Method for self-adapting spatial positioning of mean high water spring line based on elevation profile

Abstract

The application discloses an average spring high tide line self-adaptive spatial positioning method based on an elevation profile, belongs to the technical field of coastal zone remote sensing monitoring and tide level spatial analysis, and is characterized in that real terrain point cloud data of a tidal flat in a target coastal region is obtained first, a plurality of local real elevation profiles are generated through the real terrain point cloud data of the tidal flat, and corresponding local height sequences are constructed, an average spring high tide level is calculated through tide level data predicted by a nearby tide level observation station, a positional relationship between each local height sequence and the average spring high tide level is judged, a corresponding distance of the average spring high tide level on the corresponding local real elevation profile is determined, the corresponding distance is converted into a spatial coordinate, and the spatial coordinates of all average spring high tide levels are connected to obtain an average spring high tide line, and the method has the advantages that the consistency and stability of spatial positioning of the average spring high tide level under a complex tidal flat environment are improved.

Description

Technical Field

[0001] This invention relates to the field of coastal zone remote sensing monitoring and tidal level spatial analysis technology, and in particular to an adaptive spatial positioning method for mean high tide line based on elevation profile. Background Technology

[0002] In oceanography, the mean high tide level is a specific tidal characteristic value, describing the average high tide level during spring tides. Its spatial trace is called the mean high tide line. National standards such as the *Hydrological Surveying Standards*, *Chinese Nautical Chart Symbols*, and *National Basic Scale Map Symbols* use the mean high tide line as an important theoretical basis for coastline delineation. However, the mean high tide line is only submerged during spring tides or storm surges, remaining exposed for the rest of the time—less than one percent of its exposure time—posing numerous challenges to its accurate measurement and delineation.

[0003] Existing methods for extracting the mean high tide line typically rely on instantaneous waterline lines from remote sensing imagery, combined with tidal level observation data and topographic slope for estimation. One approach involves establishing the distribution relationship of waterline lines under different tidal conditions using multi-temporal remote sensing imagery, and then using profile fitting or tidal flat slope to estimate the corresponding location of the target high tide level. Another approach utilizes the relationship between digital elevation models and mean slope to shift the instantaneous waterline line towards the land. However, existing methods often rely on multi-temporal waterline lines or overall slope models, lacking direct constraints on the actual local elevation changes of the tidal flat, which can easily lead to spatial positioning errors under complex tidal flat conditions. Considering the complex topographic variations in tidal flat areas, even within the same profile, the elevation change trends in different areas can differ significantly, making it difficult to accurately represent local topographic features using a uniform slope or overall fitting method. Furthermore, due to factors such as tidal level conditions, observation range, and data gaps, some profile elevation sequences may not reach the target high tide level, and existing technologies lack robust solution mechanisms for profiles that do not cross the target high tide level. In addition, traditional LiDAR point clouds are easily affected by vegetation canopy and discrete noise, which can lead to local distortion of nearshore digital elevation models and affect the spatial positioning results of high tide levels. Summary of the Invention

[0004] To address the problems of strong multi-temporal dependence of the waterline, insufficient representation of local topography, and poor spatial positioning stability of high tide levels under complex tidal flat conditions in existing technologies, this invention provides an adaptive spatial positioning method for the mean high tide line based on elevation profiles, which can significantly improve the consistency and stability of spatial positioning of the mean high tide line under complex tidal flat conditions.

[0005] The technical solution adopted by this invention to solve the above-mentioned technical problems is: an adaptive spatial positioning method for the mean high tide line based on elevation profiles, comprising the following specific steps: S1. Obtain ICESat-2 ATL24 point cloud data of the target coastal area in recent years, and combine the noise index and local trend constraint method to denoise the obtained point cloud data to obtain the real topographic point cloud data of the tidal flats of the target coastal area. S2. Spatially thin the real topographic point cloud data of the tidal flats, construct a digital elevation model of the tidal flats, and generate multiple local real elevation profiles along the direction perpendicular to the coast. Then, construct the corresponding local elevation sequence for each local real elevation profile. S3. Obtain the tide level data predicted by the nearby tide level observation station, construct the high tide tide level sequence, and calculate the average high tide level of spring tide using a combination of frequency domain filtering and local extremum identification. S4. Determine the positional relationship between each local elevation sequence and the mean high tide level. When the local elevation sequence crosses the mean high tide level, use linear interpolation to determine the corresponding distance to the mean high tide level on the corresponding local true elevation profile. When the local elevation sequence does not cross the mean high tide level, construct a tail profile point sequence and establish a linear function. Use trend extrapolation to determine the corresponding distance to the mean high tide level on the corresponding local true elevation profile. S5. Convert the corresponding distances of the mean high tide level on each local true elevation profile into spatial coordinates, connect the spatial coordinates of each mean high tide level, and obtain the mean high tide line.

[0006] Furthermore, in step S1, the method for denoising the point cloud data is as follows: S1.1 Determine the strength of the laser beam based on the ICESat-2 satellite's direction of travel information, and select the point cloud data of the strong beam from the acquired ICESat-2 ATL24 point cloud data; S1.2, Define the noise figure η This is used to describe the degree of dispersion of point cloud elevation, and the relationship is: , in: h p Here, H0 represents the elevation of the point cloud data, Median is the median function, and Abs is the absolute value function. S1.3, Based on the noise index η The neighborhood search radius ε is adjusted according to the following formula: , Where: ε0 is the baseline neighborhood search radius, which is calculated from the median distance from each photon point to its neighboring photons; then, the point cloud data of the strong beam is initially noise-removed by combining the number of effective photons in the neighborhood; S1.4 Establish a local trend model between track distance and elevation.h Calculate the relationship between each photon point in the point cloud data of a strong beam and the local trend model Trend. h deviation e p The relation is: , in: d p The distance along the orbit of the photon point; S1.5, Based on the noise index η Determine the residual threshold τ : , Then retain the deviation. e p Photon points located within the residual threshold range are used to obtain real tidal flat topographic point cloud data for the target coastal area.

[0007] Furthermore, in step S2, the method for constructing the digital elevation model of the tidal flat is as follows: S2.1, Establish a width of Δd For each grid cell, the center position of the grid cell is used as the spatial coordinate, and the mode of the elevation of all photon points falling into each grid cell in the real topographic point cloud data of the tidal flat is used as the grid elevation. S2.2. The spatial coordinates and grid elevations of all grid cells are interpolated using the Kriging interpolation method to construct a continuous elevation field for the tidal flat, thus obtaining a digital elevation model of the tidal flat.

[0008] Furthermore, in step S2, the method for constructing the local high-order sequence is as follows: Using the coastal direction as a reference, multiple local true elevation profiles are generated on the tidal flat digital elevation model along a direction perpendicular to the coast. Then, along each local true elevation profile, the width is... Δd Sampling is performed, and the cumulative distance and corresponding elevation are recorded to obtain the corresponding local elevation sequence.

[0009] Furthermore, in step S3, the process of obtaining the mean high tide level is as follows: S3.1 Obtain tide level data predicted by nearby tide level observation stations, extract the daily highest tide level, and construct a high tide tide sequence. H HW ( t ); S3.2. Use Fast Fourier Transform to convert the high tide level sequence H HW ( t The frequency components are transformed from the time domain to the frequency domain. f kand the corresponding frequency coefficient F ( k They are respectively: , , in: k Where N is the index of the frequency component, and N is the number of days. Δt The time interval for the sequence is set to one day. t Indicates the first t Day, where e is the natural constant, i.e., the base of the natural logarithm. i The imaginary unit satisfies i 2 =-1; S3.3 Establishing a frequency selection function G ( f k ), and with frequency coefficient F ( k Multiply by the product, and then filter in the frequency domain to retain the frequency coefficients within the range of the tidal cycle. F f ( k ), Filter out other frequency coefficients: , , in: f low This represents the lowest value of the frequency component within the tidal cycle. f high This represents the highest value of the frequency component within the tidal cycle. S3.4, Filtered frequency coefficients F f ( k Performing an inverse fast Fourier transform converts the frequency domain back to the time domain, yielding a high tide level sequence that retains the cyclical variation characteristics of the spring tides. H HW_f ( t The high tide level of spring tides can be identified using the local extremum method. H HWS ,Right now H HW_f ( t spring ): , , in: t spring This is a spring tide day, where H represents the tide level. S3.5 Calculate the high tide levels of all tidal bores within the time frame of acquiring ICESat-2 ATL24 point cloud data in step S1. H HWS The mean value of the high tide level is used to obtain the mean high tide level of the spring tide. H MHWS .

[0010] Furthermore, in step S4, the method for determining the positional relationship between the local high-tide sequence and the average high-tide level is as follows: Determine the elevation of adjacent profile points in a local elevation sequence and their relationship to the mean high tide level. H MHWS The relationship between the two, and the existence of the first in the high-level procedures of the local authorities. j The and the first j The elevations of the +1 profile points are located at the mean high tide level of the spring tide. H MHWS When both sides are considered, the local elevation sequence is assumed to meet the spatial location conditions for high tide, and the corresponding distance to the mean high tide level on the local true elevation profile is determined by linear interpolation; when the local elevation sequence as a whole does not cross the mean high tide level... H MHWS If the local elevation sequence does not meet the spatial positioning conditions for high tide, the corresponding distance of the average high tide level on the local true elevation profile is determined by trend extrapolation.

[0011] Furthermore, in step S4, linear interpolation is used to determine the corresponding distance of the mean high tide level on the local true elevation profile. d MHWS The relation is: , in: d j , h j These are the first in the local high-order program sequence. j The cumulative distance and corresponding elevation of each profile point d j+1 , h j+1 These are the first in the local high-order program sequence. j +1 cumulative distance and corresponding elevation of profile points.

[0012] Furthermore, in step S4, the corresponding distance of the mean high tide level on the local true elevation profile is determined by trend extrapolation. d MHWS The specific process is as follows: Constructing the tail section point sequence W ( q ): , in: l This represents the number of profile points in the local high-order sequence. r This represents the minimum number of profile points required in a local high-order sequence. q This refers to the number of additional profile points allowed in a local high-resolution sequence. s This is the upper limit on the number of additional profile points allowed. A linear function is established in the tail section point sequence, based on the compensation factor λ and the determination coefficient R of the linear function. 2 Construct a trend evaluation function, and determine the optimal number of additional profile points allowed by maximizing the trend evaluation function. q * : ; Then, an optimal linear function is established in the optimal tail profile point sequence, and the corresponding distance of the mean high tide level on the local true elevation profile is determined by trend extrapolation. d MHWS : , in: a , b These are the slope and intercept of the optimal linear function, respectively. When the coefficient of determination of a linear function R 2 <0.5 or the slope of the optimal linear function a When <0, it is considered that the corresponding local high-order sequence does not meet the stable extrapolation condition, and high-order spatial positioning is not performed.

[0013] Furthermore, in step S5, the spatial coordinates of the average high tide level on the local true elevation profile ( x MHWS , y MHWS It is obtained through the following relational transformation: , Where (X1, Y1) and (X2, Y2) represent the coordinates of the two endpoints of the local true elevation profile, and L represents the distance between the two endpoints of the local true elevation profile.

[0014] Compared with existing technologies, the advantages of this invention are as follows: This method uses the ICESat-2 elevation profile to replace the traditional overall slope model, improving the ability to represent local topography in complex tidal flat areas. Through an elevation crossing discrimination mechanism, it achieves spatial positioning of the average high tide level based on the local true elevation profile, reducing errors caused by traditional overall fitting methods. For local true elevation profiles that do not cross the target high tide level, a tail profile point sequence trend constraint mechanism is introduced to achieve robust solution of high tide level under complex tidal flat conditions. Furthermore, by combining noise index and local trend constraints to adaptively denoise ATL24 point cloud data, the ability to restore the true topography of the tidal flat is improved. Through the synergistic effect of these multiple aspects, the consistency and stability of spatial positioning of the average high tide level in complex tidal flat environments are improved. Attached Figure Description

[0015] Figure 1 This is a flowchart of the present invention; Figure 2 This is a spatial distribution map of point cloud data of two strong ICESat-2 ATL24 beams from the ICESat-2 ATL24 point cloud data of the south bank of Hangzhou Bay in this invention. Figure 3 This is a rendering of the actual terrain restoration effect of ATL24_20210220_gt2l point cloud data according to the present invention; Figure 4 This is a rendering of the actual terrain restoration effect of ATL24_20240815_gt3r point cloud data according to the present invention; Figure 5 This is a distribution map of the actual topographic point cloud data of the southern shore of Hangzhou Bay after noise reduction. Figure 6 This is a schematic diagram illustrating the spatial thinning of point cloud data according to the present invention; Figure 7 This is a digital elevation model of tidal flats generated by Kriging interpolation according to the present invention. Figure 8 This is a sequence diagram of the tide levels predicted by the Ganpu Tide Observation Station in January 2020. Figure 9 This is a graph showing the sequence of high tide levels from January 2020 to July 2024 and the calculated average high tide levels. Figure 10 This is a spatial distribution map of the actual elevation profile on the southern bank of Hangzhou Bay. Figure 11 for Figure 10 A schematic diagram showing the distance to the mean high tide level determined by linear interpolation of the local height sequence corresponding to mid-section 36. Figure 12 for Figure 10A schematic diagram showing the distance to the mean high tide level of spring tide determined by trend extrapolation for the local height sequence corresponding to mid-section 124. Figure 13 for Figure 10 A schematic diagram showing that the local high-level sequence corresponding to the middle section 177 was not spatially located because it did not meet the stable extrapolation conditions; Figure 14 The figure shows the generation result of the mean high tide line on the south bank of Hangzhou Bay according to the present invention, wherein: (a) the figure is a spatial distribution map of the mean high tide line formed by connecting linear interpolation points and trend extrapolation points; (b) the figure is a local enlarged view of the relative position of the mean high tide line and the seawall line; and (c) the figure is a local enlarged view of the relative position of the mean high tide line and the vegetation line. Detailed Implementation

[0016] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0017] like Figure 1 As shown, the adaptive spatial positioning method for the mean high tide line based on the elevation profile includes the following specific steps: S1. Obtain the actual point cloud data of the tidal flats in the target coastal area, specifically: S1.1 Obtain ICESat-2 ATL24 point cloud data of the target coastal area in recent years, determine the strength of the laser beam based on the ICESat-2 satellite's direction of travel information, and select the point cloud data of the strong beam from the acquired ICESat-2 ATL24 point cloud data. S1.2, Define the noise figure η This is used to describe the degree of dispersion of point cloud elevation, and the relationship is: , in: h p Here, H0 represents the elevation of the point cloud data, Median is the median function, and Abs is the absolute value function. S1.3, Based on the noise index η The neighborhood search radius ε is adjusted according to the following formula: , Where: ε0 is the baseline neighborhood search radius, which is calculated from the median distance from each photon point (i.e., point cloud data, one point cloud data is one photon point) to its neighboring photon points; then, the point cloud data of the strong beam is initially noise-removed by combining the number of effective photon points in the neighborhood. S1.4 Establish a local trend model between track distance and elevation. hTo reduce the impact of discrete noise on the true topographic trend of tidal flats, the relative performance of each photon point in the point cloud data of strong beams to the local trend model Trend is calculated. h deviation e p The relation is: , in: d p The distance along the orbit of the photon point; S1.5, Based on the noise index η Determine the residual threshold τ : , Then retain the deviation. e p Photon points located within the residual threshold range are used to obtain real tidal flat topographic point cloud data of the target coastal area. S2. Spatially thin the point cloud data of the tidal flats to construct a digital elevation model of the tidal flats, and generate multiple local true elevation profiles along the direction perpendicular to the coastline. Then, construct a corresponding local elevation sequence for each local true elevation profile, specifically as follows: S2.1, Establish a width of Δd For each grid cell, the center position of the grid cell is used as the spatial coordinate, and the mode of the elevation of all photon points falling into each grid cell in the real topographic point cloud data of the tidal flat is used as the grid elevation. S2.2. The spatial coordinates and grid elevations of all grid cells are interpolated using the Kriging interpolation method to construct a continuous elevation field of the tidal flat and obtain a digital elevation model of the tidal flat. S2.3. Using the coastal direction as a reference direction, generate multiple local real elevation profiles on the tidal flat digital elevation model along the direction perpendicular to the coast. S2.4, then along each local true elevation profile by width Δd (i.e., the spacing) is sampled, the cumulative distance and corresponding elevation are recorded, and the corresponding local elevation sequence is obtained; S3. Obtain the predicted tide level data from nearby tide level observation stations, construct the high tide level sequence, and calculate the mean high tide level using a combination of frequency domain filtering and local extremum identification. Specifically: S3.1 Obtain tide level data predicted by nearby tide level observation stations, extract the daily highest tide level, and construct a high tide tide sequence. H HW ( t ); S3.2. Use Fast Fourier Transform to convert the high tide level sequence H HW (t The time domain is transformed into the frequency domain to reduce the impact of non-spring tide cycle tidal fluctuations on high tide location; its frequency components... f k and the corresponding frequency coefficient F ( k They are respectively: , , in: k Where N is the index of the frequency component, and N is the number of days. Δt The time interval for the sequence is set to one day. t Indicates the first t Day, where e is the natural constant, i.e., the base of the natural logarithm. i The imaginary unit satisfies i 2 =-1; S3.3 Establishing a frequency selection function G ( f k ), and with frequency coefficient F ( k Multiply by the product, and then filter in the frequency domain to retain the frequency coefficients within the range of the tidal cycle. F f ( k ), Filter out other frequency coefficients: , , in: f low This represents the lowest value of the frequency component within the tidal cycle. f high This represents the highest value of the frequency component within the tidal cycle. S3.4, Filtered frequency coefficients F f ( k Performing an inverse fast Fourier transform converts the frequency domain back to the time domain, yielding a high tide level sequence that retains the cyclical variation characteristics of the spring tides. H HW_f ( t The high tide level of spring tides can be identified using the local extremum method. H HWS ,Right now H HW_f ( t spring ): , , in: tspring This is a spring tide day, where H represents the tide level. S3.5 Calculate the high tide levels of all tidal bores within the time frame of acquiring ICESat-2 ATL24 point cloud data in step S1. H HWS The mean value of the high tide level is used to obtain the mean high tide level of the spring tide. H MHWS ; S4. Determine the elevation of adjacent profile points and the mean high tide level in each local elevation sequence. H MHWS Positional relationship between them: When the local high-order sequence contains the first j The and the first j The elevations of the +1 profile points are located at the mean high tide level of the spring tide. H MHWS When considering both sides, the local elevation sequence is assumed to meet the spatial positioning conditions for high tide. Linear interpolation is used to determine the corresponding distance to the mean high tide level on the local true elevation profile. d MHWS The relation is: , in: d j , h j These are the first in the local high-order program sequence. j The cumulative distance and corresponding elevation of each profile point d j+1 , h j+1 These are the first in the local high-order program sequence. j +1 cumulative distance and corresponding elevation of profile points; When the local high-level sequence as a whole does not cross the average high tide level H MHWS If the local elevation profile does not meet the spatial location conditions for high tide, then a trend extrapolation method is used to determine the corresponding distance to the mean high tide level on the local true elevation profile. d MHWS Specifically: Constructing the tail section point sequence W ( q ): , in: l This represents the number of profile points in the local high-order sequence. r This represents the minimum number of profile points required in a local high-order sequence. q This refers to the number of additional profile points allowed in a local high-resolution sequence. sThis is the upper limit on the number of additional profile points allowed. A linear function is established in the tail section point sequence, based on the compensation factor λ and the determination coefficient R of the linear function. 2 Construct a trend evaluation function, and determine the optimal number of additional profile points allowed by maximizing the trend evaluation function. q * : ; Then, an optimal linear function is established in the optimal tail profile point sequence, and the corresponding distance of the mean high tide level on the local true elevation profile is determined by trend extrapolation. d MHWS : , in: a , b These are the slope and intercept of the optimal linear function, respectively. When the coefficient of determination of a linear function R 2 <0.5 or the slope of the optimal linear function a When <0, it is considered that the corresponding local high-order sequence does not meet the stable extrapolation condition, and high-order spatial positioning is not performed; S5. Calculate the corresponding distances of the average high tide level on each local true elevation profile. d MHWS Convert to spatial coordinates ( x MHWS , y MHWS The transformation formula is: , Where (X1, Y1) and (X2, Y2) represent the coordinates of the two endpoints of the local true elevation profile, and L represents the distance between the two endpoints of the local true elevation profile. Then, connect the spatial coordinates of each mean high tide level to obtain the mean high tide line.

[0018] The following example uses the tidal flat area on the south bank of Hangzhou Bay, combined with the attached... Figure 2-14 The above adaptive spatial positioning method will be further explained and its effectiveness verified.

[0019] (1) Forty scenes of ICESat-2 ATL24 satellite point cloud data were collected from 2020 to 2024 on the south bank of Hangzhou Bay. Strong beam point cloud data were extracted based on sc_orient information, resulting in 89 along-orbit point cloud data. The spatial distribution of the gt2l point cloud data on February 20, 2021 and the gt3r point cloud data on August 15, 2024 is as follows: Figure 2 As shown; Due to the significant differences in the dispersion of point cloud data from different beams, an empirical elevation range was set to enhance the ability to restore the true topography of the tidal flats. H 0 Given a value of 20, calculate the noise index of the two point cloud datasets mentioned above. η And according to the noise index η Adaptive adjustment of neighborhood search radius ε Preliminary noise removal is performed on discrete photon points; for example, the noise in the gt2l point cloud data from February 20, 2021, is relatively low, with a noise index of [missing information]. η =0.1288, corresponding to a neighborhood search radius ε=0.2406. After initial denoising, the number of photons decreased from 17543 to 14554. Figure 3 As shown; in contrast, the gt3r point cloud data from August 15, 2024, has significantly higher noise levels, with a noise index of [insert noise level here]. η =0.6577, corresponding to the neighborhood search radius ε =0.1640. After initial noise reduction, the number of photons decreased from 52280 to 20707. Figure 4 As shown; Based on the initial denoising, to reduce the impact of discrete noise on the true topographic trend of the tidal flats, a local trend model between track distance and elevation is further established, and fine denoising is performed based on the deviation of photon points relative to the local trend model; for example, the residual threshold corresponding to the gt2l point cloud data on February 20, 2021. τ =0.25, after fine noise reduction, the number of photons further decreased to 12846, such as Figure 3 As shown; the residual threshold corresponding to the GT3R point cloud data on August 15, 2024. τ =0.35, after fine noise reduction, the number of photons further decreased to 11349, such as Figure 4 As shown.

[0020] (2) After noise reduction, the number of effective photon points of ATL24 on the south bank of Hangzhou Bay from 2020 to 2024 is 811,045, and they are densely distributed along the track, such as... Figure 5 As shown. Establish the width Δd Using a 30m grid cell, spatial thinning was performed on the real topographic point cloud data of the southern shore of Hangzhou Bay. The mode of elevation of all photon points falling within each grid cell was then used as the grid elevation to reduce the impact of local outliers on the true elevation representation of the tidal flats. For example, if 40 photon points fall within a certain grid cell, and the elevation of 4.092 m occurs most frequently (3 times), then 4.092 m is the mode of elevation for that grid cell and is therefore used as the corresponding grid elevation. Figure 6 As shown; Through the above method, a total of 30,711 grid elevation points were obtained, reducing the original 811,045 photon points to 30,711 grid elevation points. Then, using the center position of each grid cell as the spatial coordinate and the grid elevation as the elevation attribute, a continuous elevation field of the tidal flat was constructed using the Kriging interpolation method, resulting in a digital elevation model of the tidal flat with a spatial resolution of 30m. Figure 7 As shown.

[0021] (3) Obtain the predicted tide level data from the Ganpu Tide Level Observation Station from 2020 to 2024, including the predicted tide level changes in January 2020 as follows: Figure 8 As shown, the daily highest tide level is extracted to construct a high tide level sequence. H HW ( t The high tide level sequence was analyzed using Fast Fourier Transform. H HW ( t Converting from the time domain to the frequency domain, on the southern shore of Hangzhou Bay, the high tide cycle is approximately 15 days, with fluctuations generally not exceeding 3 days. Therefore, the high tide cycle range is set to 12 to 18 days, corresponding to the lowest value in the frequency range. f low =1 / 18, highest value f high =1 / 12, construct a frequency filtering function, and retain the frequency coefficients within the spring tide cycle range through frequency domain filtering. After filtering out other frequency coefficients, perform an inverse fast Fourier transform to recover the high tide level sequence in the time domain. H HW_f ( t Then, the high tide level of the spring tide was identified using the local extremum method. For example, in January 2020, two high tide levels were identified, at 3.690m and 3.210m, corresponding to 15:00 on January 13th and 15:00 on January 26th, respectively. Figure 8 As shown, the monthly high tide levels from January 2020 to July 2024 were obtained using the above method, and their average values ​​were calculated to obtain the average high tide level on the south bank of Hangzhou Bay. H MHWS =3.830 m, as shown in the attached document Figure 9 As shown.

[0022] (4) Using the coastal direction as a reference, a total of 196 local true elevation profiles were generated on the digital elevation model of the tidal flat along the direction perpendicular to the coast. Local elevation sequence was extracted at a spacing of Δd=30m. One local true elevation profile corresponds to one local elevation sequence. Among the 196 local true elevation profiles, a total of 41 local elevation sequence spans the mean high tide level of spring tides. H MHWS .For example, Figure 11In the local high-order sequence corresponding to section 36 shown, the first... j The and the first j The elevations of the +1 profile points are located at the mean high tide level of the spring tide. H MHWS Therefore, linear interpolation is used to determine the corresponding distance to the mean high tide level on both sides. For those that have not crossed the average high tide level H MHWS The local high-order sequence is used to construct the tail profile point sequence, and a linear function is established. In this example, the minimum number of profile points required is set. r =6, the maximum number of additional profile points allowed. s =10, and a trend evaluation function is constructed based on the compensation factor λ and the determination coefficient R² of the linear function. The optimal number of additional profile points allowed is determined by maximizing the trend evaluation function. For example, Figure 12 The local elevation sequence corresponding to section 124 shown does not cross the mean high tide level of a spring tide. H MHWS The corresponding distance to the average high tide level is determined by trend extrapolation. Furthermore, when the coefficient of determination of the linear function is low (R² < 0.5) or the slope of the optimal linear function is negative ( a When <0), the corresponding local high-order sequence is considered not to meet the stable extrapolation condition, and high-order spatial positioning is not performed. For example, Figure 13 The local high-order sequence corresponding to section 177 shown was not spatially located due to the poor stability of the linear function.

[0023] (5) Convert the corresponding distances of the mean high tide level on each local true elevation profile into spatial coordinates to obtain the corresponding spatial positioning points of the mean high tide level. A total of 176 high tide level spatial positioning points are obtained, including interpolated positioning points and extrapolated positioning points. Connect each spatial positioning point in sequence to obtain the mean high tide line of the south bank of Hangzhou Bay, such as... Figure 14 As shown in (a).

[0024] The results show that the mean high tide line obtained by this invention is partially located inside the seawall (see attached diagram). Figure 14 (b) shown) or within the vegetation community (attached) Figure 14 (c) shows that it cannot be directly extracted from remote sensing images of a single high tide period, further verifying the effectiveness and importance of the present invention in spatially locating the mean high tide level in complex tidal flat environments.

[0025] The scope of protection of this invention includes, but is not limited to, the above embodiments. The scope of protection is defined by the claims. Any substitutions, modifications, or improvements to this technology that are easily conceived by those skilled in the art fall within the scope of protection of this invention.

Claims

1. An adaptive spatial positioning method for the mean high tide line of spring tides based on elevation profiles, characterized in that... The specific steps include the following: S1. Obtain ICESat-2 ATL24 point cloud data of the target coastal area in recent years, and combine the noise index and local trend constraint method to denoise the obtained point cloud data to obtain the real topographic point cloud data of the tidal flats of the target coastal area. S2. Spatially thin the real topographic point cloud data of the tidal flats, construct a digital elevation model of the tidal flats, and generate multiple local real elevation profiles along the direction perpendicular to the coast. Then, construct the corresponding local elevation sequence for each local real elevation profile. S3. Obtain the tide level data predicted by the nearby tide level observation station, construct the high tide tide level sequence, and calculate the average high tide level of spring tide using a combination of frequency domain filtering and local extremum identification. S4. Determine the positional relationship between each local elevation sequence and the mean high tide level. When the local elevation sequence crosses the mean high tide level, use linear interpolation to determine the corresponding distance to the mean high tide level on the corresponding local true elevation profile. When the local elevation sequence does not cross the mean high tide level, construct a tail profile point sequence and establish a linear function. Use trend extrapolation to determine the corresponding distance to the mean high tide level on the corresponding local true elevation profile. S5. Convert the corresponding distances of the mean high tide level on each local true elevation profile into spatial coordinates, connect the spatial coordinates of each mean high tide level, and obtain the mean high tide line.

2. The adaptive spatial positioning method for the mean high tide line based on elevation profile as described in claim 1, characterized in that: In step S1, the method for denoising the point cloud data is as follows: S1.1 Determine the strength of the laser beam based on the ICESat-2 satellite's direction of travel information, and select the point cloud data of the strong beam from the acquired ICESat-2 ATL24 point cloud data; S1.2, Define the noise figure η This is used to describe the degree of dispersion of point cloud elevation, and the relationship is: ， in: h p Here, H0 represents the elevation of the point cloud data, Median is the median function, and Abs is the absolute value function. S1.3, Based on the noise index η The neighborhood search radius ε is adjusted according to the following formula: ， Where: ε0 is the baseline neighborhood search radius, which is calculated from the median distance from each photon point to its neighboring photons; then, the point cloud data of the strong beam is initially noise-removed by combining the number of effective photons in the neighborhood; S1.4 Establish a local trend model between track distance and elevation. h Calculate the relationship between each photon point in the point cloud data of a strong beam and the local trend model Trend. h deviation e p The relation is: ， in: d p The distance along the orbit of the photon point; S1.5, Based on the noise index η Determine the residual threshold τ : ， Then retain the deviation. e p Photon points located within the residual threshold range are used to obtain real tidal flat topographic point cloud data for the target coastal area.

3. The adaptive spatial positioning method based on the mean high tide line of the elevation profile as described in claim 1, characterized in that: In step S2, the method for constructing the digital elevation model of the tidal flat is as follows: S2.1, Establish a width of Δd For each grid cell, the center position of the grid cell is used as the spatial coordinate, and the mode of the elevation of all photon points falling into each grid cell in the real topographic point cloud data of the tidal flat is used as the grid elevation. S2.

2. The spatial coordinates and grid elevations of all grid cells are interpolated using the Kriging interpolation method to construct a continuous elevation field for the tidal flat, thus obtaining a digital elevation model of the tidal flat.

4. The adaptive spatial positioning method based on the mean high tide line of the elevation profile as described in claim 3, characterized in that: In step S2, the method for constructing the local high-order sequence is as follows: Using the coastal direction as a reference, multiple local true elevation profiles are generated on the tidal flat digital elevation model along a direction perpendicular to the coast. Then, along each local true elevation profile, the width is... Δd Sampling is performed, and the cumulative distance and corresponding elevation are recorded to obtain the corresponding local elevation sequence.

5. The adaptive spatial positioning method for the mean high tide line based on elevation profile as described in claim 1, characterized in that: In step S3, the process of obtaining the mean high tide level is as follows: S3.1 Obtain tide level data predicted by nearby tide level observation stations, extract the daily highest tide level, and construct a high tide tide sequence. H HW ( t ); S3.

2. Use Fast Fourier Transform to convert the high tide level sequence H HW ( t The frequency components are transformed from the time domain to the frequency domain. f k and the corresponding frequency coefficient F ( k They are respectively: ， ， in: k Where N is the index of the frequency component, and N is the number of days. Δt The time interval for the sequence is set to one day. t Indicates the first t Day, where e is the natural constant, i.e., the base of the natural logarithm. i The imaginary unit satisfies i 2 =-1; S3.3 Establishing a frequency selection function G ( f k ), and with frequency coefficient F ( k Multiply by the product, and then filter in the frequency domain to retain the frequency coefficients within the range of the tidal cycle. F f ( k ), Filter out other frequency coefficients: ， ， in: f low This represents the lowest value of the frequency component within the tidal cycle. f high This represents the highest value of the frequency component within the tidal cycle. S3.4, Filtered frequency coefficients F f ( k Performing an inverse fast Fourier transform converts the frequency domain back to the time domain, yielding a high tide level sequence that retains the cyclical variation characteristics of the spring tides. H HW_f ( t The high tide level of spring tides can be identified using the local extremum method. H HWS ,Right now H HW_f ( t spring ): ， ， in: t spring This is a spring tide day, where H represents the tide level. S3.5 Calculate the high tide levels of all tidal bores within the time frame of acquiring ICESat-2 ATL24 point cloud data in step S1. H HWS The mean of the high tide level is obtained. H MHWS .

6. The adaptive spatial positioning method for the mean high tide line based on elevation profile as described in claim 1, characterized in that: In step S4, the method for determining the positional relationship between the local high-level sequence and the mean high tide level is as follows: Determine the elevation of adjacent profile points in a local elevation sequence and their relationship to the mean high tide level. H MHWS The relationship between the two, and the existence of the first in the high-level procedures of the local authorities. j The and the first j The elevations of the +1 profile points are located at the mean high tide level of the spring tide. H MHWS When both sides are considered, the local elevation sequence is assumed to meet the spatial location conditions for high tide, and the corresponding distance to the mean high tide level on the corresponding local true elevation profile is determined by linear interpolation; when the local elevation sequence as a whole does not cross the mean high tide level... H MHWS If the local elevation sequence does not meet the spatial positioning conditions for high tide, the corresponding distance of the average high tide level on the local true elevation profile is determined by trend extrapolation.

7. The adaptive spatial positioning method for the mean high tide line based on elevation profile as described in claim 6, characterized in that: In step S4, linear interpolation is used to determine the corresponding distance of the mean high tide level on the local true elevation profile. d MHWS The relation is: ， in: d j , h j These are the first in the local high-order program sequence. j The cumulative distance and corresponding elevation of each profile point d j+1 , h j+1 These are the first in the local high-order program sequence. j +1 cumulative distance and corresponding elevation of profile points.

8. The adaptive spatial positioning method for the mean high tide line based on the elevation profile as described in claim 6, characterized in that: In step S4, the distance corresponding to the mean high tide level on the local true elevation profile is determined by trend extrapolation. d MHWS The specific process is as follows: Constructing the tail section point sequence W ( q ): ， in: l This represents the number of profile points in the local high-order sequence. r This represents the minimum number of profile points required in a local high-order sequence. q This refers to the number of additional profile points allowed in a local high-resolution sequence. s This is the upper limit on the number of additional profile points allowed. A linear function is established in the tail section point sequence, based on the compensation factor λ and the determination coefficient R of the linear function. 2 Construct a trend evaluation function, and determine the optimal number of additional profile points allowed by maximizing the trend evaluation function. q * : ； Then, an optimal linear function is established in the optimal tail profile point sequence, and the corresponding distance of the mean high tide level on the local true elevation profile is determined by trend extrapolation. d MHWS : ， in: a , b These are the slope and intercept of the optimal linear function, respectively. When the coefficient of determination of a linear function R 2 <0.5 or the slope of the optimal linear function a When <0, it is considered that the corresponding local high-order sequence does not meet the stable extrapolation condition, and high-order spatial positioning is not performed.

9. The adaptive spatial positioning method for the mean high tide line based on elevation profile as described in claim 1, characterized in that: In step S5, the spatial coordinates of the average high tide level on the local true elevation profile are ( x MHWS , y MHWS It is obtained through the following relational transformation: ， Where (X1, Y1) and (X2, Y2) represent the coordinates of the two endpoints of the local true elevation profile, and L represents the distance between the two endpoints of the local true elevation profile.

Images