A method and system for storm surge analysis of long-term sea level data

Abstract

The application discloses a storm surge analysis method and system for long-time-series sea level data, and comprises the following steps: S1, acquiring long-time-series sea level measured data; S2, calculating the sea level component of each year according to the long-time-series sea level measured data; S3, performing tidal harmonic analysis on the long-time-series sea level measured data from which the sea level component is removed to obtain storm surge time-series data; and S4, performing characteristic analysis on the storm surge time-series data to obtain a storm surge analysis result. Compared with the prior art, the application separates and acquires storm surges from long-time-series sea level measured data of a coastal tide station, and further analyzes the storm surge distribution during typhoon and non-typhoon periods, the relationship between the storm surge and typhoon characteristics, the top ten historical storm surge events, and the interdecadal storm surge change trend.

Description

Technical Field

[0001] This invention relates to the field of meteorological and hydrological early warning technology, and more specifically, to a method and system for analyzing storm surge data from long-term sea level data. Background Technology

[0002] Storm surge, also known as storm water rise, is a common marine disaster in coastal areas. It is a sea-level rise induced by wind and pressure gradients on the ocean surface during typhoons, and is the most direct manifestation of the typhoon's influence and forcing on the ocean. Previously, storm surge was often extracted by subtracting the astronomical tide level from the measured tide level, neglecting the impact of sea-level changes. Against the backdrop of human activities and global climate change, sea-level rise has become increasingly significant, and its impact on storm surge has received widespread attention from the scientific community. Therefore, when extracting storm water rise from long-term measured sea-level data, the influence of sea-level changes must be removed first.

[0003] Existing technologies propose a method, system, device, and storage medium for early warning of storm surge flooding and overtopping. The method includes the following steps: acquiring real-time water level data and real-time wave height data; constructing a three-dimensional storm surge mathematical model; acquiring typhoon wind field information, upstream flow boundary information, and offshore tidal boundary information; obtaining storm surge water level prediction data based on the three-dimensional storm surge mathematical model; acquiring wave run-up calculation formulas and overtopping calculation formulas; calculating target wave run-up data and target overtopping data; and issuing an overtopping warning based on the comparison between the target wave run-up data and target overtopping data and the overtopping crest elevation. However, this approach does not eliminate the sea level component and tidal component, making it impossible to accurately obtain the storm surge volume.

[0004] Therefore, in light of the above requirements and the shortcomings of existing technologies, this application proposes a method and system for analyzing storm surge data in long-term sea level data. Summary of the Invention

[0005] This invention provides a method and system for analyzing storm surges from long-term sea level data. It separates and obtains storm surges from long-term sea level data measured at coastal tide gauge stations, and further analyzes the distribution of storm surges during typhoon and non-typhoon periods, the relationship between storm surges and typhoon characteristics, the top ten historical storm surge events, and the interdecadal trend of storm surge changes.

[0006] The primary objective of this invention is to solve the aforementioned technical problems. The technical solution of this invention is as follows:

[0007] The first aspect of this invention provides a method for analyzing storm surge data from long-term sea level data, the method comprising the following steps:

[0008] S1. Obtain long-term sea level measurement data.

[0009] S2. Calculate the annual sea level component based on long-term sea level measurement data.

[0010] S3. Tidal harmonic analysis is performed on the long-term sea level measurement data after removing the sea level component to obtain storm surge time series data.

[0011] S4. Perform feature analysis on storm surge time series data to obtain storm surge analysis results.

[0012] Further, step S1 specifically involves: obtaining measured sea level data from a public database and preprocessing the obtained data; the preprocessing process is as follows: removing data with a quality control identifier of 0 from the measured sea level data, retaining only data with a quality control identifier of 1, detecting abrupt changes in the mean of the data and correcting the average difference of the abrupt change points to obtain long-term measured sea level data.

[0013] It should be noted that a quality control flag of 0 indicates that the data is unavailable, while a quality control flag of 1 indicates that the data is available.

[0014] Furthermore, the method for detecting abrupt changes in the mean of the data is the Pruned Exact Linear Time (PELT) method. This method defines the criterion function as the sum of the costs of all segmented segments, with the aim of minimizing the value of the criterion function. The mathematical expression of the Pruned Exact Linear Time method is as follows:

[0015]

[0016] Where Γ is the index of a set of variable points, i.e., Γ = {t1, t2, ...}, t is the variable point index, y is the input data, V is the criterion function, pen is the measure of the piecewise complexity of variable points, and C is the measure of the sub-signal. The loss function for model fit is the mean-shift model C. L2 As the loss function, pen uses a linear penalty as a constraint on the number of variable points:

[0017]

[0018] pen L2 (Γ)=β|Γ|

[0019] in, Let Γ be the empirical time-series mean between sub-signals a and b, β be a smoothing parameter that measures complexity and good fit, where β > 0, and |Γ| be the cardinality of the change points.

[0020] Further, the specific process of step S2 is as follows: Divide the long-term sea level measurement data by year, calculate the annual median sea level for each year, and subtract the annual median sea level from the long-term sea level measurement data to obtain the annual sea level component. Its mathematical expression is:

[0021]

[0022] Where X' represents the data after removing the sea level component, i represents the descending sequence number of the sea level data for each year, j represents the year, and N represents the number of time-series sea level data points for each tide gauge station each year.

[0023] Further, the specific process of step S3 is as follows: tidal harmonic analysis is calculated using MATLAB and tidal components are removed; the number of tidal fractions is automatically selected and the tidal harmonic constant is calculated using the S_Tide model; and the predicted tide level is obtained by subtracting the tidal harmonic analysis from the long-term sea level measured data after removing the sea level components.

[0024] It should be noted that the predicted tide level is based on storm surge time series data.

[0025] Furthermore, the mathematical expression of step S3 is as follows:

[0026] H storm_surge =H′-H tide

[0027]

[0028] Among them, H storm_surge H represents the storm surge component, and H' represents the long-term sea level measurement data after removing the sea level component. tide Let Z be the tidal component, Z be the number of tidal constituents in the tidal harmonic analysis, and H be the tidal component. Z Let σ be the tidal amplitude. Z For the tidal angular velocity, g Z The tidal lag angle is T, where T is time.

[0029] Furthermore, before step S4, it is necessary to obtain the storm surge periods affecting the target station and to divide the storm surge data into typhoon and non-typhoon periods. The specific calculation process is as follows:

[0030]

[0031]

[0032] Where S is the spherical distance between the target station and the typhoon center, r is the Earth's radius, taken as 6371.137 km; φ1 and φ2 are the latitudes of the two points, and λ1 and λ2 are the longitudes of the two points, all in radians.

[0033] Furthermore, in step S4, the specific content of feature analysis includes:

[0034] S41. Calculate the empirical cumulative distribution function and Gaussian probability function of storm surge data during typhoon and non-typhoon periods, respectively.

[0035] S42. Calculate the relationship between storm surge and typhoon characteristics during typhoon season.

[0036] S43. Calculate the maximum storm surge increase and its maximum water level in chronological order, and then visualize the result.

[0037] Further, in step S41, the empirical cumulative distribution function and the Gaussian probability function are specifically as follows:

[0038]

[0039]

[0040] Where K is the Gaussian kernel function, x i Here, n represents the number of time-series data points, and h represents the bandwidth.

[0041] The relationship between storm surge and typhoon characteristics during typhoon season includes: storm surge during typhoon season and the distance between the typhoon and the target monitoring station at the corresponding time, the minimum air pressure at the typhoon center, the radius of maximum wind speed, and the maximum wind speed information. The radius of maximum wind speed is calculated using the following Willoughby formula:

[0042]

[0043] Among them, R max V is the radius of maximum wind speed. max θ represents the maximum wind speed near the typhoon center, and θ represents the latitude of the typhoon center.

[0044] Furthermore, the calculation of the maximum storm surge and its maximum water level by order of time includes: calculating the historical maximum storm surge events and their corresponding water levels, the monthly maximum storm surge and maximum water level, and the annual maximum storm surge and maximum water level; the visualization methods include: drawing distribution maps, drawing trend maps, and performing linear regression fitting on the trends.

[0045] The second aspect of the present invention provides a storm surge analysis system for long-term sea level data. The system is used in the storm surge analysis method for long-term sea level data and includes: a data acquisition module, a data processing module, and a feature analysis module.

[0046] The data acquisition module acquires long-term sea level measurement data and inputs it into the data processing module. The data processing module calculates the sea level component for each year based on the long-term sea level measurement data. The long-term sea level measurement data after removing the sea level component is subjected to tidal harmonic analysis to obtain storm surge time series data, which is then input into the feature analysis module. The feature analysis module performs feature analysis on the storm surge time series data and outputs storm surge analysis results.

[0047] Compared with the prior art, the beneficial effects of the technical solution of the present invention are:

[0048] This invention provides a storm surge analysis method and system for long-term sea level data. By removing the sea level component and performing tidal harmonic analysis on the collected data, the influence of sea level rise on the measured sea level data in the long-term sea level data can be eliminated. Compared with the traditional method of subtracting the astronomical tide level from the measured tide level, this method is more scientific and reasonable and has good universality. Attached Figure Description

[0049] Figure 1 This is a flowchart of a storm surge analysis method for long-term sea level data according to the present invention.

[0050] Figure 2 This is a flowchart of step S4 of the storm surge analysis method for long-term sea level data according to the present invention.

[0051] Figure 3 This is a time series diagram showing the original measured sea level and the sea level component removed from the station in one embodiment of the present invention.

[0052] Figure 4 This is an example of an embodiment of the present invention, showing the empirical cumulative distribution function (ECDF) of storm surge during typhoon and non-typhoon periods at a weather station.

[0053] Figure 5 This is a graph of the probability density function (PDF) of storm surge during typhoon and non-typhoon periods at a monitoring station in one embodiment of the present invention. PDF can describe the distribution of values ​​of a random variable.

[0054] Figure 6 This is a scatter plot showing the relationship between storm surge and typhoon characteristics at a monitoring station during a typhoon season, according to one embodiment of the present invention.

[0055] Figure 7 This is a bar chart showing the top ten largest storm surges and corresponding water levels at a monitoring station in one embodiment of the present invention.

[0056] Figure 8This is a distribution map of the maximum monthly storm surge and maximum water level at a monitoring station in one embodiment of the present invention.

[0057] Figure 9 This is a diagram showing the annual maximum storm surge and maximum water level change at a monitoring station in one embodiment of the present invention.

[0058] Figure 10 This is a schematic diagram of a storm surge analysis system for long-term sea level data according to the present invention. Detailed Implementation

[0059] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0060] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0061] Example 1

[0062] like Figure 1 As shown, this invention provides a method for storm surge analysis of long-term sea level data. This method includes the following steps:

[0063] S1. Obtain long-term sea level measurement data.

[0064] S2. Calculate the annual sea level component based on long-term sea level measurement data.

[0065] S3. Tidal harmonic analysis is performed on the long-term sea level measurement data after removing the sea level component to obtain storm surge time series data.

[0066] S4. Perform feature analysis on storm surge time series data to obtain storm surge analysis results.

[0067] Step S1 specifically involves: obtaining measured sea level data from a public database and preprocessing the data; the preprocessing process includes: removing data with a quality control flag of 0 from the measured sea level data, retaining only data with a quality control flag of 1, detecting abrupt changes in the data mean and correcting the average difference of abrupt change points to obtain long-term measured sea level data. A quality control flag of 0 indicates that the data is unusable, and a quality control flag of 1 indicates that the data is usable.

[0068] In this embodiment, the dataset obtained is the actual sea level data from the Sewells Point sea level station in Virginia, USA, in the Global Extreme Sea Level Analysis dataset GESLA-3, which is read by the GeslaDataset function in the Python third-party library gesla.

[0069] The method for detecting abrupt changes in the mean of the data is the Pruned Exact Linear Time (PELT) method. This method defines the criterion function as the sum of the costs of all segmented fragments, aiming to minimize the value of the criterion function. The mathematical expression of the Pruned Exact Linear Time method is as follows:

[0070]

[0071] Where Γ is the index of a set of variable points, i.e., Γ = {t1, t2, ...}, t is the variable point index, y is the input data, V is the criterion function, pen is the measure of the piecewise complexity of variable points, and C is the measure of the sub-signal. The loss function for model fit is the mean-shift model C. L2 As the loss function, pen uses a linear penalty as a constraint on the number of variable points:

[0072]

[0073] pen L2 (Γ)=β|Γ|

[0074] in, Let Γ be the empirical time-series mean between sub-signals a and b, β be a smoothing parameter that measures complexity and good fit, where β > 0, and |Γ| be the cardinality of the change points.

[0075] The specific process of step S2 is as follows: Divide the long-term sea level measurement data by year, calculate the annual median sea level for each year, and subtract the annual median sea level from the long-term sea level measurement data to obtain the annual sea level component. The mathematical expression for this is:

[0076]

[0077] Where X' represents the data after removing the sea level component, i represents the descending sequence number of the sea level data for each year, j represents the year, and N represents the number of time-series sea level data points for each tide gauge station each year.

[0078] The specific process of step S3 is as follows: MATLAB is used to calculate tidal harmonic analysis and remove tidal components. The S_Tide model is used to automatically select the number of tidal constituents and calculate the tidal harmonic constant. The predicted tide level is obtained by subtracting the tidal harmonic analysis from the long-term measured sea level data after removing the sea level component. The mathematical expression of step S3 is as follows:

[0079] H storm_surge =H′-H tide

[0080]

[0081] Among them, H storm_surge H represents the storm surge component, and H' represents the long-term sea level measurement data after removing the sea level component. tide Let Z be the tidal component, Z be the number of tidal constituents in the tidal harmonic analysis, and H be the tidal component. Z Let σ be the tidal amplitude. Z For the tidal angular velocity, g Z The tidal lag angle is T, where T is time.

[0082] It should be noted that in Python, based on the 'mark' identifier and timestamp indicating whether the typhoon affects the target station, storm surge data are added to the typhoon and non-typhoon DataFrames respectively, in order to conduct further results analysis.

[0083] Before step S4, it is also necessary to obtain the storm surge periods affecting the target station and to divide the storm surge data into typhoon and non-typhoon periods. The specific calculation process is as follows:

[0084]

[0085]

[0086] Where S is the spherical distance between the target station and the typhoon center, r is the Earth's radius, taken as 6371.137 km; φ1 and φ2 are the latitudes of the two points, and λ1 and λ2 are the longitudes of the two points, all in radians.

[0087] like Figure 2 As shown, the specific content of feature analysis in step S4 includes:

[0088] S41. Calculate the empirical cumulative distribution function and Gaussian probability function of storm surge data during typhoon and non-typhoon periods, respectively.

[0089] S42. Calculate the relationship between storm surge and typhoon characteristics during typhoon season.

[0090] S43. Calculate the maximum storm surge increase and its maximum water level in chronological order, and then visualize the result.

[0091] It should be noted that this step is based on the storm surge data frames obtained by S3 during typhoon and non-typhoon periods, and uses the `stats.cumfreq` function from the Python third-party library SciPy to calculate, for example... Figure 4 The Empirical Cumulative Distribution Function (ECDF) shown is calculated using the gaussian_kde function in the Python third-party library SciPy, as follows: Figure 5 The Gaussian kernel probability density function (PDF, which describes the distribution of values ​​of a random variable) and the ks_2samp function from the Python third-party library SciPy are used to perform a two-sample Kolmogorov-Smirnov test on storm surges during typhoon and non-typhoon periods. The pyplot.scatter function from the matplotlib library is used to plot a scatter plot of the relationship between storm surge and typhoon characteristics; the pyplot.bar function from the matplotlib library is used to plot the top ten historical storm surge events; the heatmap function from the seaborn library is used to plot the monthly maximum storm surge and maximum water level distribution at the target station; and the pyplot function from the matplotlib library is used to plot the annual maximum storm surge and maximum water level trend at the target station. These methods are used to analyze the storm surge characteristics of the target station.

[0092] In step S41, the empirical cumulative distribution function and the Gaussian probability function are specifically as follows:

[0093]

[0094]

[0095] Where K is the Gaussian kernel function, x i Here, n represents the number of time-series data points, and h represents the bandwidth.

[0096] The relationship between storm surge and typhoon characteristics during typhoon season includes: storm surge during typhoon season and the distance between the typhoon and the target monitoring station at the corresponding time, the minimum air pressure at the typhoon center, the radius of maximum wind speed, and the maximum wind speed information. The radius of maximum wind speed is calculated using the following Willoughby formula:

[0097]

[0098] Among them, R max V is the radius of maximum wind speed.max θ represents the maximum wind speed near the typhoon center, and θ represents the latitude of the typhoon center.

[0099] The calculation of the maximum storm surge and its maximum water level by sorting by time sequence includes: calculating the historical maximum storm surge events and their corresponding water levels, the monthly maximum storm surge and maximum water level, and the annual maximum storm surge and maximum water level; the visualization methods include: drawing distribution maps, drawing trend maps, and performing linear regression fitting on the trends.

[0100] It should be noted that, compared with the prior art, the present invention has the following advantages:

[0101] This invention takes into account the impact of sea-level rise on the time-series sea-level measurement data of tide gauge stations, and is more scientific and reasonable than the traditional method of subtracting the astronomical tide level from the measured tide level.

[0102] The method for removing sea level and tidal components from the time-series sea level measurement data of tide gauge stations is simple and easy to understand, and has good universality.

[0103] The programming process involved in this invention can be implemented through different programming languages ​​such as Python, and has wide portability and applicability, so that it can be used on different systems and platforms.

[0104] Example 2

[0105] Based on the above embodiment 1, combined with Figures 3-9 This embodiment details the specific process of storm surge analysis based on long-term sea level data from the Sewells Point station in Virginia, USA, using the GESLA-3 (Global Extreme Sea Level Analysis) dataset.

[0106] Based on this dataset, the specific implementation steps of the present invention are as follows.

[0107] Step 1: Perform quality control on the data, identify and remove abrupt changes and outliers in the time-series sea level data. The specific steps are as follows:

[0108] The `GeslaDataset` function from the Python third-party library `gesla` was used to read the measured sea level data from the SewellsPoint station in the GESLA-3 dataset. The GESLA-3 dataset contains a quality control flag `use_flag`, where `use_flag` of 1 indicates data availability, and `use_flag` of 0 indicates data unavailability. Sea level data corresponding to `use_flag` of 1 were replaced with NaN values; only sea level data with `use_flag` of 1 were retained and used.

[0109] To check for sea level anomalies in the original time-series sea level data, the PELT function from the Python third-party library ruptures was used to detect abrupt changes in the data mean and to correct the average difference of abrupt change points to adjust the identified jumps or shifts.

[0110] Step Two: After processing in Step One, as follows Figure 3 The solid line represents high-quality time-series data divided by year, with the median sea level for each year calculated. The time-series data is then subtracted from the corresponding median to eliminate the impact of interannual sea-level rise. After removing the sea-level component, as shown... Figure 3 As shown by the dashed line, the influence of sea level on water level has been eliminated.

[0111] Step 3: Use MATLAB's S_Tide tool to perform tidal harmonic analysis on the time series data after removing the sea level component in Step 2. The number of independent points in the S_Tide model is set to 1, and the calculated tidal amplitude and lag angle are constants, which is the same as the result of the classical harmonic analysis using T_Tide; the number of tidal constituents is automatically selected according to the length of the time series sea level data using the "autoselected" mode.

[0112] Step 4: Subtract the predicted tide level obtained from the harmonic analysis in Step 3 from the time series data in Step 2 to obtain the storm surge time series data. Combine two methods to examine the storm surge time series data and remove outliers: first, check whether the water level changes before and after the outlier are normal, i.e., the water level rises within a few hours rather than rapidly; second, determine whether it is during a typhoon season based on publicly available NCEP / NCAR historical sea level pressure analysis datasets.

[0113] Step 5: Using the HURDAT2 North Atlantic Hurricane dataset as the basis for distinguishing between typhoon and non-typhoon periods, the spherical distance between the typhoon center and the target monitoring station is calculated using the Haversine formula, with the Earth's radius taken as 6371.137 km. Typhoons and their periods with a distance less than a set threshold (500 km) are defined as impact periods, while those greater than the set threshold (500 km) are non-impact periods. A new 'mark' column is created in the row corresponding to typhoon periods and assigned the value 'mark' = 1; a new 'mark' column is created in the row corresponding to non-typhoon periods and assigned the value 'mark' = 0. Typhoon and non-typhoon periods are distinguished according to the 'mark' identifier, and the storm surge time-series data extracted in Step 4 is added to the respective DataFrames for typhoon and non-typhoon periods based on the timestamp.

[0114] Step Six: Based on the storm surge data obtained in Step Five for typhoon and non-typhoon periods, conduct analyses on the distribution of storm surge during typhoon and non-typhoon periods, the relationship between storm surge and typhoon characteristics, the top ten historical storm surge events, and the interdecadal trend of storm surge. The specific process is as follows.

[0115] DataFrames representing typhoon and non-typhoon periods were sorted in ascending order of storm surge values. The empirical cumulative distribution function (ECDF) was calculated using the `stats.cumfreq` function from the Python third-party library SciPy. The Gaussian kernel probability density function (PDF) was calculated using the `gaussian_kde` function from SciPy. The Kolmogorov-Smirnov test was performed on the two samples of storm surge values ​​during typhoon and non-typhoon periods using the `ks_2samp` function from SciPy, calculating the Kolmogorov-Smirnov statistic and p-value. The results were plotted using the `pyplot` function. Figure 4 and Figure 5 The ECDF and PDF maps of storm surges during typhoon and non-typhoon periods shown are from the Sewells Point weather station. A Kolmogorov-Smirnov test p-value less than 0.05 indicates significant differences in storm surge distribution between typhoon and non-typhoon periods; a p-value greater than 0.05 indicates similar storm surge distributions between typhoon and non-typhoon periods. Figure 4 and Figure 5 It is known that, compared with the non-typhoon season, the storm surge at Sewell Point station originates from different distributions and shows significant differences. The peak values ​​(i.e., modes) of the PDF during the typhoon season and the non-typhoon season are approximately 0.0m and 0.2m, respectively, indicating that the storm surge during the typhoon season and the non-typhoon season are mainly concentrated at 0.0m and 0.2m. Compared with the non-typhoon season, the longer right tail of the PDF during the typhoon season indicates that the storm surge is greater during the typhoon season.

[0116] A scatter plot of the relationship between storm surge and typhoon characteristics (distance from the station to the typhoon center, minimum central pressure, maximum wind speed radius, and maximum wind speed) at the Sewells Point weather station was plotted using the `pyplot.scatter` function from the Python third-party library matplotlib. The maximum wind speed radius was calculated according to the Willoughby (2004) formula. A linear regression curve was fitted using the `LinearRegression` function from the Python third-party library sklearn, and the correlation coefficient R (ranging from -1 to 1) and p-value were calculated using the `pearsonr` function from SciPy. A higher R value (closer to 1) indicates a stronger positive correlation, a lower R value (closer to -1) indicates a stronger negative correlation, and a p-value of 0 indicates no linear relationship. A p-value < 0.05 was considered statistically significant; otherwise, it was not statistically significant. Figure 6 It can be seen that the storm surge increases in an approximately linear manner as the distance between the monitoring station and the typhoon center decreases, the central pressure decreases, the radius of maximum wind speed decreases, and the maximum wind speed increases, and this pattern is statistically significant.

[0117] The storm surge values ​​during the typhoon season are sorted in descending order, and the storm surge and corresponding water level are filtered based on the first occurrence of the typhoon name, i.e., the maximum storm surge and corresponding water level for each typhoon event. The storm surge intensity level induced by the typhoons with the top ten historical maximum storm surges is evaluated using the custom functions `storm_surge_grade` and `get_color` in Python, and a color mapping is defined. For example, `storm_surge_grade` is defined as five levels, with the maximum storm surge H... storm_surge >2.5m is considered extremely strong, 2.0m <H storm_surge ≤2.5m is considered strong, 1.5m <H storm_surge ≤2.0m is considered strong, 1.0m <H storm_surge ≤1.5m is considered medium, 0.5m <H storm_surge ≤1.0m is considered general; the corresponding color mapping is defined as red, orange, yellow, blue, and green, respectively. The `pyplot.bar` function from the Python third-party library matplotlib is used to plot the image as shown below. Figure 7 The bar chart shown represents the ten largest storm surges and corresponding water levels in history at the Sewells Point weather station. The largest storm surge at Sewells Point was approximately 2.499 m, corresponding to a water level of approximately 3.780 m. The second and third largest surges were induced by Typhoons Isabel and Donna, at 1.710 m and 1.610 m, respectively. The fourth to tenth largest storm surges ranged from 1.100 to 1.400 m. All ten largest storm surges in history were greater than 1.0 m, indicating moderate to strong storms.

[0118] The maximum storm surge and maximum water level during the typhoon season each month were extracted using the `resample` function from the Pandas library in Python. These values ​​were stored in a new DataFrame_1 with the year as the index and the month as the column. The `heatmap` function from the seaborn library in Python was then used to plot the data. Figure 8 The map shows the monthly maximum storm surge and maximum water level distribution at Sewells Point. It can be seen that storm surge at Sewells Point is mainly concentrated from August to October, with the historical maximum storm surge occurring in August 1933; only one storm surge occurred in February and December. The corresponding maximum water level increased from 3.38 m in 1936 to 3.74 m in 2003. Overall, from 1927 to 2021, there was a trend of decreasing maximum storm surge and increasing maximum water level.

[0119] The maximum storm surge and maximum water level for each typhoon season were extracted using the `resample` function from the Pandas library in Python and stored in a new DataFrame_2 indexed by the year of the target station. The `pyplot` function from the matplotlib library in Python was then used to plot the data as shown below. Figure 9 The graph shows the annual maximum storm surge and annual maximum water level variation at the Sewells Point weather station, and the linear regression trend is fitted using the LinearRegression function from the Python third-party library sklearn. Figure 9 It can be seen that the maximum storm surge during the typhoon season shows a downward trend, with a decrease rate of about 5.62 mm / a; while the maximum water level during the typhoon season shows an upward trend, with an increase rate of about 3.40 mm / a.

[0120] Example 3

[0121] like Figure 10 As shown, the present invention also provides a storm surge analysis system for long-term sea level data. The system is used for the storm surge analysis method for long-term sea level data and includes: a data acquisition module, a data processing module, and a feature analysis module.

[0122] The data acquisition module acquires long-term sea level measurement data and inputs it into the data processing module. The data processing module calculates the sea level component for each year based on the long-term sea level measurement data. The long-term sea level measurement data after removing the sea level component is subjected to tidal harmonic analysis to obtain storm surge time series data, which is then input into the feature analysis module. The feature analysis module performs feature analysis on the storm surge time series data and outputs storm surge analysis results.

[0123] In the embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0124] Alternatively, if the above embodiments of the present invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiments of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device.

[0125] The methods described in the various embodiments of the present invention may be performed in whole or in part. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

[0126] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. The icons depicting structural positional relationships in the accompanying drawings are for illustrative purposes only and should not be construed as limiting this patent. Those skilled in the art can make other variations or modifications based on the above description. It is neither necessary nor possible to exhaustively describe all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A method for analyzing storm surge in long-term sea level data, characterized in that, Includes the following steps: S1. Obtain long-term sea level measurement data; S2. Calculate the annual sea level component based on long-term sea level measurement data; specifically: divide the long-term sea level measurement data by year, calculate the annual median sea level for each year, and subtract the annual median sea level from the long-term sea level measurement data to obtain the annual sea level component. The mathematical expression is as follows: in, For the data after removing the sea level component, The sea level data for each year are numbered in descending order. Where N is the year, and N is the number of time-series sea level data points collected by the tide gauge station chief each year; S3. Tidal harmonic analysis is performed on the long-term sea-level measured data after removing the sea-level component to obtain storm surge time-series data. Specifically, MATLAB is used to calculate tidal harmonic analysis and remove tidal components. The S_Tide model is used to automatically select the number of tidal constituents and calculate the tidal harmonic constant. The predicted tide level is obtained by subtracting the tidal harmonic analysis from the long-term sea-level measured data after removing the sea-level component. Its mathematical expression is as follows: in, As a component of storm surge, For long-term sea level measurement data after removing the sea level component, For tidal components, This represents the number of tidal constituents in the tidal harmonic analysis. For tidal amplitude, The tidal angular velocity, To determine the tide ebb angle, It involves time; it also requires obtaining the storm surge periods affecting the target monitoring station and dividing the storm surge data into typhoon and non-typhoon periods. The specific calculation process is as follows: in, The spherical distance between the target station and the typhoon center. The radius of the Earth is taken as 6371.137 km; and The latitudes of the two points are respectively. and These are the longitudes of the two points, both in radians. S4. Perform feature analysis on storm surge time series data to obtain storm surge analysis results.

2. The storm surge analysis method for long-term sea level data according to claim 1, characterized in that, Step S1 specifically involves: obtaining measured sea level data from a public database and preprocessing the obtained data; the preprocessing process is as follows: removing data with a quality control identifier of 0 from the measured sea level data, retaining only data with a quality control identifier of 1, detecting abrupt changes in the mean of the data and correcting the average difference of the abrupt change points to obtain long-term measured sea level data.

3. The storm surge analysis method for long-term sea level data according to claim 2, characterized in that, The method for detecting abrupt changes in the mean of the data is the precise pruning linear-time method, and the mathematical expression of the precise pruning linear-time method is as follows: in, An index for a set of variable points , For variable point indexes, For input data, V For the criterion function, A measure of the complexity of piecewise transformations. C To measure sub-signals The loss function for model fit is the mean-shift model. As a loss function; Then a linear penalty is used as the variable number constraint: in, Sub-signal a and b The empirical mean of the time series between them A smoothing parameter to measure complexity and fit, where , The base number is the variable point.

4. The storm surge analysis method for long-term sea level data according to claim 1, characterized in that, In step S4, the specific content of feature analysis includes: S41. Calculate the empirical cumulative distribution function and Gaussian probability function of storm surge data during typhoon and non-typhoon periods, respectively. S42. Calculate the relationship between storm surge and typhoon characteristics during typhoon season; S43. Calculate the maximum storm surge increase and its maximum water level in chronological order, and then visualize the result.

5. A storm surge analysis method for long-term sea level data according to claim 4, characterized in that, In step S41, the empirical cumulative distribution function and the Gaussian probability function are specifically as follows: in, For Gaussian kernel function, It is time series data. The number of time series data. For bandwidth; The relationship between storm surge during typhoon season and typhoon characteristics includes: storm surge during typhoon season and the corresponding distance between the typhoon and the target monitoring station, the minimum air pressure at the typhoon center, the maximum wind speed radius, and the maximum wind speed information. The maximum wind speed radius is calculated using the following Willoughby formula: in, The radius of maximum wind speed. This is the maximum wind speed near the typhoon center. This refers to the latitude of the typhoon's center.

6. The storm surge analysis method for long-term sea level data according to claim 4, characterized in that, The calculation of the maximum storm surge and its maximum water level by sorting by time sequence includes: calculating the historical maximum storm surge events and their corresponding water levels, the monthly maximum storm surge and maximum water level, and the annual maximum storm surge and maximum water level; the visualization methods include: drawing distribution maps, drawing trend maps, and performing linear regression fitting on the trends.

7. A storm surge analysis system for long-term sea level data, the system being used in the storm surge analysis method for long-term sea level data as described in any one of claims 1-6, characterized in that, It includes: a data acquisition module, a data processing module, and a feature analysis module; The data acquisition module acquires long-term sea level measurement data and inputs it into the data processing module. The data processing module calculates the sea level component for each year based on the long-term sea level measurement data. Specifically, the long-term sea level measurement data is divided by year, and the median sea level for each year is calculated sequentially. The sea level component for each year is obtained by subtracting the median sea level for each year from the long-term sea level measurement data. The mathematical expression for this is: in, For the data after removing the sea level component, The sea level data for each year are numbered in descending order. Let N be the year, and N be the number of long-term sea-level data points from tide gauge stations each year. Tidal harmonic analysis is performed on the long-term sea-level measured data after removing the sea-level component to obtain storm surge time-series data, which is then input into the feature analysis module. Specifically, MATLAB is used to calculate tidal harmonic analysis and remove tidal components. The S_Tide model is used to automatically select the number of tidal constituents and calculate the tidal harmonic constant. The predicted tide level is obtained by subtracting the tidal harmonic analysis from the long-term sea-level measured data after removing the sea-level component. Its mathematical expression is: in, As a component of storm surge, For long-term sea level measurement data after removing the sea level component, For tidal components, This represents the number of tidal constituents in the tidal harmonic analysis. For tidal amplitude, The tidal angular velocity, To determine the tide ebb angle, It involves time; it also requires obtaining the storm surge periods affecting the target monitoring station and dividing the storm surge data into typhoon and non-typhoon periods. The specific calculation process is as follows: in, The spherical distance between the target station and the typhoon center. The radius of the Earth is taken as 6371.137 km; and The latitudes of the two points are respectively. and The longitudes of the two points are shown in radians; the feature analysis module performs feature analysis on the storm surge time series data and outputs the storm surge analysis results.

Images