Typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy
By adopting a multi-parameter comprehensive weighting strategy in typhoon monitoring, combining PWV and other meteorological parameters, a linear weighting model is established, which solves the problems of insufficient accuracy and poor stability in the existing methods, and achieves more accurate and stable typhoon monitoring results. In the eastern coastal areas of China.
Patent Information
- Application Number
- CN202411377042.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-30
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2044-09-30
AI Technical Summary
The existing typhoon monitoring methods ignore the influence of multiple parameters and parameter differences, resulting in insufficient accuracy and poor stability. Especially in typhoon monitoring in eastern China's eastern coastal areas, it is difficult to accurately describe typhoon movement.
Using a typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy, the discrete coefficients of each parameter are calculated by establishing a typhoon database and meteorological parameter data set, and a multi-parameter linear weighting model is established, combining PWV, air pressure, rainfall, wind speed, temperature and other parameters. , perform linear weighting and normalization processing to improve the accuracy and stability of monitoring results.
Effectively improve the accuracy and stability of typhoon monitoring, overcome the problems of high measurement costs, low spatiotemporal resolution and environmental impact in traditional methods, provide more accurate typhoon path forecast, and reduce the deviation of single PWV to typhoon path forecast .
Smart Images

Figure CN118897932A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of meteorological monitoring, and in particular to a typhoon monitoring method based on PWV and a multi-parameter comprehensive weighting strategy. Background Art
[0002] Precipitable Water Vapor (PWV) refers to the amount of precipitation resulting from the condensation of all water vapor in a column of air per unit cross-sectional area between the ground and the top of the atmosphere. It reflects the changing water vapor content in the atmosphere. During extreme weather events such as typhoons, the spatiotemporal state of atmospheric water vapor fluctuates dramatically. In recent years, with global warming, typhoons have become more frequent, posing a serious threat to human life and property. Therefore, timely and accurate understanding of the spatiotemporal distribution of atmospheric water vapor is crucial for weather and climate research.
[0003] Traditional PWV acquisition technologies primarily include radiosondes (RS), microwave radiometers, and lidar. RS can observe atmospheric parameters such as temperature and pressure up to 30 km, and the uncertainty of the retrieved PWV is 0–1.5 mm. However, these instruments suffer from high cost and low temporal and spatial resolution. Microwave radiometers measure atmospheric radiation energy, providing real-time atmospheric conditions and vertical distribution information. However, they are expensive, require regular calibration, and are easily affected by precipitation. Lidar offers high vertical resolution and observation accuracy, but the measurement cost is high and measurements are easily affected by precipitation and dense clouds, making it difficult to obtain long time series datasets suitable for meteorological analysis. In recent years, the potential of using PWV to monitor typhoon movement has been demonstrated. However, current typhoon motion monitoring algorithms primarily focus on PWV, ignoring the influence of other meteorological parameters such as temperature and wind speed. This makes it difficult to provide a detailed description of typhoon movement, resulting in insufficient accuracy.
[0004] Compared with traditional PWV monitoring methods, PWV derived from ERA5 (ECMWF Reanalysis v5) reanalysis data has the advantage of high temporal and spatial resolution (1 hour, 0.25˚ × 0.25˚). The ERA5 reanalysis dataset is a high-quality global meteorological dataset provided by the European Centre for Medium-Range Weather Forecasts (ECMWF), covering a wide range of regions and a long time span (from 1940 to the present). Therefore, this study investigated PWV and other meteorological parameters derived from ERA5 reanalysis data and proposed an improved typhoon track monitoring model. This model not only addresses the high cost and low temporal and spatial resolution of traditional PWV acquisition methods, but also effectively reduces the bias of single PWV in typhoon track forecasts. Summary of the Invention
[0005] In response to the deficiencies in the prior art, the present invention proposes a typhoon monitoring method based on PWV and a multi-parameter comprehensive weighted strategy to improve the accuracy and stability of typhoon monitoring. This method adopts a multi-parameter linear weighted model, incorporates multiple meteorological parameters into the analysis system, and targets the performance differences of different parameters in typhoon monitoring. By calculating the distribution center of PWV, air pressure, rainfall, wind speed, temperature and other parameters during each typhoon between 2018 and 2022 and the discrete coefficient of the typhoon position, the weight is calculated by taking the mean, thereby effectively improving the accuracy and stability of the monitoring results. The present invention can monitor typhoons in the eastern coastal areas of China, solving the problems of insufficient accuracy and poor stability caused by the existing methods ignoring the influence of multiple parameters and parameter differences.
[0006] In order to solve the above technical problems, the technical solution of the present invention is:
[0007] A typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy, the specific steps are as follows:
[0008] Step 1: Establish a typhoon database. Using the Zhejiang Provincial Water Resources Department's real-time typhoon track release website, collect data on typhoons that made landfall along China's eastern coast from 2018 to 2022, including information such as typhoon occurrence time, center location, and typhoon severity. This data is categorized by year and typhoon name and saved in Excel format on a local computer.
[0009] Step 2: Create a meteorological parameter dataset. First, download hourly meteorological data for the eastern coast of China (105°E–135°E, 10°N–40°N) from the ERA5 (ECMWF Reanalysis v5) reanalysis dataset for 2018–2022. This data includes air pressure, rainfall, wind speed, temperature, and specific humidity. Specific humidity is pressure-stratified data; wind speed data is presented as a U component (longitude) and a V component (latitude), which must be calculated using vectors and formulas; temperature data is in Kelvin (K) and must be converted to Celsius (°C). All data are categorized by year and variable name and saved as nc files on a local computer. Next, download meteorological parameter data for several RS stations along the eastern coast of China (105°E–135°E, 10°N–40°N) from the University of Wyoming RS dataset for 2018–2022, including air pressure and specific humidity. This data is categorized by year and variable name and saved as an Excel spreadsheet on a local computer.
[0010] Step 3: Extract and verify ERA5-PWV. First, calculate PWV using the RS meteorological parameter data for several stations downloaded in Step 2 and the corresponding ERA5 meteorological parameter data at different times and spaces. Then, use RS-PWV as a benchmark to verify the accuracy and usability of ERA5-PWV.
[0011] Step 4: Calculate the distance difference between the parameter peak and the typhoon center. First, use the ERA5 meteorological parameter data in the meteorological database to calculate the spatial distribution of PWV during each typhoon. Then, extract the latitude and longitude corresponding to the hourly peak values of PWV, air pressure, rainfall, wind speed, and temperature. Use the Haversine formula to calculate the distance between the center of each parameter peak and the typhoon center. Finally, calculate the mean and variance of each parameter distance difference series, and use the dispersion coefficient formula to calculate the dispersion coefficient of each parameter during each typhoon.
[0012] Step 5: Establish a multi-parameter weighted model. Calculate the mean of the coefficient of variation for the distance difference between the PWV, air pressure, rainfall, wind speed, and temperature peak center and the typhoon center, processed in Step 3. Subtract 1 and take the absolute value as the weight. Using PWV as the key parameter and a linear weighting formula, establish two-, three-, and four-parameter PWV weighted models. Furthermore, normalize the weights of each parameter to ensure that their sum is 1.
[0013] Step 6: Predict the typhoon's location. Using the weighted model established in Step 4, input data such as PWV, air pressure, rainfall, wind speed, and temperature within a specific time and space range. Linearly transform the output to the interval [0, 1], and extract the longitude and latitude corresponding to the maximum value as the predicted typhoon location.
[0014] Through the above algorithm, the present invention effectively solves the problems of low accuracy and poor stability caused by ignoring the influence of multiple parameters and the performance differences of different parameters in existing methods, and provides a new idea for typhoon monitoring.
[0015] The present invention has the following characteristics and beneficial effects:
[0016] Compared with existing typhoon monitoring methods based on radiosondes and lidar, the adoption of the above-mentioned technical solution not only solves problems such as high measurement costs and susceptibility to environmental influences, but also overcomes the shortcomings of low spatial and temporal resolution and inability to meet large-scale measurement needs. Compared with the existing single PWV typhoon monitoring method, this method not only improves monitoring accuracy through multi-parameter weighted constraints, but also solves the problem of monitoring instability caused by ignoring the influence of other meteorological parameters. In addition, considering the differences in the performance of different parameters in typhoon monitoring, the weighting coefficient is calculated using long time series data, further ensuring the accuracy and stability of the typhoon monitoring model. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0018] Figure 1 A flow chart for establishing a typhoon and meteorological parameter database in an embodiment of the present invention.
[0019] Figure 2 Flowchart of a typhoon monitoring method based on the PWV integral formula, discrete coefficient and weighting strategy in an embodiment of the present invention.
[0020] Figure 3 Schematic diagram comparing the results of the two-parameter combination model based on PWV and any other parameter.
[0021] Figure 4 Schematic diagram comparing the results of the three-parameter combination model based on PWV and any other two parameters.
[0022] Figure 5 Schematic diagram comparing the results of the four-parameter combination model based on PWV and any other three parameters. DETAILED DESCRIPTION
[0023] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.
[0024] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings.
[0025] On the contrary, the present invention covers any alternatives, modifications, equivalents, and solutions that fall within the spirit and scope of the present invention as defined by the claims. Furthermore, to facilitate a better understanding of the present invention, certain specific details are described in detail below in the detailed description of the present invention. Those skilled in the art will be able to fully understand the present invention without these details.
[0026] Example 1
[0027] The present invention provides a typhoon monitoring method based on PWV and a multi-parameter comprehensive weighting strategy, comprising the following steps:
[0028] Step 1: Establish a typhoon database and a meteorological parameter data set within the same region and the same time region respectively. The typhoon database includes the typhoon occurrence time, center location and typhoon level, and the meteorological parameter data set includes air pressure, rainfall, wind speed, temperature and specific humidity.
[0029] Specifically, such as Figure 1 As shown, in this embodiment, data on typhoons that made landfall along China's eastern coast from 2018 to 2022 is obtained from the Zhejiang Provincial Water Resources Department's real-time typhoon path release website. This data includes information such as typhoon occurrence time, center location, and typhoon level. This data is categorized by year and typhoon name and stored in an Excel spreadsheet on a local computer.
[0030] Furthermore, according to the established typhoon database, meteorological data such as air pressure, rainfall, temperature, wind speed, and specific humidity for the eastern coast of China (105˚E–135˚E, 10˚N–40˚N) from 2018 to 2022 were downloaded from the ERA5 reanalysis dataset and stored in the local computer in the form of nc files according to year and variable name.
[0031] In a further configuration of this embodiment, a meteorological parameter dataset is established, which extracts the meteorological parameter dataset of the eastern coast of China (105˚E~135˚E, 10˚N~40˚N) during each typhoon from 2018 to 2022, mainly including air pressure, rainfall, wind speed and temperature, and converts them into standard units respectively. The original wind speed observation values are the U-component (longitude) and the V-component (latitude) at 10m. These two components need to be combined to convert to the 10m wind speed. The conversion formula is as follows:
[0032]
[0033] Where, Indicates wind speed Towards the weight, Indicates wind speed Towards the weight, Indicates the synthesized wind speed in units of .
[0034] The original observed temperature is in Kelvin , in order to standardize the units, it is converted, the formula is as follows:
[0035]
[0036] Where, Expressed in Kelvin Temperature data in units of Indicates conversion to degrees Celsius Temperature data.
[0037] Step 2: Select any The PWV was calculated based on the meteorological parameter data of RS observations at each station and the ERA5 meteorological parameter data of the corresponding time and space. The RS-PWV calculated based on the meteorological parameter data of RS observations was used as a benchmark to verify the ERA5-PWV calculated based on the ERA5 meteorological parameter data.
[0038] In this embodiment, RS meteorological parameter data for stations 45004 and 47102 from 2018 to 2022, including specific humidity and air pressure, are downloaded from the University of Wyoming website and stored in an Excel spreadsheet on a local computer according to year and variable name.
[0039] It should be noted that PWV is defined as the total amount of water vapor contained in an air column per unit area. PWV is the most abundant greenhouse gas on Earth, accounting for 70% of the total atmospheric radiation absorption. It changes greatly in extreme weather such as typhoons and can be used as an important parameter for monitoring typhoon activities.
[0040] Specifically, the PWV calculation process is divided into two steps. First, using the meteorological database established in step 2, the RS observation data of the two stations 45004 and 47102 within the study time range and the corresponding ERA5 meteorological parameter data at different times and spaces, including air pressure and specific humidity, are extracted, and the PWV is calculated using the formula. The formula is as follows:
[0041]
[0042] Where, Represents the acceleration due to gravity in units of ; Indicates the surface air pressure in units of ; is the specific humidity of the specified pressure layer, in units of ; represents the pressure difference; Indicates the number of layers; and yes Layer and Surface pressure of the layer; and Respectively represent Layer and Specific humidity of the layer.
[0043] Furthermore, this embodiment also includes testing the accuracy and usability of ERA5-PWV.
[0044] Specifically, using RS-PWV as a benchmark, the usability of ERA5-PWV was tested using two criteria: the root mean square error (RMSE) and the mean absolute error (MAE). RMSE measures the average difference between the statistically predicted value (ERA5-PWV) and the true value (RS-PWV). Mathematically, it is the standard deviation of the residuals, representing the distance between the regression line and the data points, and is calculated as follows:
[0045]
[0046] Where, Indicates the total number of data; and represent the predicted value and the true value respectively.
[0047] To eliminate the influence of gross errors, MAE was also used to evaluate the performance of ERA5-PWV, which was defined as the mean absolute difference between the predicted value (ERA5-PWV) and the true value of the data (RS-PWV).
[0048]
[0049] Where m represents the total number of data; and Represent the predicted value and the true value respectively; Represents absolute value operation.
[0050] Furthermore, since RS data has a 12-hour temporal resolution, ERA5-PWV was compared using the same time interval. In meteorological research, an acceptable range for RMSE (root mean square error) is generally considered to be between 0.5 mm and 2 mm, and an acceptable range for MAE (mandatory error) is generally considered to be between 0.3 mm and 1.5 mm. When the RMSE and MAE errors are within acceptable ranges, ERA5-PWV and RS-PWV are well-matched and can be used for further analysis and calculations.
[0051] If the test results do not meet the requirements, recalculate the RMSE and MAE year by year in different time periods to find out the abnormal years, and use the deviation correction formula to correct the error. The formula is as follows:
[0052]
[0053] Where, represents the mean deviation, Indicates the abnormal ERA5-PWV values, represents the value of RS-PWV at the corresponding moment, Indicates the total number of abnormal data, Recalculate the RMSE and MAE of the corrected ERA5-PWV. If the RMSE and MAE after bias correction are within the acceptable range, continue with the subsequent analysis. Otherwise, repeat the correction.
[0054] Step 3: Calculate the spatial distribution of PWV during each typhoon period based on ERA5 meteorological parameter data. Then, extract the longitude and latitude corresponding to the peak center of the typhoon parameter. Use the haversine formula to calculate the distance between the peak center of the typhoon parameter and the typhoon center. Finally, calculate the mean and variance of the typhoon parameter distance difference sequence. Use the dispersion coefficient formula to calculate the dispersion coefficient of the typhoon parameter during each typhoon period.
[0055] The typhoon parameters include hourly PWV, air pressure, rainfall, wind speed, and temperature.
[0056] Specifically, in this example, after verifying the accuracy and availability of ERA5-PWV within the study time range, the spatial distribution of PWV along the eastern coast of China (105˚E~135˚E, 10˚N~40˚N) during each typhoon from 2018 to 2022 was calculated according to the PWV calculation formula.
[0057] Further, such as Figure 2 As shown, based on meteorological parameter data and ERA5-PWV, the longitude and latitude corresponding to the peak centers of hourly PWV, air pressure, rainfall, wind speed, and temperature during each typhoon period were extracted. Combined with the typhoon center position at the corresponding epoch, the distance difference between the distribution center of each parameter and the typhoon center was calculated using the Haversine Formula. The Haversine Formula is a formula used to calculate the shortest distance (great circle distance) between two points on a sphere. It is commonly used in geographic information systems (GIS) and navigation to calculate the distance between two points on the earth. Its principle is based on spherical trigonometry and the calculation formula is as follows:
[0058]
[0059] Where, represents the sine function; represents the cosine function; Indicates the difference between the latitudes of two places; Indicates the difference between the longitudes of two places; and is the latitude of two places; is the equatorial radius of the Earth, which is 6371 kilometers in this embodiment.
[0060] The coefficient of dispersion is then calculated for the distance difference series between the peak distribution center of each parameter and the typhoon center. The coefficient of dispersion is the ratio of the standard deviation to the mean. It is a dimensionless unit and can directly compare data sets of different units or orders of magnitude. At the same time, when comparing the degree of dispersion of different data sets, the variance can only reflect the absolute degree of dispersion, while the coefficient of dispersion can reflect the relative degree of dispersion, that is, the degree of dispersion of the data around the mean. In this way, even if the mean of the data set is different, an effective comparison can be made. The formula is as follows:
[0061]
[0062] Where, represents the standard deviation; represents the mean; The number of data representing the time series; It represents the coefficient of dispersion. It is understandable that the smaller the value, the smaller the data fluctuation, that is, the better the performance of meteorological parameters in typhoon monitoring.
[0063] Step 4: Establish a multi-parameter linear weighted model based on the dispersion coefficients of typhoon parameters.
[0064] Specifically, we first extract the weighting coefficient based on the calculated dispersion coefficient. Considering that the performance of the same parameter in different typhoon monitoring varies, we take the mean of the dispersion coefficient of each parameter as the basis for calculating the weight. The smaller the dispersion coefficient, the better the performance of the parameter in typhoon monitoring, and the larger the corresponding weight. Therefore, the weight of each parameter is determined according to the following formula:
[0065]
[0066] Where, represents the mean of the corresponding parameter dispersion coefficient; Represents the weight. Including weight coefficients corresponding to PWV, air pressure, rainfall, wind speed, and temperature.
[0067] Then, a linear weighted model was established with PWV as the key parameter, and the weighting coefficients were normalized to ensure that the sum of the weights of the parameters in the weighted model was 1. The model is as follows:
[0068]
[0069] Where, represents the results of the combined model; Refers to the weight coefficient of the corresponding parameter. In this embodiment, PWV: 0.52, wind speed: 0.59, air pressure: 0.48, temperature: 0.47, rainfall: 0.38; Represents variable: air pressure ( ), rainfall ( )、wind speed( ) and temperature ( ).
[0070] Furthermore, for the convenience of comparison, the model output result range is transformed to [0,1] using the linear transformation formula. The linear transformation formula is as follows:
[0071]
[0072] Where, Represents the original output of the model; Indicates the minimum value of the original output result; Indicates the maximum value of the original output result; The result after linear transformation is shown in Figure 2. Finally, the peak position of the combined model result is extracted every hour, which is the predicted typhoon position.
[0073] In this example, the average (median) difference between the distance between the weighted model's predicted typhoon center and the actual typhoon center is used as an indicator to reflect the model's reliability. The median is robust and unaffected by extreme values, making it a more reliable indicator of the typical level of forecast error. The coefficient of dispersion, a second indicator reflecting the overall volatility and stability of forecast results, accounts for the relative volatility of forecast errors and better describes the consistency and stability of the model's forecast results. A smaller coefficient of dispersion indicates less volatility in the model's forecast results and greater stability. The model inputs are the eight typhoons that made landfall along China's eastern coast between 2018 and 2022 (Maria, Ampil, Mangkhut, Bailu, Lekima, Hagupit, In-Fa, and Chaba).
[0074] The specific performance is as follows: Figure 3 As can be seen, the distribution of forecast deviations is represented by a box plot. The coefficient of dispersion is represented by an orange star line. It can be seen that the average level of range deviation for the single PWV model is approximately 200 km. Both the range deviation and the coefficient of dispersion change after adding other meteorological parameters. For example, the combination of PWV and W outperforms the single PWV model, as evidenced by the significantly reduced width of the box plot. Furthermore, the coefficient of dispersion for the combined PWV and W model decreases from 0.53 to 0.47, indicating an overall reduction in forecast deviation for the combined PWV and W model. In general, with the exception of the combined PWV and P2 model, the coefficients of dispersion for the other two-parameter models are superior to those of the single PWV model.
[0075] pass Figure 4It can be seen that, except for the combination model that includes P2, the other combination models outperform the single PWV model. For example, the coefficient of variation of PWV-P1-W is reduced to 0.47. This means that combining temperature and pressure can improve typhoon monitoring performance compared to using PWV alone.
[0076] pass Figure 5 As can be seen, the coefficient of dispersion of the PWV-P1-WT combined model decreases from 0.53 to 0.46, an improvement of approximately 13.21% compared to the single PWV model. Here, P1 represents air pressure, P2 represents rainfall, W represents wind speed, and T represents temperature.
[0077] Step 5: Using the established multi-parameter linear weighted model, input PWV, air pressure, rainfall, wind speed, and temperature within a certain time and space range; linearly transform the output result into the [0,1] interval, and extract the longitude and latitude corresponding to the maximum value of the result as the predicted typhoon position.
[0078] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. It will be apparent to those skilled in the art that various changes, modifications, substitutions, and variations of these embodiments, including components, without departing from the principles and spirit of the present invention are still within the scope of protection of the present invention.
Claims
1. A typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy, characterized in that: The steps include: Step 1, respectively establishing a typhoon database and a meteorological parameter data set in the same region and the same time region, wherein the typhoon database includes the typhoon occurrence time, center location and typhoon level, and the meteorological parameter data set includes air pressure, rainfall, wind speed, temperature and specific humidity; Step 2: Select the meteorological parameter data observed by RS at any n stations and the ERA5 meteorological parameter data of the corresponding time and space, calculate PWV respectively, and use the RS-PWV calculated by the meteorological parameter data observed by RS as the benchmark to verify the ERA5-PWV calculated by the ERA5 meteorological parameter data; Step 3: Calculate the spatial distribution of ERA5-PWV during each typhoon according to the ERA5 meteorological parameter data, then extract the longitude and latitude corresponding to the typhoon parameters, and use the haversine formula to calculate the distance between the peak center of the typhoon parameters and the typhoon center. Finally, calculate the mean and variance of the typhoon parameter distance difference sequence, and use the dispersion coefficient formula to calculate the dispersion coefficient of the typhoon parameters during each typhoon. The typhoon parameters include hourly ERA5-PWV, air pressure, rainfall, wind speed, and temperature; Step 4: Establish a multi-parameter linear weighted model based on the discrete coefficients of typhoon parameters; Step 5: Using the established multi-parameter linear weighted model, input ERA5-PWV, air pressure, rainfall, wind speed, and temperature within a certain time and space range; linearly transform the output result into the [0,1] interval, and extract the longitude and latitude corresponding to the maximum value of the result as the predicted typhoon position.
2. A typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy according to claim 1, characterized in that: The data in the typhoon database are classified according to year and typhoon name, and saved in the local computer in the form of Excel tables.
3. A typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy according to claim 1, characterized in that: The wind speed in the meteorological parameter data set exists in the form of a longitude component and a latitude component, and is calculated by a vector and a formula.
4. A typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy according to claim 1, characterized in that: The temperature data in the meteorological parameter data set is in Kelvin and needs to be converted to Celsius.
5. The typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy according to claim 1 is characterized in that: The ERA5 meteorological parameter data in the meteorological parameter data set is saved in the form of an nc file, and the RS observed meteorological parameter data is saved in the form of an Excel table.
6. The typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy according to claim 1 is characterized in that: The PWV is calculated as follows: ; In the formula, Represents the acceleration due to gravity in units of ; It represents the surface pressure in ; is the specific humidity of the specified pressure layer, in units of ; represents the pressure difference; Indicates the number of layers; and yes Layer and Surface pressure of the layer; and Respectively represent Layer and Specific humidity of the layer.
7. A typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy according to claim 6, characterized in that: The step 2 also includes a test of PWV, taking RS-PWV as a benchmark, and testing the obtained ERA5-PWV by using two criteria: root mean square error and mean absolute error.
8. A typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy according to claim 7, characterized in that: If the ERA5-PWV test results do not meet the requirements, recalculate the RMSE and MAE year by year in different time periods to find out the abnormal years, and use the deviation correction formula to correct the error. The formula is as follows: ; ; In the formula, represents the mean deviation, Indicates abnormal ERA5-PWV values, represents the value of RS-PWV at the corresponding moment, N represents the total number of abnormal data, Represents the corrected data. Recalculate RMSE and MAE for the corrected ERA5-PWV. If the RMSE and MAE after deviation correction are within the acceptable range, the subsequent analysis can be continued, otherwise the correction is repeated.
9. A typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy according to claim 7, characterized in that: In step 3, the distance between the peak center of the typhoon parameter and the typhoon center is calculated using the haversine formula as follows: ; In the formula, represents the sine function; represents the cosine function; It represents the difference between the latitudes of two places; Indicates the difference between the longitudes of two places; and is the latitude of two places; is the equatorial radius of the Earth.
10. A typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy according to claim 9, characterized in that: In step 3, the discrete coefficient formula is used to calculate the discrete coefficient of the typhoon parameters during each typhoon as follows: ; In the formula, represents standard deviation; represents the mean; The number of data representing the time series; represents the coefficient of dispersion.
11. A typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy according to claim 10, characterized in that: The multi-parameter linear weighted model is as follows: ; In the formula, represents the results of the combined model; refers to the weight coefficient of the corresponding parameter; Represents a variable.
12. A typhoon monitoring method based on PWV and multi-parameter comprehensive weighting strategy according to claim 11, characterized in that: In step 5, the method of applying the multi-parameter linear weighted model to predict the typhoon path is: First, the weighting coefficient is extracted according to the discrete coefficient, and the mean of the discrete coefficient of each parameter is taken as the basis for calculating the weight: ; Then, the calculated weight coefficients and various parameter variables are input into the model, and the results are output; Finally, the linear transformation formula is used to transform the model output result range to [0,1]. The linear transformation formula is as follows: ; In the formula, Represents the original output of the model; Indicates the minimum value of the original output result; Indicates the maximum value of the original output result; It represents the result after linear transformation, and finally the peak position of the hourly combined model result is extracted, which is the predicted typhoon position.
Citation Information
Patent Citations
Near-earth weighted average temperature information acquisition method based on multiple meteorological factors
CN113639893A
High-temporal-spatial-resolution near-surface air temperature reconstruction method and system and equipment
CN114019579A
Water quality multi-parameter spectral data Stacking fusion model and water quality multi-parameter measurement method
CN114894725A
PWV reconstruction method and system based on polynomial correction and neural network
CN115993668A
Typhoon disaster forecasting method and system based on ensemble learning
CN117828490A
Cited By
Multi-source rainfall forecast fusion method based on typhoon similarity
CN119355848A
