Analysis and dynamic simulation of topographic evolution in the Yangtze River Estuary based on multi-source data fusion

Abstract

The present invention discloses a method for terrain evolution analysis and dynamic simulation of the Yangtze River Estuary based on multi-source data fusion, which relates to the technical field of terrain evolution analysis, and includes the following steps: according to the monitoring needs of the tidal flat area, the satellite remote sensing platform sets a preliminary expected monitoring cycle to regularly obtain image data of the tidal flat area to achieve real-time monitoring of the tidal flat area; after each acquisition of a new satellite image, the new satellite image is aligned with the satellite image acquired last time. The present invention divides the tidal flat area into two categories: fast evolution and slow evolution, maintains the initial cycle for the stable change area, saves resources; shortens the cycle for the drastic change area, improves the data acquisition frequency and response speed, effectively solves the problem of traditional monitoring lag, can timely capture the rapid changes in terrain, optimize monitoring efficiency, reduce the risk of ecological damage, and provide accurate decision-making support for tidal flat protection and management.

Description

Technical Field

[0001] The present invention relates to the technical field of terrain evolution analysis, and in particular to a terrain evolution analysis and dynamic simulation method of the Yangtze River Estuary based on multi-source data fusion. Background Art

[0002] The analysis and dynamic simulation of the topographic evolution of the Yangtze River Estuary based on multi-source data fusion refers to the process of comprehensively analyzing the topographic changes in the Yangtze River Estuary region by integrating data from multiple sources (such as remote sensing data, geological exploration data, tidal monitoring data, hydrological and meteorological data, etc.), and establishing a dynamic simulation model to predict future topographic changes in the region. This process first involves fusing data from different sensors, equipment or technologies to improve the timeliness, accuracy and completeness of the data; then, through data processing and analysis, the regularity and trend of the topographic evolution of the Yangtze River Estuary, such as riverbed sedimentation, tidal influence, coastal erosion and other phenomena, are identified. Ultimately, through dynamic simulation, the impact of topographic changes in the Yangtze River Estuary can be evaluated under different scenarios, helping decision makers to take appropriate management and protection measures to prevent disasters and optimize regional resource utilization and environmental protection.

[0003] Tidal flats are shallow areas formed by the accumulation of sediments under the action of water flow in tidal zones such as estuaries and coasts. They are usually located in the intertidal zone or below the low tide level. Tidal flats are an important ecosystem with rich biodiversity and are the habitat of many water birds, fish and other plants and animals. They also have important ecological functions, such as filtering pollutants, regulating water quality, mitigating the impact of ocean storm surges, and protecting inland areas from seawater erosion. In addition, tidal flats are important natural resources.

[0004] The role of tidal flat terrain evolution analysis and dynamic simulation is to gain an in-depth understanding of the changing patterns and development trends of tidal flat areas, thereby providing a scientific basis for the rational development and protection of tidal flat resources. By analyzing the historical evolution of tidal flat terrain, we can reveal the change patterns of estuaries and coastal areas due to the combined effects of natural factors (such as tides, currents, hydrodynamics, etc.) and human activities (such as reclamation, engineering construction, etc.). Dynamic simulation can predict future trends in tidal flat terrain changes based on historical data and current conditions, and assess its impact on the ecological environment, resource utilization, engineering construction, etc. Such analysis and simulation are helpful in formulating sustainable utilization plans for tidal flat resources, optimizing engineering planning, reducing environmental risks, protecting ecosystems, and ensuring the long-term stability of tidal flat functions.

[0005] The prior art has the following deficiencies:

[0006] Satellite images play an important role in the analysis and dynamic simulation of tidal flat terrain evolution. They can provide long-term monitoring data in a wide range of areas, help identify tidal flat distribution, tidal changes, and terrain evolution trends. Through multispectral sensors, satellites can regularly obtain images of tidal flat areas and capture surface changes in different periods. These data not only help analyze changes in tidal flat area, but also monitor the impact of factors such as tides, currents, and storm surges on tidal flat surface deposition and scouring, thereby revealing the dynamic changes of tidal flat terrain. In the prior art, satellite images are usually obtained through regular monitoring in the analysis and dynamic simulation of tidal flat terrain evolution. If the images of the tidal flat area obtained by two adjacent satellites change greatly, and the data is still obtained by regular monitoring, regular monitoring depends on a fixed time interval, so it may not be possible to capture the rapid dynamics of tidal flat terrain changes in time (such as sudden floods, storm surges, or drastic changes in terrain caused by human activities). If the tidal flat area has changed significantly between two monitoring cycles, especially catastrophic changes (such as sudden scouring of the tidal flat or deposition of a large amount of soil), delayed monitoring will lead to delayed reaction and response. This may result in the damage to the tidal flat ecological environment not being discovered in a timely manner, thus missing the best opportunity for prevention or response measures, causing serious ecological losses or human property losses.

[0007] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0008] The purpose of the present invention is to provide a method for analyzing and dynamically simulating the topographic evolution of the Yangtze River Estuary based on multi-source data fusion. The tidal flat area is divided into a fast evolution process and a slow evolution process according to the actual changes in the topographic evolution of the tidal flat area. For areas with relatively stable changes, the initial monitoring cycle is maintained to save resources; while for areas with drastic changes, the monitoring cycle is dynamically shortened, and the data acquisition frequency and response speed are significantly improved. This dynamic adjustment strategy effectively solves the hysteresis problem existing in the traditional regular monitoring method, and can timely capture the rapid changes of the tidal flat topography (such as sudden deposition, scouring or ecological risks), thereby ensuring that key changes can be quickly identified, and providing accurate and real-time decision support for the protection of the tidal flat ecological environment and resource management. This flexible monitoring method not only optimizes the monitoring efficiency, but also significantly reduces the risk of ecological environmental damage, and realizes efficient and targeted monitoring resource allocation to solve the problems in the above-mentioned background technology.

[0009] In order to achieve the above object, the present invention provides the following technical solution: a method for analyzing and dynamically simulating the topographic evolution of the Yangtze River Estuary based on multi-source data fusion, comprising the following steps:

[0010] According to the monitoring needs of the tidal flat area, the satellite remote sensing platform sets a preliminary expected monitoring cycle to regularly obtain image data of the tidal flat area to achieve real-time monitoring of the tidal flat area;

[0011] After each acquisition of a new satellite image, the new satellite image is registered with the previously acquired satellite image to ensure spatial alignment of the two satellite images. After registration, the image difference analysis method is used to compare the two adjacent satellite images, focusing on areas with terrain changes;

[0012] After completing the image comparison, the key features of the tidal flat terrain changes are extracted. After analyzing the extracted key features, the analyzed features are passed as input data to the pre-trained convolutional neural network, and the convolutional neural network is used to analyze and predict the change trend of the tidal flat terrain;

[0013] Through the prediction results of convolutional neural network, the terrain changes in the tidal flat area are divided into two categories: fast evolution process and slow evolution process;

[0014] For tidal flat areas predicted to be undergoing a slow evolution process, the initially set expected monitoring cycle should be maintained to conduct regular monitoring of the tidal flat areas;

[0015] For tidal flat areas predicted to be in a rapid evolution process, the expected monitoring period is shortened based on the predicted results of the terrain evolution of the tidal flat areas to respond to the rapid changes in the tidal flat areas in a timely manner.

[0016] Preferably, the new satellite image is registered with the previously acquired satellite image to ensure spatial alignment of the two images. The specific steps are as follows:

[0017] Before image registration, the acquired satellite images are preprocessed;

[0018] Points that accurately correspond to geographic features in the two satellite images are selected as registration control points. In order to ensure that the registration control points are widely distributed and representative, multiple points located at the edge and center of the image are selected to achieve global alignment;

[0019] After the control points are selected, the registration algorithm is used to align the images. Specifically, the geometric transformation relationship between the corresponding control points in the two satellite images is calculated, and the displacement, rotation or scaling of one of the satellite images is adjusted until the two satellite images completely overlap at the control points.

[0020] Preferably, after completing the image comparison, key features of the change in the tidal flat terrain are extracted, wherein the extracted key features include the fluctuation of the tidal flat surface reflectance between two adjacent image acquisitions and the change of plant coverage in the tidal flat ecosystem. After analyzing the fluctuation of the tidal flat surface reflectance between two adjacent image acquisitions and the change of plant coverage in the tidal flat ecosystem, a reflectance fluctuation quantification value and a plant coverage decrease quantification value are generated respectively. The reflectance fluctuation quantification value is used to quantify the change amplitude of the tidal flat surface reflectance, reflecting the distribution change of the tidal flat surface material. The plant coverage decrease quantification value is used to quantify the degree of reduction in plant coverage in the tidal flat ecosystem, reflecting the deterioration of the health of the tidal flat vegetation.

[0021] Preferably, after obtaining the reflectivity fluctuation quantification values ​​and the plant coverage decrease quantification values ​​generated by analyzing the changes in the tidal flat terrain, the analyzed reflectivity fluctuation quantification values ​​and the plant coverage decrease quantification values ​​are passed as input data to a pre-trained convolutional neural network, and a terrain evolution index is generated based on the convolutional neural network, and the changing trend of the tidal flat terrain is predicted by the terrain evolution index.

[0022] Preferably, the terrain evolution index generated when analyzing and predicting the change trend of the tidal flat terrain through the convolutional neural network is compared and analyzed with a preset terrain evolution index reference threshold, and the terrain change situation of the tidal flat area is divided. The specific division steps are as follows:

[0023] If the terrain evolution index generated when analyzing and predicting the changing trend of the tidal flat terrain is greater than the pre-set terrain evolution index reference threshold, the terrain change in the tidal flat area is classified as a rapid evolution process;

[0024] If the terrain evolution index generated when analyzing and predicting the changing trend of the tidal flat terrain is less than or equal to a preset terrain evolution index reference threshold, the terrain change in the tidal flat area is classified as a slow evolution process.

[0025] Preferably, the specific steps of analyzing the fluctuation of the reflectivity of the tidal flat surface between two adjacent image acquisitions and generating the reflectivity fluctuation quantization value are as follows:

[0026] The spectral data difference analysis of two adjacent remote sensing images of the tidal flat area is performed to extract the local reflectivity gradient change value of each pixel. The calculation expression of the local reflectivity gradient change value is: ,in, Pixel The local reflectivity gradient change value, i and j represent the row index and column index of the pixel in the remote sensing image, respectively. Indicates that in the image at time t, the pixel The reflectivity value in band b, Indicates that in the image at time t+1, the pixel The reflectance value in band b, B is the total number of bands in the multispectral image, and t and t+1 represent the time of two remote sensing image acquisitions;

[0027] For the gradient results of reflectivity changes, we further analyze their spatial change patterns and assign weights to each pixel to reflect the contribution of local changes to the overall area. The calculation formula of the weight matrix is: ,in, Represents pixel The spatial variation weight of Represents the Euclidean distance from the pixel to the center of the tidal flat area. , are the coordinates of the region center, is the spatial weight parameter, which controls the influence of pixels far from the center. and Represents pixel points The horizontal and vertical coordinates of

[0028] The local reflectivity gradient change value and spatial variation weight of each pixel are combined to calculate the reflectivity fluctuation quantization value of the entire tidal flat area. The calculation expression is: ,in, is the reflectivity fluctuation quantification value of the tidal flat area, M and N are the number of pixel rows and columns of the tidal flat area image, respectively. is the maximum reflectance change of the pixel point in each spectral band, reflecting the extreme change effect in spectral change. .

[0029] Preferably, the specific steps of analyzing the change in plant coverage in the tidal flat ecosystem between two adjacent image acquisitions to generate a quantitative value of plant coverage decrease are as follows:

[0030] The normalized vegetation index is calculated by the near-infrared band and the red light band in the remote sensing image to quantify the plant coverage of each pixel. The normalized vegetation index can effectively reflect the growth status and density of vegetation. As the characteristic of tidal flat plant coverage, the calculation result will form a pixel-level plant coverage distribution map, providing basic data for subsequent analysis. The calculation expression of the normalized vegetation index is: , where Indicates the coordinate point corresponding to time t The normalized difference vegetation index on Indicates the coordinate point corresponding to time t The near-infrared reflectance on the surface reflects the photosynthetic activity of vegetation. Indicates the coordinate point corresponding to time t The red reflectance on the surface reflects the amount of red light absorbed by the vegetation;

[0031] By calculating the time change rate of the normalized vegetation index, the change in plant coverage between two consecutive image acquisitions is extracted, the dynamic change of plant coverage during the entire monitoring period is quantified, and short-term drastic changes or long-term stable changes are captured. The calculation expression of the change in plant coverage is: , where Indicates the coordinate point The change in plant coverage reflects the dynamic trend. represents the instantaneous rate of change of plant coverage, which is calculated by the derivative of time t, and They respectively represent the start time and end time of the interval between two adjacent image acquisitions;

[0032] By taking a weighted average of the changes in all pixels in the tidal flat area and combining the ecological sensitivity and environmental exposure factors of each pixel, a quantitative value of plant coverage decrease is generated. The quantitative value of plant coverage decrease reflects the overall reduction in tidal flat vegetation coverage. The calculation expression of the quantitative value of plant coverage decrease is: , where is a quantitative value for the decline in plant cover, quantifying the overall severity of the reduction in tidal flat vegetation. is a weight factor used to measure the ecological sensitivity of each pixel. It represents the regional impact factor, which represents the environmental exposure degree of the pixel point, and R is the set of all pixels in the tidal flat area.

[0033] Preferably, for the tidal flat area predicted to be in a rapid evolution process, the expected monitoring period is shortened based on the predicted results of the terrain evolution of the tidal flat area. The specific steps are as follows:

[0034] For the case where the tidal flat area is predicted to be a rapid evolution process, the terrain evolution difference between the terrain evolution index and the pre-set terrain evolution index reference threshold is first calculated, and the calculation expression is:

[0035] , is the topographic evolution index of the current tidal flat area, which is predicted by the convolutional neural network model. It is a pre-set reference threshold of terrain evolution index, which is used to distinguish the rate of change of tidal flat terrain. It is the difference in terrain evolution, representing the degree of terrain change;

[0036] After obtaining the terrain evolution difference, the expected monitoring period of the tidal flat area is adjusted according to the terrain evolution difference. In order to respond to the rapid changes of the tidal flat terrain in time, the monitoring period is dynamically shortened according to the size of the terrain evolution index. The expression for adjusting the expected monitoring period is: , where is the actual monitoring period after adjustment, is the expected monitoring period initially set, is a scaling factor that controls the degree of shortening of the monitoring period. represents the maximum topographic evolution index observed during past tidal flat monitoring periods, serving as a benchmark to assess the relative rate of change in the current area;

[0037] The adjusted actual monitoring cycle is applied to the monitoring process of the tidal flat area. The shortened monitoring cycle will achieve more frequent image acquisition through the satellite remote sensing platform, detect the changes in the tidal flat area in a timely manner, and reset the time interval for acquiring images. The specific expression is: , where The next monitoring time is The time point of the last monitoring.

[0038] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0039] The present invention divides the tidal flat area into a fast evolution process and a slow evolution process according to the actual changes in the terrain evolution of the tidal flat area. For areas with relatively stable changes, the initial monitoring cycle is maintained to save resources; while for areas with drastic changes, the monitoring cycle is dynamically shortened, significantly improving the data collection frequency and response speed. This dynamic adjustment strategy effectively solves the hysteresis problem existing in the traditional regular monitoring method, and can timely capture the rapid changes in the tidal flat terrain (such as sudden sedimentation, scouring or ecological risks), thereby ensuring that key changes can be quickly identified, and providing accurate and real-time decision support for the protection of the tidal flat ecological environment and resource management. This flexible monitoring method not only optimizes the monitoring efficiency, but also significantly reduces the risk of ecological environmental damage, and realizes efficient and targeted monitoring resource allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0041] Figure 1 The present invention is a flow chart of the method for analyzing and dynamically simulating the topographic evolution of the Yangtze River Estuary based on multi-source data fusion. DETAILED DESCRIPTION

[0042] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0043] The present invention provides Figure 1 The method for analyzing and dynamically simulating the topographic evolution of the Yangtze River Estuary based on multi-source data fusion includes the following steps:

[0044] According to the monitoring needs of the tidal flat area, the satellite remote sensing platform sets a preliminary expected monitoring cycle to regularly obtain image data of the tidal flat area to achieve real-time monitoring of the tidal flat area;

[0045] The expected monitoring cycle refers to the time interval for obtaining satellite images in the tidal flat area monitoring plan, which is initially set based on the characteristics of tidal flat topography changes and research needs. The expected monitoring cycle is usually determined based on the topographic dynamics and seasonal changes of the tidal flat area and the influence of natural factors such as tides and storm surges. It is to ensure that sufficient remote sensing data can be obtained at appropriate time intervals to comprehensively evaluate the long-term evolution of the tidal flat topography.

[0046] The expected monitoring cycle should meet the following basic requirements:

[0047] First, the expected monitoring period needs to be set to be able to capture key changes in the tidal flat area, such as tidal influence, scouring, sedimentation and other topographic change characteristics;

[0048] Second, the expected monitoring period should take into account the feasibility of technology and resources, including the observation frequency of the satellite platform, data processing capabilities, and budget constraints;

[0049] Finally, the expected monitoring cycle needs to balance the frequency and cost of data acquisition to avoid waste of resources due to overly frequent monitoring, while ensuring the adequacy of monitoring data to support effective tidal flat topography analysis and decision-making.

[0050] After each acquisition of a new satellite image, the new satellite image is registered with the previously acquired satellite image to ensure spatial alignment of the two satellite images. After registration, the image difference analysis method (such as change detection technology) is used to compare the two adjacent satellite images, focusing on areas with terrain changes;

[0051] The new satellite image is registered with the previously acquired satellite image to ensure that the two satellite images are spatially aligned. The specific steps are as follows:

[0052] Before image registration, the acquired satellite images are preprocessed;

[0053] This includes steps such as geometric correction, radiation correction and atmospheric correction. Geometric correction ensures that the geographic location of the image is accurately aligned with the actual geographic coordinate system to avoid spatial offset caused by satellite attitude changes or sensor errors; radiation correction adjusts the radiation differences when shooting with different sensors to ensure consistent brightness and tone of the image; atmospheric correction eliminates atmospheric interference and ensures the accuracy of image reflectance data. The preprocessed image can provide accurate pixel values ​​and spatial information, providing a reliable data source for subsequent alignment.

[0054] Select the registration control points. These control points are geographical features in the image that can be accurately matched in both images, such as rivers, roads, buildings and other stable and easily identifiable objects. When selecting control points, they should be widely distributed and representative. It is best to select multiple points located at the edge and center of the image to achieve global alignment. These control points are extracted through manual recognition or automated algorithms to ensure that the positions of the control points in the two images are highly consistent, providing basic data for the registration algorithm. Selecting accurate control points is crucial because it directly affects the accuracy and reliability of the registration results.

[0055] After the control points are selected, the registration algorithm is used to align the images. Common methods include algorithms based on feature point matching (such as SURF, SIFT) or global optimization algorithms based on grayscale values. The algorithm adjusts the displacement, rotation or scaling of one of the satellite images by calculating the geometric transformation relationship between the corresponding control points in the two images until the two satellite images completely overlap at the control points. In the process of applying the algorithm, error analysis may be required to detect the residuals between the registered images and adjust the algorithm parameters to further reduce the registration error. Through refined optimization steps, ensure that the accuracy of the final registration result meets the requirements for terrain analysis, especially in high-resolution satellite images, where registration accuracy is critical.

[0056] Satellite remote sensing platforms regularly acquire images of tidal flats twice or more, which can reflect the topographic features of the tidal flat surface, such as land deposition, scouring, and plant coverage. In order to understand the dynamic changes of the tidal flats, image difference analysis methods (such as change detection technology) are applied to compare two adjacent images. Change detection technology is a technology that identifies surface changes by analyzing the differences between two or more images. It is usually used to monitor environmental changes, urban expansion, agricultural land changes, and other fields.

[0057] Specifically, the change detection method can compare the same areas of two images through different algorithms (such as pixel-level difference method, method based on change vector analysis, post-classification change detection method, etc.), calculate the amount of change between the images, and identify the parts that have changed in the tidal flat area. These changes can be caused by tidal effects, storm surges, sediment transport, or terrain changes caused by human activities. For example, in the tidal flat area, the rise and fall of tides may cause changes in the coastline, or sudden heavy rains and storm surges may cause drastic changes in the terrain. Through change detection technology, these change areas can be accurately identified, and then the nature, scope and cause of the changes can be analyzed, providing data support for subsequent dynamic simulation and decision-making.

[0058] This image difference analysis method is crucial for dynamic monitoring of tidal flat terrain, because the terrain changes in tidal flat areas are usually slow and periodic, but can change dramatically under certain conditions (such as strong storms, tidal changes, etc.). By comparing adjacent images, these changes can be discovered in a timely manner, ensuring the timeliness and accuracy of monitoring data, and providing a scientific basis for ecological protection, resource management, and disaster warning.

[0059] This step is crucial for image comparison and analysis, especially in areas with large terrain changes such as tidal flats. Image registration can accurately align the changed areas of the features. After registration, image difference analysis methods (such as change detection technology) are used to compare the two adjacent images, focusing on the areas of terrain changes. This method can identify changes in the tidal flat area, such as land deposition or scouring, and determine whether these changes are in line with the expected terrain evolution trend.

[0060] After completing the image comparison, the key features of the tidal flat terrain changes are extracted. After analyzing the extracted key features, the analyzed features are passed as input data to the pre-trained convolutional neural network, and the convolutional neural network is used to analyze and predict the change trend of the tidal flat terrain;

[0061] After completing the image comparison, the key features of the tidal flat topography changes are extracted, wherein the extracted key features include the fluctuation of the tidal flat surface reflectance (especially the multispectral data obtained through remote sensing images) between two adjacent image acquisitions and the change of plant coverage in the tidal flat ecosystem. After analyzing the fluctuation of the tidal flat surface reflectance between two adjacent image acquisitions and the change of plant coverage in the tidal flat ecosystem, the reflectance fluctuation quantification value and the plant coverage decrease quantification value are generated respectively. The reflectance fluctuation quantification value is used to quantify the change amplitude of the tidal flat surface reflectance, reflecting the distribution change of the tidal flat surface material (such as mud and sediment), thereby revealing the dynamic change of the tidal flat topography and the stability of the ecological environment. The plant coverage decrease quantification value is used to quantify the degree of reduction of plant coverage in the tidal flat ecosystem, reflecting the deterioration of the health of the tidal flat vegetation, thereby revealing the extent to which the tidal flat ecosystem may be damaged.

[0062] After obtaining the reflectivity fluctuation quantification values ​​and plant coverage decrease quantification values ​​generated by analyzing the changes in tidal flat terrain, the analyzed reflectivity fluctuation quantification values ​​and plant coverage decrease quantification values ​​are passed as input data to a pre-trained convolutional neural network, and a terrain evolution index is generated based on the convolutional neural network. The changing trend of the tidal flat terrain is predicted by the terrain evolution index.

[0063] Convolutional neural network (CNN) is a type of convolutional neural network that is widely used in image processing, visual recognition and computer vision. Its core idea is to effectively extract the spatial features of images through local receptive fields, shared weights and pooling operations. The structure of CNN usually includes multiple convolutional layers, pooling layers and fully connected layers. When processing remote sensing images, the convolutional layer can extract local features in the image through convolution kernels (filters), the pooling layer further performs feature dimensionality reduction, and the final fully connected layer makes classification or regression decisions based on the pooled features. The advantage of CNN is that it can automatically learn features from the original image without the need for manually designed feature extraction methods, which makes CNN perform well when processing high-dimensional and complex image data.

[0064] For the analysis of tidal flat terrain evolution, convolutional neural networks can be used as a powerful tool to automatically extract spatial features from remote sensing images. By inputting satellite images taken at different time points, quantified values ​​of reflectivity fluctuations, quantified values ​​of plant coverage decline, etc. into CNN, the model can identify the terrain change pattern of the tidal flat area and predict future terrain evolution. Specifically, CNN is able to identify subtle changes in tidal flat terrain (such as sediment deposition, scouring, or changes in vegetation coverage) from images through convolutional layers, and gradually learn more complex terrain evolution laws in deep networks. Through training, the network can understand the relationship between different tidal flat terrain features, thereby making accurate predictions of terrain change trends.

[0065] In the prediction of tidal flat terrain evolution, the role of the pre-trained convolutional neural network is to extract and learn the features of the input remote sensing data (such as multispectral data of satellite images, quantitative values ​​of reflectivity fluctuations, quantitative values ​​of plant coverage decline, etc.) in order to generate a terrain evolution index that reflects the evolution trend of tidal flat terrain. The key task of the convolutional neural network in this process is to automatically extract valuable spatial features from the input data and perform step-by-step reasoning and prediction through a multi-layer network structure.

[0066] The training of convolutional neural networks usually relies on a large amount of labeled data, which contains information on the historical changes of tidal flat topography. By learning from these labeled data, CNN can automatically identify the patterns of topographic changes in the tidal flat area and make rapid predictions based on existing patterns when new monitoring data arrives. For example, the input quantified values ​​of reflectivity fluctuations and plant coverage decline can be identified by CNN as important signals of tidal flat topography changes. The network will learn the physical or ecological change trends hidden behind these changes, such as tidal flat sedimentation, erosion, and vegetation loss. Through the training and optimization of convolutional neural networks, the system can effectively predict the future evolution of tidal flat topography. The prediction results can help relevant departments to promptly detect abnormal changes in tidal flat ecosystems and take corresponding protective measures.

[0067] Large fluctuations in the reflectivity of the tidal flat surface usually indicate that the tidal flat topography is evolving rapidly and may be accompanied by serious damage to the ecological environment. Reflectivity fluctuations reflect changes in the distribution and characteristics of materials on the tidal flat surface (such as silt, sediment, vegetation, etc.). Large fluctuations may be caused by strong natural or human interference, such as heavy rains, storm surges, flood erosion, rapid migration of sediments, or large-scale reclamation. These changes will lead to instability in the physical structure of the tidal flat, reduction or disappearance of vegetation cover, and even affect its ecological functions (such as water purification, biological habitat provision, etc.). Therefore, monitoring the fluctuation amplitude of reflectivity can not only reveal the rapid evolution of tidal flat topography, but also warn of possible ecological crises, providing a key basis for tidal flat protection and disaster response.

[0068] The specific steps of analyzing the fluctuation of the reflectivity of the tidal flat surface between two adjacent image acquisitions and generating the reflectivity fluctuation quantification value are as follows:

[0069] The spectral data difference analysis of two adjacent remote sensing images of the tidal flat area is performed to extract the local reflectivity gradient change value of each pixel. The calculation expression of the local reflectivity gradient change value is: ,in, Pixel The local reflectivity gradient change value, i and j represent the row index and column index of the pixel in the remote sensing image, respectively. Indicates that in the image at time t, the pixel The reflectivity value in band b, Indicates that in the image at time t+1, the pixel The reflectance value in band b, B is the total number of bands of the multispectral image (such as red, green, blue, near infrared, etc.), and t and t+1 represent the time of two remote sensing image acquisitions;

[0070] This step can capture the reflectance variation gradient of each pixel in multiple spectral dimensions, providing a preliminary quantification of the material changes on the tidal flat surface.

[0071] For the gradient results of reflectivity changes, we further analyze their spatial change patterns and assign weights to each pixel to reflect the contribution of local changes to the overall area. The calculation formula of the weight matrix is: ,in, Represents pixel The spatial variation weight of Represents the Euclidean distance from the pixel to the center of the tidal flat area. , are the coordinates of the region center, is the spatial weight parameter, which controls the influence of pixels far from the center. and Represents pixel points The horizontal and vertical coordinates of

[0072] This step calculates the weight matrix to weight the local reflectivity change gradient of the tidal flat area according to its spatial position, giving higher importance to the core area.

[0073] The local reflectivity gradient change value and spatial variation weight of each pixel are combined to calculate the reflectivity fluctuation quantization value of the entire tidal flat area. The calculation expression is: ,in, is the reflectivity fluctuation quantification value of the tidal flat area, M and N are the number of pixel rows and columns of the tidal flat area image, respectively. is the maximum reflectance change of the pixel point in each spectral band, reflecting the extreme change effect in spectral change. ;

[0074] Through this step, the results of spatial weighting and spectral gradient changes are combined to generate a reflectance fluctuation quantification value. The reflectance fluctuation quantification value not only captures the variation amplitude of the tidal flat surface reflectance, but also reflects the impact of spatial variation patterns and extreme spectral changes on the evolution of tidal flat topography.

[0075] It can be seen from the reflectivity fluctuation quantification value that the larger the performance value of the reflectivity fluctuation quantification value generated after analyzing the fluctuation of the tidal flat surface reflectivity between two adjacent image acquisitions, the greater the change in the tidal flat surface reflectivity between two adjacent image acquisitions, which usually means that the tidal flat terrain is undergoing rapid evolution. Large changes in reflectivity fluctuation values ​​may be caused by factors such as rapid sediment migration, silt scouring, reduced vegetation cover, or human intervention. These changes are often a sign of instability in the tidal flat ecosystem. Large reflectivity fluctuation quantification values ​​further indicate that the tidal flat ecological environment may have a high risk of damage, such as continued degradation of geomorphic structure, loss of biological habitats, or weakening of water quality filtration function. Therefore, by monitoring the changes in reflectivity fluctuation quantification values, it is possible to effectively identify areas of rapid evolution of tidal flat terrain, thereby providing early warning for ecological protection and emergency response.

[0076] The rapid decline rate of plant cover in tidal flat ecosystems usually indicates that the tidal flat topography is undergoing rapid evolution and may reflect that the tidal flat ecological environment is being severely damaged. Plant cover is an important indicator of the health of tidal flat ecosystems. When vegetation cover declines rapidly, it may mean that the tidal flat surface has changed significantly due to strong sedimentation, scouring, erosion or pollution. This change may be caused by natural factors (such as storm surges, floods, and sea level rise) or human activities (such as reclamation, mining, and engineering construction). The reduction of vegetation will weaken the ecological functions of the tidal flat, such as flood control, sand fixation, and pollution filtration, and will also lead to problems such as habitat loss and biodiversity decline. Therefore, the rapid decline in vegetation cover is an important warning of the deterioration of the health of the tidal flat ecosystem, which requires timely monitoring and the adoption of protection and restoration measures.

[0077] The specific steps for analyzing the changes in plant coverage in the tidal flat ecosystem between two consecutive image acquisitions to generate a quantitative value of plant coverage decrease are as follows:

[0078] The normalized vegetation index (NDVI) is calculated by using the near infrared band (NIR) and red light band (R) in remote sensing images to quantify the plant coverage of each pixel. The normalized vegetation index (NDVI) can effectively reflect the growth status and density of vegetation. As the basic characteristic of tidal flat plant coverage, the calculation result will form a pixel-level plant coverage distribution map, providing basic data for subsequent analysis. The calculation expression of the normalized vegetation index is: , where Indicates the coordinate point corresponding to time t The Normalized Difference Vegetation Index (NDVI) on Indicates the coordinate point corresponding to time t The near-infrared reflectance on the surface reflects the photosynthetic activity of vegetation. Indicates the coordinate point corresponding to time t The red reflectance on the surface reflects the amount of red light absorbed by the vegetation;

[0079] By calculating the time change rate of the normalized difference vegetation index (NDVI), the change in plant coverage between two consecutive image acquisitions is extracted, the dynamic change of plant coverage during the entire monitoring period is quantified, and short-term drastic changes or long-term stable changes are captured. The calculation expression of the change in plant coverage is: , where Indicates the coordinate point The change in plant coverage reflects the dynamic trend. represents the instantaneous rate of change of plant coverage, which is calculated by the derivative of time t, and They respectively represent the start time and end time of the interval between two adjacent image acquisitions;

[0080] By taking a weighted average of the changes in all pixels in the tidal flat area, and combining the ecological sensitivity of each pixel (such as the priority of the ecological core area) and the environmental exposure factors (such as tidal frequency and salinity level), a quantitative value of plant coverage decrease is generated. The quantitative value of plant coverage decrease reflects the overall reduction in tidal flat vegetation coverage. The calculation expression of the quantitative value of plant coverage decrease is: , where is a quantitative value for the decline in plant cover, quantifying the overall severity of the reduction in tidal flat vegetation. is a weight factor used to measure the ecological sensitivity of each pixel. It represents the regional impact factor, which represents the environmental exposure degree of the pixel point, and R is the set of all pixels in the tidal flat area;

[0081] The regional impact factor is a parameter used to quantify the exposure of each pixel in the tidal flat area to specific environmental conditions, reflecting the environmental pressure or impact of the area. It comprehensively considers environmental factors closely related to the topography and ecological changes of the tidal flat, such as tidal frequency, salinity level, storm surge impact, hydrodynamic conditions, etc. These factors will directly or indirectly affect the growth of tidal flat plants and changes in vegetation coverage.

[0082] Regional impact factors can be determined through multi-source data fusion and weighted calculation. First, relevant environmental parameters (such as tidal monitoring records, salinity distribution, hydrodynamic model output, etc.) are extracted from remote sensing images, ground observation data, and meteorological data. Then, weights are assigned according to the intensity of the impact of each environmental parameter on the tidal flat topography or ecosystem, and a comprehensive regional impact factor is calculated using linear or nonlinear combinations. For example, for areas with frequent tides, their exposure is higher, so a greater weight is given in the calculation; while for areas with low salinity or stable hydrodynamics, their exposure is lower, and the weight of the regional impact factor is relatively small. Finally, the regional impact factor of each pixel is determined by the comprehensive calculation of these parameters, providing an accurate measure of environmental exposure.

[0083] It can be seen from the quantitative value of plant coverage decline that the larger the performance value of the quantitative value of plant coverage decline generated by analyzing the changes in plant coverage in the tidal flat ecosystem between two adjacent image acquisitions, the more significant the reduction in vegetation coverage in the tidal flat area between two adjacent image acquisitions. This phenomenon reflects the rapid evolution of the tidal flat topography, which may be caused by severe sedimentation, scouring, erosion, pollution or human interference. The rapid decline in vegetation coverage weakens the ecological functions of the tidal flat, such as sand fixation, flood control, and water purification, and may also lead to ecological problems such as habitat loss and biodiversity reduction. Therefore, the larger the performance value of the quantitative value of plant coverage decline, not only indicates that the severity of the tidal flat topography change has increased, but also warns that there is a greater risk of damage to the tidal flat ecological environment, and timely monitoring and protection measures are needed.

[0084] The convolutional neural network is not specifically limited here, and can quantify the reflectivity fluctuation of the tidal flat area. and quantification of plant cover decline Comprehensive analysis to generate terrain evolution index In order to realize the technical solution of the present invention, the present invention provides a specific implementation method;

[0085] Terrain Evolution Index The generated expression is:

[0086]

[0087] , where , are the quantitative values ​​of reflectivity fluctuation in the tidal flat area. and quantification of plant cover decline The preset scaling factor of , Both are greater than 0.

[0088] It can be seen from the terrain evolution index that the greater the performance value of the reflectivity fluctuation quantification value generated by analyzing the fluctuation of the reflectivity of the tidal flat surface between two adjacent image acquisitions, the greater the performance value of the plant coverage decrease quantification value generated by analyzing the change of plant coverage in the tidal flat ecosystem between two adjacent image acquisitions, that is, the greater the performance value of the terrain evolution index generated by analyzing and predicting the change trend of the tidal flat terrain through the convolutional neural network, the faster the rate of tidal flat terrain evolution. This change usually indicates that there may be a greater risk of damage to the tidal flat ecological environment; on the contrary, if the performance value of the terrain evolution index is small, it means that the tidal flat terrain changes slowly, the ecosystem is more stable, and the risk of ecological environment damage is small.

[0089] Through the prediction results of convolutional neural network, the terrain changes in the tidal flat area are divided into two categories: fast evolution process and slow evolution process;

[0090] The terrain evolution index generated by analyzing and predicting the change trend of the tidal flat terrain through the convolutional neural network is compared with the pre-set terrain evolution index reference threshold to divide the terrain change situation of the tidal flat area. The specific division steps are as follows:

[0091] If the terrain evolution index generated when analyzing and predicting the changing trend of the tidal flat terrain is greater than the pre-set terrain evolution index reference threshold, the terrain change in the tidal flat area is classified as a rapid evolution process;

[0092] Rapid evolution processes refer to areas where the tidal flat topography changes dramatically, such as changes caused by natural factors such as sudden storm surges and floods caused by heavy rains, or drastic changes in topography caused by human activities (such as reclamation and mining).

[0093] If the terrain evolution index generated when analyzing and predicting the changing trend of the tidal flat terrain is less than or equal to the pre-set terrain evolution index reference threshold, the terrain change of the tidal flat area is classified as a slow evolution process;

[0094] The slow evolution process refers to areas where changes are relatively gentle and stable, such as long-term natural sedimentation and tidal influences.

[0095] Through classification, corresponding monitoring strategies can be formulated according to different types of evolution trends, thereby effectively responding to the different needs of tidal flat topography changes.

[0096] For tidal flat areas predicted to be undergoing a slow evolution process, the initially set expected monitoring cycle should be maintained to conduct regular monitoring of the tidal flat areas;

[0097] Areas with slow evolution processes usually do not change much, and extending the monitoring period can reduce resource waste and still maintain long-term tracking of topographic changes. For example, tidal influences on mudflats and seasonal sediment changes are relatively slow processes. Therefore, continuing to monitor at a predetermined period (such as every month or every two months) is sufficient to meet the needs of tracking changes in such areas and ensure the continuity and stability of data.

[0098] For tidal flat areas predicted to be in a rapid evolution process, the expected monitoring period is shortened based on the predicted results of the terrain evolution of the tidal flat areas, so as to respond to the rapid changes in the tidal flat areas in a timely manner;

[0099] The purpose of this step is to dynamically adjust the expected monitoring cycle according to the predicted results of the tidal flat terrain evolution in order to respond more quickly to the rapid changes in the tidal flat area. When the model predicts that a tidal flat area is in a rapid evolution process, it means that the terrain in the area is changing rapidly, which may involve sudden ecological risks or environmental disasters such as storm surges, erosion or sedimentation. By shortening the monitoring cycle, more frequent and timely data collection can be achieved to ensure that these changes are captured in a timely manner, thereby providing more efficient and accurate monitoring methods for ecological protection, resource management and disaster warning in tidal flat areas. In short, this step helps to improve the response speed of monitoring, avoid the failure to detect and deal with drastic changes in tidal flat areas in a timely manner, and reduce potential ecological and environmental risks.

[0100] For tidal flat areas predicted to be in a rapid evolution process, the expected monitoring period is shortened based on the predicted results of the terrain evolution of the tidal flat area. The specific steps are as follows:

[0101] For the case where the tidal flat area is predicted to be a rapid evolution process, the terrain evolution difference between the terrain evolution index and the pre-set terrain evolution index reference threshold is first calculated, and the calculation expression is:

[0102] , is the topographic evolution index of the current tidal flat area, which is predicted by the convolutional neural network model. It is a pre-set reference threshold of terrain evolution index, which is used to distinguish the rate of change of tidal flat terrain. It is the difference in terrain evolution, representing the degree of terrain change;

[0103] When the terrain evolves differently When the value of is large, it means that the change speed of the tidal flat topography is much higher than the normal change, and it is predicted to be a rapid evolution process. According to this difference in topographic evolution, the next step will be to adjust the monitoring cycle.

[0104] After obtaining the terrain evolution difference, the expected monitoring period of the tidal flat area is adjusted according to the terrain evolution difference. In order to respond to the rapid changes of the tidal flat terrain in time, the monitoring period is dynamically shortened according to the size of the terrain evolution index. The expression for adjusting the expected monitoring period is: , where is the actual monitoring period after adjustment, is the expected monitoring period initially set, It is a scaling factor used to control the shortening degree of the monitoring period. It can be adjusted according to actual needs and is usually a constant less than 1 (such as 0.1-0.5). represents the maximum topographic evolution index observed during past tidal flat monitoring periods, serving as a benchmark to assess the relative rate of change in the current area;

[0105] Through this step, the monitoring period can be dynamically adjusted according to the size of the terrain evolution index of the tidal flat area. The larger the area), the shorter the monitoring period is to ensure timely detection and response to the rapidly changing tidal flat environment.

[0106] The adjusted actual monitoring cycle is applied to the monitoring process of the tidal flat area. The shortened monitoring cycle will achieve more frequent image acquisition through the satellite remote sensing platform, detect the changes in the tidal flat area in a timely manner, and reset the time interval for acquiring images. The specific expression is: , where The next monitoring time is is the time point of the last monitoring;

[0107] The core of this step is to ensure that changes in tidal flat topography can be captured quickly through more frequent image acquisition, and timely response measures (such as issuing alarms, conducting more in-depth analysis, etc.) can be taken to improve the management efficiency of tidal flat ecosystems and the ability to respond to emergencies.

[0108] Through the above-mentioned analysis and dynamic simulation method of the topographic evolution of the Yangtze River Estuary based on multi-source data fusion, the monitoring cycle of the tidal flat area can be dynamically adjusted, thereby improving the flexibility and timeliness of monitoring. This method divides the tidal flat area into a fast evolution process and a slow evolution process according to the actual changes in the topographic evolution. For areas with relatively stable changes, the initial monitoring cycle is maintained to save resources; while for areas with drastic changes, the monitoring cycle is dynamically shortened, significantly improving the data collection frequency and response speed. This dynamic adjustment strategy effectively solves the lag problem existing in traditional regular monitoring methods, and can capture the rapid changes of tidal flat topography (such as sudden sedimentation, scouring or ecological risks) in a timely manner, thereby ensuring that key changes can be quickly identified, and providing accurate and real-time decision support for the protection of the tidal flat ecological environment and resource management. This flexible monitoring method not only optimizes the monitoring efficiency, but also significantly reduces the risk of ecological and environmental damage, and realizes efficient and targeted monitoring resource allocation.

[0109] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0110] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0111] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0112] In the several embodiments provided in the present application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0113] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0114] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0115] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0116] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for analyzing and dynamically simulating the topographic evolution of the Yangtze River Estuary based on multi-source data fusion, characterized in that: The following steps are involved: According to the monitoring needs of the tidal flat area, the satellite remote sensing platform sets a preliminary expected monitoring cycle to regularly obtain image data of the tidal flat area to achieve real-time monitoring of the tidal flat area; After each acquisition of a new satellite image, the new satellite image is registered with the previously acquired satellite image to ensure spatial alignment of the two satellite images. After registration, the image difference analysis method is used to compare the two adjacent satellite images, focusing on areas with terrain changes; After completing the image comparison, the key features of the tidal flat terrain changes are extracted. After analyzing the extracted key features, the analyzed features are passed as input data to the pre-trained convolutional neural network, and the convolutional neural network is used to analyze and predict the change trend of the tidal flat terrain; Through the prediction results of convolutional neural network, the terrain changes in the tidal flat area are divided into two categories: fast evolution process and slow evolution process; For tidal flat areas predicted to be undergoing a slow evolution process, the initially set expected monitoring cycle should be maintained to conduct regular monitoring of the tidal flat areas; For tidal flat areas predicted to be in a rapid evolution process, the expected monitoring period is shortened based on the predicted results of the terrain evolution of the tidal flat areas, so as to respond to the rapid changes in the tidal flat areas in a timely manner; After completing the image comparison, the key features of the tidal flat terrain changes are extracted, wherein the extracted key features include the fluctuation of the tidal flat surface reflectance between two adjacent image acquisitions and the change of the plant coverage in the tidal flat ecosystem. After analyzing the fluctuation of the tidal flat surface reflectance between two adjacent image acquisitions and the change of the plant coverage in the tidal flat ecosystem, a reflectance fluctuation quantification value and a plant coverage reduction quantification value are generated respectively. The reflectance fluctuation quantification value is used to quantify the change amplitude of the tidal flat surface reflectance, reflecting the distribution change of the tidal flat surface material. The plant coverage reduction quantification value is used to quantify the reduction degree of the plant coverage in the tidal flat ecosystem, reflecting the deterioration of the health status of the tidal flat vegetation. After obtaining the reflectivity fluctuation quantification value and the plant coverage decrease quantification value generated by analyzing the tidal flat terrain changes, the analyzed reflectivity fluctuation quantification value and the plant coverage decrease quantification value are passed as input data to the pre-trained convolutional neural network, and the terrain evolution index is generated based on the convolutional neural network. The change trend of the tidal flat terrain is predicted through the terrain evolution index; The terrain evolution index generated by analyzing and predicting the change trend of the tidal flat terrain through the convolutional neural network is compared with the pre-set terrain evolution index reference threshold to divide the terrain change situation of the tidal flat area. The specific division steps are as follows: If the terrain evolution index generated when analyzing and predicting the changing trend of the tidal flat terrain is greater than the pre-set terrain evolution index reference threshold, the terrain change in the tidal flat area is classified as a rapid evolution process; If the terrain evolution index generated when analyzing and predicting the changing trend of the tidal flat terrain is less than or equal to the pre-set terrain evolution index reference threshold, the terrain change of the tidal flat area is classified as a slow evolution process; For tidal flat areas predicted to be in a rapid evolution process, the expected monitoring period is shortened based on the predicted results of the terrain evolution of the tidal flat area. The specific steps are as follows: For the case where the tidal flat area is predicted to be a rapid evolution process, the terrain evolution difference between the terrain evolution index and the pre-set terrain evolution index reference threshold is first calculated, and the calculation expression is: , is the topographic evolution index of the current tidal flat area, which is predicted by the convolutional neural network model. It is a pre-set reference threshold of terrain evolution index, which is used to distinguish the rate of change of tidal flat terrain. It is the difference in terrain evolution, representing the degree of terrain change; After obtaining the terrain evolution difference, the expected monitoring period of the tidal flat area is adjusted according to the terrain evolution difference. In order to respond to the rapid changes of the tidal flat terrain in time, the monitoring period is dynamically shortened according to the size of the terrain evolution index. The expression for adjusting the expected monitoring period is: , where is the actual monitoring period after adjustment, is the expected monitoring period initially set, is a scaling factor that controls the degree of shortening of the monitoring period. represents the maximum topographic evolution index observed during past tidal flat monitoring periods, serving as a benchmark to assess the relative rate of change in the current area; The adjusted actual monitoring cycle is applied to the monitoring process of the tidal flat area. The shortened monitoring cycle will achieve more frequent image acquisition through the satellite remote sensing platform, detect the changes in the tidal flat area in a timely manner, and reset the time interval for acquiring images. The specific expression is: , where The next monitoring time is The time point of the last monitoring.

2. The method for analyzing and dynamically simulating the topographic evolution of the Yangtze River Estuary based on multi-source data fusion according to claim 1 is characterized in that: The new satellite image is registered with the previously acquired satellite image to ensure spatial alignment of the two images. The specific steps are as follows: Before image registration, the acquired satellite images are preprocessed; Points that accurately correspond to geographic features in the two satellite images are selected as registration control points. In order to ensure that the registration control points are widely distributed and representative, multiple points located at the edge and center of the image are selected to achieve global alignment; After the control points are selected, the registration algorithm is used to align the images. Specifically, the geometric transformation relationship between the corresponding control points in the two satellite images is calculated, and the displacement, rotation or scaling of one of the satellite images is adjusted until the two satellite images completely overlap at the control points.

3. The method for analyzing and dynamically simulating the topographic evolution of the Yangtze River Estuary based on multi-source data fusion according to claim 1 is characterized in that: The specific steps of analyzing the fluctuation of the reflectivity of the tidal flat surface between two adjacent image acquisitions and generating the reflectivity fluctuation quantification value are as follows: The spectral data difference analysis of two adjacent remote sensing images of the tidal flat area is performed to extract the local reflectivity gradient change value of each pixel. The calculation expression of the local reflectivity gradient change value is: ,in, Pixel The local reflectivity gradient change value, i and j Respectively represent the row index and column index of the pixel in the remote sensing image, Indicates that in the image at time t, the pixel In the band b The reflectivity value, Indicates that in the image at time t+1, the pixel In the band b The reflectivity value, B is the total number of bands of the multispectral image, t and t+1 represent the time of two remote sensing image acquisitions; For the gradient results of reflectivity changes, we further analyze their spatial change patterns and assign weights to each pixel to reflect the contribution of local changes to the overall area. The calculation formula of the weight matrix is: ,in, Represents pixel The spatial variation weight of Represents the Euclidean distance from the pixel to the center of the tidal flat area. , are the coordinates of the region center, is the spatial weight parameter, which controls the influence of pixels far from the center. and Represents pixel points The horizontal and vertical coordinates of The local reflectivity gradient change value and spatial variation weight of each pixel are combined to calculate the reflectivity fluctuation quantization value of the entire tidal flat area. The calculation expression is: ,in, is the quantitative value of the reflectivity fluctuation of the tidal flat area, M and N are the number of pixel rows and columns of the tidal flat area image, is the maximum reflectance change of the pixel point in each spectral band, reflecting the extreme change effect in the spectral change. .

4. The method for analyzing and dynamically simulating the topographic evolution of the Yangtze River Estuary based on multi-source data fusion according to claim 1 is characterized in that: The specific steps for analyzing the changes in plant coverage in the tidal flat ecosystem between two consecutive image acquisitions to generate a quantitative value of plant coverage decrease are as follows: The normalized vegetation index is calculated by the near-infrared band and the red light band in the remote sensing image to quantify the plant coverage of each pixel. The normalized vegetation index can effectively reflect the growth status and density of vegetation. As the characteristic of tidal flat plant coverage, the calculation result will form a pixel-level plant coverage distribution map, providing basic data for subsequent analysis. The calculation expression of the normalized vegetation index is: , where Indicates the coordinate point corresponding to time t The normalized difference vegetation index on Indicates the coordinate point corresponding to time t The near-infrared reflectance on the surface reflects the photosynthetic activity of vegetation. Indicates the coordinate point corresponding to time t The red reflectance on the surface reflects the amount of red light absorbed by the vegetation; By calculating the time change rate of the normalized vegetation index, the change in plant coverage between two consecutive image acquisitions is extracted, the dynamic change of plant coverage during the entire monitoring period is quantified, and short-term drastic changes or long-term stable changes are captured. The calculation expression of the change in plant coverage is: , where Indicates the coordinate point The change in plant coverage reflects the dynamic trend. represents the instantaneous rate of change of plant coverage, which is calculated by the derivative of time t, and They respectively represent the start time and end time of the interval between two adjacent image acquisitions; By taking a weighted average of the changes in all pixels in the tidal flat area and combining the ecological sensitivity and environmental exposure factors of each pixel, a quantitative value of plant coverage decrease is generated. The quantitative value of plant coverage decrease reflects the overall reduction in tidal flat vegetation coverage. The calculation expression of the quantitative value of plant coverage decrease is: , where is a quantitative value for the decline in plant cover, quantifying the overall severity of the reduction in tidal flat vegetation. is a weight factor used to measure the ecological sensitivity of each pixel. Represents the regional impact factor, representing the degree of environmental exposure of the pixel point. R It is the set of all pixels in the tidal flat area.

Images