts, artificial precipitation, climate change, and other research. The specific results are as follows:

[0079] 1) Taking advantage of the high precision and high temporal resolution of ground-based GNSS-MET observations of atmospheric precipitable water, a three-dimensional water vapor product based on GNSS-PWV fusion was developed by developing multi-data fusion technology, and has been globally shared on the pilot platform since 2019.

[0080] 2) After verification, the relative humidity error of the fused three-dimensional water vapor profile at altitudes below 800hPa is less than 20%, while the error of high-level water vapor is relatively large. Through analysis, it is found that the relative humidity error is large because the water vapor content at high levels is low and the relative humidity of water vapor is low.

[0081] 3) Through the application of important weather process cases, it can be seen that GNSS observation of water vapor distribution and its three-dimensional fusion products have important reference value in the weather diagnosis and analysis of typhoons and severe convective processes, and can effectively reflect the precipitation change trend of offshore typhoons and the severe convection triggering area of ​​hail weather.

[0082] The ground-based GNSS-MET service has been included in the assessment of the observation service of the China Meteorological Administration. In the future, the construction of stations will be more and more complete, and the spatial resolution of the stations will be greatly improved. The ground-based GNSS-MET observation elements are single, and the conventional method can only observe the precipitation of the entire atmosphere, which is insufficient for the analysis of changes in atmospheric structure and stratification stability. This patent effectively utilizes the characteristics of the GNSS-PWV product, and through the development of technology that integrates with sounding data, each observation of each GNSS-MET station can be integrated to generate a water vapor profile, providing diagnostic analysis products for precipitation forecasting, and realizing dynamic three-dimensional monitoring of humidity based on GNSS-PWV. At present, the technical process has been realized, and the real-time business has been uploaded to the urban meteorological service pilot platform. At present, the product system is running stably.

[0083] At present, the three-dimensional water vapor fusion products based on GNSS-PWV must rely on the initial profile, and the water vapor profile structure accuracy is required to be high. Fusion relies on sounding data (including radiosonde and ground-based microwave radiometer). Since the distribution of sounding stations and ground-based microwave radiometers is small and the observation frequency is low, the horizontal resolution of the fusion product is seriously affected. It is possible to consider introducing high-resolution satellite remote sensing products. Ground-based GNSS water vapor observations play an important role in the observation and forecast of severe weather. The development of comprehensive application methods based on ground-based GNSS combined with other detection information has broad prospects.

[0084] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.

Claims

1. A three-dimensional water vapor distribution fusion method based on GNSS-PWV, characterized in that: The following steps are involved: S1. Obtain the GNSS-PWV distribution of atmospheric precipitable water at each observation time of the ground-based GNSS-MET meteorological station; S2. Obtain the initial three-dimensional water vapor field profile, and directly obtain the initial three-dimensional atmospheric water vapor field, mainly including the initial water vapor profile field, by using radiosonde equipment, ground-based microwave radiometers or numerical model forecast products; S3. Develop 3D water vapor fusion technology based on GNSS-PWV to obtain the 3D distribution of water vapor observations at each observation time of each ground-based GNSS station; The water vapor distribution of each pressure layer above GNSS-MET station i is obtained by formula (1): Where, PWV i PWV is the precipitable water volume of the whole atmosphere observed by ground-based GNSS meteorological observations GNSS-PWV. c represents the atmospheric precipitable water calculated from the initial field profile of the station, q c (k) is the initial field profile; where, In formula (2), ρ is the water density, g is the gravitational acceleration, p0 is the ground pressure at the station, and k represents the kth layer of the atmosphere; Step S3: In order to more accurately obtain the three-dimensional profile structure of water vapor above the ground-based GNSS-MET weather station, the Cressman objective analysis method is introduced to perform grid analysis on the sounding stations that are not far apart; Where, the subscript i is the number of m radiosonde stations within the radius R, and d i is the distance between the ith sounding station and the GNSS-MET station, W i is the distance weight factor, and the specific calculation formula is: The water vapor profile with radiosonde data uses the radiosonde data as the initial field, while the water vapor profile with no radiosonde data uses the ground-based microwave radiometer data or numerical prediction field as the initial field.

2. The three-dimensional water vapor distribution fusion method based on GNSS-PWV as claimed in claim 1, characterized in that: In step S1, the GNSS-PWV distribution of the entire atmospheric precipitable water is obtained by relying on the ground-based GNSS meteorological observation equipment. The atmospheric precipitable water GNSS-PWV distribution observation equipment is a ground-based GNSS-MET meteorological station, which can obtain GNSS-PWV once every hour or half an hour.

3. The 3D water vapor distribution fusion method based on GNSS-PWV as claimed in claim 2, characterized in that: Step S3 takes into account that directly adjusting the relative humidity will cause the atmospheric analysis field to be oversaturated. Since the GNSS-MET station contains ground temperature, air pressure and humidity, the relationship between temperature and height is used as the constraint of relative humidity. The atmospheric saturated relative humidity profile is calculated according to the temperature profile as q si (k); if q i (k)>q si (k), then q i (k) = q si (k); According to the Magnus saturated water vapor pressure empirical formula, the saturated water vapor pressure E si (k), and then calculate q si (k) is used as the saturation constraint value. For the specific method, refer to formula (3) and formula (4): Where t(k) is the k-th atmospheric temperature in degrees Celsius, and P(k) is the k-th atmospheric pressure. The atmospheric precipitable water volume after the fusion of the updated water vapor profile is calculated as PWV i (j)j is the jth iteration, set |PWV i (j)-PWV i (j+1)|<ε; where ε is the convergence index, and the ε value is set according to the specific conditions of each ground-based GNSS-MET meteorological station.

4. The three-dimensional water vapor distribution fusion method based on GNSS-PWV as claimed in claim 3, characterized in that: In step S3, the value of the profile convergence index ε is set to 0.1 mm.

Images