Mangrove forest growth monitoring method and mangrove forest growth monitoring system

Abstract

The invention provides a mangrove forest growth monitoring method and system, and the method comprises the steps: obtaining a mangrove forest multi-temporal remote sensing image of a target monitoring region, extracting a spectral feature index sum of the mangrove forest multi-temporal remote sensing image, and obtaining a leaf area index based on a leaf area index model; obtaining a water body suspended sediment concentration inversion result based on a sediment content inversion model; performing spatial statistical analysis on the leaf area index and the inversion result of the suspended sediment concentration in the water body to obtain a spatial statistical index; according to the leaf area index and the water body suspended sediment concentration inversion result, the stress degree of water body suspended sediment on mangrove forest growth is analyzed, and mangrove forest growth condition change parameters are obtained; and according to the spatial statistical index and the mangrove forest growth condition change parameter, performing stress degree grading on the target monitoring area, and generating a corresponding early warning instruction. The problem that the influence of suspended sediment in the water body on mangrove forest growth cannot be recognized in time can be solved, and the mangrove forest growth monitoring accuracy is improved.

Description

Technical Field

[0001] The present application relates to the field of geographic information technology, and in particular, to a method and system for monitoring the growth of mangroves. Background Art

[0002] Mangroves are the core barrier of coastal ecosystem, with key functions such as wave attenuation and shore protection, carbon sequestration and storage, and water purification. However, affected by the superposition of human activities (such as reclamation and port construction) and natural factors (such as sea-level rise and extreme climate), mangroves are often faced with water and sediment environment stresses such as sediment deposition in water body and rising water turbidity, resulting in growth retardation or even degradation. Therefore, real-time monitoring of the growth status of mangroves and early warning of stress risks are the core requirements for coastal ecological protection and restoration.

[0003] At present, conventional methods mainly rely on remote sensing technology and static threshold segmentation of vegetation indices (such as NDVI) to classify and monitor mangroves. However, the impact of periodic tidal inundation on the growth of mangroves is not considered. During the tidal inundation process, suspended sediment in the water body will rapidly deposit in the mangrove area. When the suspended sediment in the water body stays at a high concentration for a long time, it will cause certain harm to the growth of mangroves. Due to the lag of suspended sediment in the water body on the growth of mangroves, for example, when the root system of mangroves lacks oxygen due to the accumulation of suspended sediment in the water body and the canopy has not shown symptoms at the initial stage, this hidden stress cannot be recognized in time when relying on vegetation indices to monitor the growth of mangroves, resulting in the problem of low monitoring accuracy of mangroves. Summary of the Invention

[0004] The present application aims at the problems in the prior art, and provides a method and system for monitoring the growth of mangroves, which can effectively improve the monitoring accuracy of mangroves by analyzing the stress degree of suspended sediment in the water body on the growth of mangroves.

[0005] A method for monitoring the growth of mangroves includes:

[0006] Obtaining multi-temporal remote sensing images of a target monitoring area, and extracting spectral feature indices of the multi-temporal remote sensing images, wherein the target monitoring area includes a mangrove area and a non-mangrove area;

[0007] Classifying the multi-temporal remote sensing images based on a preset mangrove classification model according to the spectral feature indices to generate a multi-temporal remote sensing image of mangroves only including the mangrove area;

[0008] Analyzing the canopy coverage degree of mangroves based on a preset leaf area index model according to the spectral feature indices of the multi-temporal remote sensing image of mangroves to obtain a leaf area index;

[0009] Extract the sediment-sensitive reflectance in the multi-temporal remote sensing images of the mangroves, and based on a preset sediment content inversion model, invert the water suspended sediment content in the water body to obtain the inversion result of the water suspended sediment concentration. Among them, the sediment-sensitive reflectance is the remote sensing reflectance of the band in the multi-temporal remote sensing images of the mangroves that is sensitive to the water suspended sediment;

[0010] Conduct spatial statistical analysis on the leaf area index and the inversion result of the water suspended sediment concentration to obtain a spatial statistical index;

[0011] According to the leaf area index and the inversion result of the water suspended sediment concentration, based on a preset mangrove-water suspended sediment stress coupling model, analyze the stress degree of the water suspended sediment on the growth of the mangroves to obtain the change parameters of the mangrove growth status;

[0012] According to the spatial statistical index and the change parameters of the mangrove growth status, based on a preset stress level division rule, divide the stress degree level of the target monitoring area;

[0013] According to the stress degree level division result of the target monitoring area, generate a warning instruction corresponding to the stress degree level, and transmit the warning instruction to the mangrove growth warning device.

[0014] This application also provides a mangrove growth monitoring system, including:

[0015] Multi-temporal remote sensing image acquisition module: used to acquire the multi-temporal remote sensing images of the target monitoring area and extract the spectral feature index of the multi-temporal remote sensing images. Among them, the target monitoring area includes the mangrove area and the non-mangrove area;

[0016] Mangrove classification module: used to classify the multi-temporal remote sensing images based on a preset mangrove classification model according to the spectral feature index to generate the multi-temporal remote sensing images of the mangroves that only include the mangrove area;

[0017] Leaf area index calculation module: used to analyze the canopy coverage degree of the mangroves based on a preset leaf area index model according to the spectral feature index of the multi-temporal remote sensing images of the mangroves to obtain the leaf area index;

[0018] Water suspended sediment concentration inversion module: used to extract the sediment-sensitive reflectance in the multi-temporal remote sensing images of the mangroves, and based on a preset sediment content inversion model, invert the water suspended sediment content in the water body to obtain the inversion result of the water suspended sediment concentration. Among them, the sediment-sensitive reflectance is the remote sensing reflectance of the band in the multi-temporal remote sensing images of the mangroves that is sensitive to the water suspended sediment;

[0019] Spatial statistical index calculation module: used to perform spatial statistical analysis on the inversion results of the leaf area index and the water body suspended sediment concentration, and obtain the spatial statistical index;

[0020] Mangrove growth status change parameter calculation module: used to analyze the stress degree of water body suspended sediment on mangrove growth based on the preset mangrove-water body suspended sediment stress coupling model according to the inversion results of the leaf area index and the water body suspended sediment concentration, and obtain the mangrove growth status change parameter;

[0021] Stress degree level division module: used to divide the stress degree level of the target monitoring area based on the preset stress level division rule according to the spatial statistical index and the mangrove growth status change parameter;

[0022] Early warning instruction transmission module: used to generate an early warning instruction corresponding to the stress degree level according to the stress degree level division result of the target monitoring area, and transmit the early warning instruction to the mangrove growth early warning device.

[0023] Compared with the prior art, the present application uses multi-temporal remote sensing images and analyzes the canopy coverage degree of mangroves by introducing a leaf area index model to obtain the leaf area index as a direct index of the mangrove growth state. At the same time, based on the sediment-sensitive reflectance and sediment content inversion model, the water body suspended sediment content in the water body is accurately inverted. By combining the spatial statistical analysis of the leaf area index and the water body suspended sediment content, the mangrove area affected by suspended sediment is determined, and based on the preset mangrove-water body suspended sediment stress coupling model, the stress degree of water body suspended sediment on mangrove growth is analyzed, realizing the dynamic division of the stress level in the mangrove area, and automatically generating early warning instructions, effectively solving the problem that the hidden stress of water body suspended sediment cannot be recognized in time when relying on vegetation indices for mangrove growth monitoring, and significantly improving the accuracy of mangrove growth monitoring.

[0024] In order to understand the present application more clearly, the specific implementation manners of the present application will be described below in conjunction with the accompanying drawings. Brief description of the drawings

[0025] Figure 1 It is a flowchart of a mangrove growth monitoring method of the present application;

[0026] Figure 2 It is a flowchart of the method for constructing a mangrove classification model in a mangrove growth monitoring method of the present application;

[0027] Figure 3 It is a flowchart of the method for evaluating the accuracy of a mangrove classification model in a mangrove growth monitoring method of the present application;

[0028] Figure 4It is a flowchart of the method for constructing a leaf area index model in a mangrove growth monitoring method of the present application;

[0029] Figure 5 It is a flowchart of the method for constructing a sediment content inversion model in a mangrove growth monitoring method of the present application;

[0030] Figure 6 It is a flowchart of the method for constructing a sediment-sensitive reflectance-optical parameter coupling model in a mangrove growth monitoring method of the present application;

[0031] Figure 7 It is a flowchart of the method for spatially statistical analyzing the inversion results of the leaf area index and the water body suspended sediment concentration in a mangrove growth monitoring method of the present application;

[0032] Figure 8 It is a flowchart of the method for constructing a mangrove-water body suspended sediment stress coupling model in a mangrove growth monitoring method of the present application;

[0033] Figure 9 It is a flowchart of the method for classifying the stress degree level of a target monitoring area in a mangrove growth monitoring method of the present application;

[0034] Figure 10 It is a schematic diagram of a mangrove growth monitoring system of the present application. Specific embodiments

[0035] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present application.

[0036] It should be understood that the schematic drawings are not drawn to scale. The flowcharts used in the present application illustrate the operations implemented according to some embodiments of the present application. It should be understood that the operations in the flowcharts may not be implemented in sequence, and steps without logical context relationships may be reversed or implemented simultaneously. In addition, those skilled in the art can add one or more other operations to the flowchart or remove one or more operations from the flowchart under the guidance of the content of the present application.

[0037] References to "embodiments" in this specification mean that the particular features, structures, or characteristics described in connection with the embodiments can be included in at least one embodiment of the present application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0038] Embodiment 1

[0039] Please refer to Figure 1 , Figure 1 , which is a flowchart of a method for monitoring the growth of mangroves according to the present application. The present application provides a method for monitoring the growth of mangroves, specifically including the following steps:

[0040] S1: Obtain multi-temporal remote sensing images of the target monitoring area, and extract the spectral feature indices of the multi-temporal remote sensing images, where the target monitoring area includes mangrove areas and non-mangrove areas;

[0041] S2: Classify the multi-temporal remote sensing images based on a preset mangrove classification model according to the spectral feature indices to generate multi-temporal remote sensing images of mangroves that only include the mangrove areas;

[0042] S3: Analyze the canopy coverage degree of mangroves based on a preset leaf area index model according to the spectral feature indices of the multi-temporal remote sensing images of mangroves to obtain the leaf area index;

[0043] S4: Extract the sediment-sensitive reflectance in the multi-temporal remote sensing images of mangroves, and invert the water body suspended sediment content in the water body based on a preset sediment content inversion model to obtain the inversion result of the water body suspended sediment concentration, where the sediment-sensitive reflectance is the remote sensing reflectance of the band sensitive to water body suspended sediment in the multi-temporal remote sensing images of mangroves;

[0044] S5: Conduct spatial statistical analysis on the leaf area index and the inversion result of the water body suspended sediment concentration to obtain a spatial statistical index;

[0045] S6: Analyze the stress degree of water body suspended sediment on the growth of mangroves based on a preset mangrove-water body suspended sediment stress coupling model according to the leaf area index and the inversion result of the water body suspended sediment concentration to obtain the change parameters of the growth status of mangroves;

[0046] S7: Classify the stress degree level of the target monitoring area based on a preset stress level division rule according to the spatial statistical index and the change parameters of the growth status of mangroves;

[0047] S8: Generate a warning instruction corresponding to the stress level based on the stress level classification result of the target monitoring area, and transmit the warning instruction to the mangrove growth warning device;

[0048] Compared with the prior art, this solution uses multi-temporal remote sensing images to accurately distinguish mangrove areas from non-mangrove areas, avoiding the influence of vegetation in non-mangrove areas on mangrove monitoring. Secondly, by introducing a leaf area index model to analyze the canopy coverage of mangroves, the leaf area index is obtained as a direct indicator of mangrove growth status. At the same time, based on the sediment-sensitive reflectance and sediment content inversion model, the suspended sediment content in the water body is accurately inverted. By combining spatial statistical analysis of the leaf area index and the suspended sediment content in the water body, the mangrove areas affected by suspended sediment are determined, and based on a preset mangrove-water suspended sediment stress coupling model, the stress degree of water suspended sediment on mangrove growth is analyzed, realizing the dynamic classification of the stress level in the mangrove area and automatically generating a warning instruction, effectively solving the problem that the hidden stress of water suspended sediment cannot be timely identified when relying on vegetation indices for mangrove growth monitoring, and significantly improving the accuracy of mangrove growth monitoring.

[0049] The mangrove growth monitoring method of the present invention can be executed by the following computer system, which includes a mangrove growth monitoring database server, a data acquisition server, and a mangrove growth monitoring server. The mangrove growth monitoring database server is used to store information such as multi-temporal remote sensing images of the target monitoring area, so as to construct the mangrove growth monitoring database.

[0050] The data acquisition server is used to obtain the multi-temporal remote sensing image information of the target monitoring area from the mangrove growth monitoring database server and send it to the mangrove growth monitoring server for processing.

[0051] The mangrove growth monitoring server executes the mangrove growth monitoring method of the present invention, classifies the stress level of the target monitoring area, generates a warning instruction corresponding to the stress level, and transmits it to the mangrove growth warning device.

[0052] For step S1, in this embodiment, the target monitoring area includes a mangrove area and a non-mangrove area. The multi-temporal remote sensing image can be the seasonal images of Landsat satellites or Sentinel-2 satellites. Exemplarily, in March (spring), June (summer), September (autumn), and December (winter) of each year, the time span covers a complete annual cycle.

[0053] Preferentially use multi-temporal remote sensing images of Sentinel-2 satellite with a temporal resolution of 5 days and a spatial resolution of 10m or 20m. The multi-temporal remote sensing images of Sentinel-2 satellite include a blue band Band2 with a central wavelength of 490nm, a green band Band3 with a central wavelength of 560nm, a red band Band4 with a central wavelength of 665nm, a red edge band Band7 with a central wavelength of 783nm, a wide near-infrared band Band8 with a central wavelength of 842nm, a narrow near-infrared band Band8A with a central wavelength of 865nm, and shortwave infrared bands, etc., which can be obtained from the official website of the European Space Agency.

[0054] The spectral feature index is a numerical index calculated from the remote sensing reflectance of specified bands in the multi-temporal remote sensing images, including the normalized difference vegetation index, water index, and mangrove extraction index. The mangrove area refers to the geographical area where mangrove vegetation grows, and the non-mangrove area refers to other geographical areas except the mangrove area, including other types of vegetation, water bodies, construction land, etc.

[0055] Preprocessing such as radiometric calibration, geometric precise correction, and atmospheric correction can be performed on the multi-temporal remote sensing images of the satellite in sequence. Among them, radiometric calibration is to convert the original digital values recorded by the sensor into radiance or apparent reflectance; the purpose of geometric precise correction is to eliminate geometric distortions caused by sensor attitude, platform movement, or terrain undulation to ensure accurate spatial position of the image; atmospheric correction is to remove the influence of atmospheric scattering, absorption, etc. and convert the apparent reflectance into the true surface reflectance.

[0056] For step S2, in this embodiment, the mangrove multi-temporal remote sensing image only including the mangrove area refers to the multi-temporal remote sensing image obtained after labeling the pixels in the mangrove area in the multi-temporal remote sensing image as 1 and the pixels in the non-mangrove area as 0. Please refer to Figure 2 , Figure 2 This is the method flowchart for constructing a mangrove classification model in a mangrove growth monitoring method of this application. Constructing the mangrove classification model includes:

[0057] S21: Obtain historical multi-temporal remote sensing images of several target monitoring areas. Among them, the historical mangrove multi-temporal remote sensing images include: green band, shortwave infrared band, red edge band, wide near-infrared band, and narrow near-infrared band;

[0058] S22: Label the mangrove area and non-mangrove area in the historical multi-temporal remote sensing images to obtain a standard historical multi-temporal remote sensing image;

[0059] S23: Calculate the water body index and the mangrove extraction index based on the remote sensing reflectance of the green light band, short-wave infrared band, red edge band, wide near-infrared band, and narrow near-infrared band, using a preset water body index calculation formula and a mangrove extraction index calculation formula;

[0060] S24: Use the normalized vegetation index, water body index, and mangrove extraction index as input features, and the standard historical multi-temporal remote sensing image as the output feature to construct sample data, and divide the sample data into a training sample data set and a test sample data set according to a ratio of 8:2;

[0061] S25: Construct the mangrove classification model based on the input features and output features of the training sample data set using a machine learning algorithm;

[0062] S26: Based on the mangrove classification model, obtain the test results of the mangrove multi-temporal remote sensing image corresponding to the test sample data set according to the input features of the test sample data set;

[0063] S27: Evaluate the accuracy of the mangrove classification model by counting the number of correctly classified mangrove areas and non-mangrove areas in the test results of the mangrove multi-temporal remote sensing image to obtain the accuracy evaluation result of the mangrove classification model;

[0064] S28: When the accuracy evaluation result of the mangrove classification model is lower than the preset accuracy threshold of the mangrove classification model, use an optimization algorithm to adjust the hyperparameters of the mangrove classification model.

[0065] For steps S21 and S22, the historical multi-temporal remote sensing image of the target monitoring area refers to the satellite multi-temporal remote sensing image of the target monitoring area obtained through remote sensing technology at multiple past time points or time periods. Software such as ArcGIS or QGIS can be used to label the mangrove area and non-mangrove area in the historical multi-temporal remote sensing image based on manual visual interpretation, such as by information such as color difference and texture features. Specifically, the mangrove area is labeled as 1 and the non-mangrove area is 0 to obtain the standard historical multi-temporal remote sensing image.

[0066] For step S23, the historical mangrove multi-temporal remote sensing image includes: green light band, short-wave infrared band, red edge band, wide near-infrared band, and narrow near-infrared band. The water body index calculation formula is:

[0067] NDWI = (ρ3 - ρ8) / (ρ3 + ρ8)

[0068] In the formula, NDWI is the water body index, ρ3 is the remote sensing reflectance of the green light band, and ρ8 is the remote sensing reflectance of the wide near-infrared band;

[0069] The calculation formula for the mangrove extraction index is as follows:

[0070]

[0071] In the formula, NDMI is the mangrove extraction index, and ρ SWIR is the remote sensing reflectance of the short-wave infrared band, ρ7 is the remote sensing reflectance of the red-edge band, ρ8 is the remote sensing reflectance of the wide near-infrared band, and ρ 8A is the remote sensing reflectance of the narrow near-infrared band.

[0072] For steps S24 and S25, the training sample dataset is used to construct a mangrove classification model, enabling the mangrove classification model to learn the spectral characteristics of mangrove areas and non-mangrove areas. The test sample dataset is used to evaluate the performance of the mangrove classification model. The machine learning algorithms include support vector machine SVM and random forest.

[0073] By using the normalized vegetation index, water body index, and mangrove extraction index of the training sample dataset as input features, and the standard historical multi-temporal remote sensing images of the training sample dataset as output features to provide a supervision signal, machine learning algorithms such as vector machine SVM or random forest are used to establish a mapping relationship between the input features and output features. By iteratively adjusting the model hyperparameters, such as the penalty parameter and kernel function parameter in vector machine SVM or the number and depth of trees in random forest, etc., the difference between the predicted label and the true label is minimized to construct the mangrove classification model. Among them, the predicted label refers to the training result of the multi-temporal remote sensing images of mangroves corresponding to the training sample dataset output by the model during the training process, and the true label refers to the standard historical multi-temporal remote sensing images of the training sample dataset.

[0074] For steps S26 and S27, please refer to Figure 3 , Figure 3 which is the flowchart of the method for evaluating the accuracy of the mangrove classification model in a mangrove growth monitoring method of this application. The accuracy of the mangrove classification model is evaluated by statistically counting the number of correctly classified mangrove areas and non-mangrove areas in the test results of the multi-temporal remote sensing images of mangroves, and the accuracy evaluation result of the mangrove classification model is obtained, including:

[0075] S271: Randomly generate several test points in the test results of the multi-temporal remote sensing images of mangroves;

[0076] S272: Statistically count the number of test points correctly classified as mangrove areas and the number of test points correctly classified as non-mangrove areas in the test results of the multi-temporal remote sensing images of mangroves;

[0077] S273: Based on the total number of the checkpoints, the number of checkpoints correctly classified in the mangrove area, and the number of checkpoints correctly classified in the non-mangrove area, obtain the accuracy evaluation result of the mangrove classification model according to the following formula:

[0078]

[0079] In the formula, OA is the accuracy evaluation result of the mangrove classification model, x 11 is the number of checkpoints correctly classified in the mangrove area, x 22 is the number of checkpoints correctly classified in the non-mangrove area, and N is the total number of checkpoints.

[0080] For steps S271 - S273, a geographic information system software such as ArcGIS can be used to read the test results of the mangrove multi-temporal remote sensing images, and tools such as creating random points can be used to randomly generate the checkpoints. In this embodiment, the total number of the checkpoints can be set to 200. To avoid the random points being too dense, the minimum distance between two random checkpoints can also be set.

[0081] The judgment rule can specifically be that those skilled in the art, based on the inventive concept of this application and combined with the common general knowledge in the art, manually judge whether each checkpoint in the test results of the mangrove multi-temporal remote sensing images belongs to the mangrove area or the non-mangrove area, and perform statistical calculations on the number of checkpoints correctly classified in the mangrove area and the number of checkpoints correctly classified in the non-mangrove area.

[0082] In other embodiments, the accuracy of the mangrove classification model can also be evaluated by calculating coefficients such as Kappa and F1-Score to obtain the accuracy evaluation result of the mangrove classification model.

[0083] For step S28, the accuracy threshold of the mangrove classification model is preferably set to 85%, and the second optimization algorithm can be an iterative algorithm such as the grid search method, Bayesian optimization, genetic algorithm, gradient boosting, etc. Of course, the accuracy threshold of the mangrove classification model can be adaptively modified according to actual requirements.

[0084] For step S3, in one embodiment, based on the spectral characteristic index of the mangrove multi-temporal remote sensing image and a preset leaf area index model, the canopy coverage degree of the mangrove is analyzed to obtain the leaf area index corresponding to each season. For the leaf area index corresponding to each season, the change amount of the leaf area index between two seasons can also be obtained based on the interpolation method. The leaf area index model refers to a model used to estimate the leaf area density of the mangrove canopy according to the spectral characteristic index of the mangrove multi-temporal remote sensing image. The leaf area index refers to half of the total area of plant leaves on a unit ground surface area and is used to measure the mangrove canopy structure. Please refer to Figure 4 , Figure 4 which is the flowchart of the method for constructing the leaf area index model in a mangrove growth monitoring method of this application. Constructing the leaf area index model includes:

[0085] S31: Obtain a number of historical mangrove multi-temporal remote sensing images and the measured values of the leaf area index in the same period as them. Among them, the historical mangrove multi-temporal remote sensing images include a wide near-infrared band and a red band;

[0086] S32: According to the remote sensing reflectance of the near-infrared band and the red band, based on the normalized difference vegetation index calculation formula, obtain the normalized difference vegetation index. The normalized difference vegetation index calculation formula is:

[0087]

[0088] In the formula, NDVI is the normalized difference vegetation index, ρ8 is the remote sensing reflectance of the wide near-infrared band, and ρ4 is the remote sensing reflectance of the red band;

[0089] S33: According to the measured value of the leaf area index and the normalized difference vegetation index, establish an empirical inversion model through linear regression to obtain the leaf area index model, and its expression is:

[0090] LAI = A × NDVI + b1

[0091] In the formula, LAI is the leaf area index, NDVI is the normalized difference vegetation index, A is the weight coefficient of the normalized difference vegetation index, and b1 is the second empirical coefficient.

[0092] For step S31, the historical mangrove multi-temporal remote sensing image refers to the satellite multi-temporal remote sensing image of the target monitoring area obtained by remote sensing technology at multiple past time points or time periods. The measured value of the leaf area index refers to the leaf area index obtained by field measurement, and the time point for obtaining the measured value of the leaf area index is synchronized with the time point for obtaining the historical mangrove multi-temporal remote sensing image.

[0093] For step S33, the leaf area index model is a unary linear regression model. The weight coefficient of the normalized difference vegetation index and the second empirical coefficient can be calculated by minimizing the sum of the squares of the residuals between the predicted value and the measured value, and methods such as calculating the coefficient of determination and the root mean square error are used to evaluate the leaf area index model.

[0094] In other embodiments, the leaf area index model can also be a multiple linear regression equation. The historical multi-temporal remote sensing images of mangroves also include the blue band. Based on the remote sensing reflectance of the near-infrared band, the red band, and the blue band, and according to the calculation formula of the enhanced vegetation index, the enhanced vegetation index is obtained:

[0095] EVI = G×(ρ8 - ρ4) / (ρ8 + C1×ρ4 - C2×ρ2 + L)

[0096] In the formula, EVI is the enhanced vegetation index, ρ8 is the remote sensing reflectance of the wide near-infrared band, ρ4 is the remote sensing reflectance of the red band, ρ2 is the remote sensing reflectance of the blue band, G is the gain factor, which can be taken as 2.5, C1 and C2 are the atmospheric correction coefficients, which can be taken as 6.0 and 7.5 respectively, and L is the background adjustment parameter, which can be taken as 1.0.

[0097] Combining the measured value of the leaf area index, the normalized difference vegetation index, and the enhanced vegetation index, a multiple linear regression model is constructed. The specific expression is:

[0098] LAI = A1×NDVI + A2×EVI + b1

[0099] In the formula, A1 and A2 are the regression coefficients of the normalized difference vegetation index and the enhanced vegetation index respectively, and can also be obtained by minimizing the sum of the squares of the residuals between the predicted value and the measured value.

[0100] For step S4, in this embodiment, based on a preset sediment content inversion model, the suspended sediment content in the water body is inverted to obtain the inversion result of the suspended sediment concentration corresponding to each season. For the inversion result of the suspended sediment concentration corresponding to each season, the change amount of the inversion result of the suspended sediment concentration between two seasons can also be obtained based on the interpolation method. Please refer to Figure 5 , Figure 5 This is the flowchart of the method for constructing a sediment content inversion model in a mangrove growth monitoring method of this application. Constructing the sediment content inversion model includes:

[0101] S41: Obtain the historical multi-temporal remote sensing images of the mangrove area and the measured concentration of suspended sediment in the water body in the same period.

[0102] S42: Based on the sediment-sensitive reflectance in the historical multi-temporal remote sensing images of mangroves, and based on a preset sediment-sensitive reflectance - optical parameter coupling model, obtain the absorption coefficient of suspended sediment in water and the backscattering coefficient of suspended sediment in water;

[0103] S43: Based on the sediment-sensitive reflectance in the historical multi-temporal remote sensing images of mangroves, obtain the suspended sediment index in water based on the calculation formula of the suspended sediment index in water;

[0104] S44: Based on the measured concentration of suspended sediment in water, the inversion values of the absorption coefficient of suspended sediment in water and the backscattering coefficient of suspended sediment in water, and the suspended sediment index in water, construct the sediment content inversion model through linear regression:

[0105] CSS = k1·NDBI + k1·a sed (λ) + k3·b b,sed (λ) + b2

[0106] In the formula, CSS is the measured concentration of suspended sediment in water or the inversion result of the suspended sediment concentration in water, NDBI is the suspended sediment index in water, a sed (λ) is the absorption coefficient of the suspended sediment in water at wavelength λ, b b,sed (λ) is the backscattering coefficient of the suspended sediment in water at wavelength λ, k1, k2, and k3 are the weight coefficients of the suspended sediment index, the absorption coefficient of the suspended sediment in water, and the backscattering coefficient of the suspended sediment in water respectively, and b2 is the first empirical coefficient.

[0107] For step S42, please refer to Figure 6 , Figure 6 which is the flowchart of the method for constructing a sediment-sensitive reflectance - optical parameter coupling model in a mangrove growth monitoring method of this application. Constructing the sediment-sensitive reflectance - optical parameter coupling model includes:

[0108] S421: Simplify the water body radiation field into an upward radiation flux and a downward radiation flux, and establish a two-stream approximate radiation transfer equation set:

[0109]

[0110] In the formula, Lu(z,λ) is the upward radiation flux at water depth z and wavelength λ, Ld(z,λ) is the downward radiation flux at water depth z and wavelength λ, Ku(z,λ) is the upward diffuse attenuation coefficient at water depth z and wavelength λ, Kd(z,λ) is the downward diffuse attenuation coefficient at water depth z and wavelength λ, Su(z,λ) is the upward source function at water depth z and wavelength λ, and Sd(z,λ) is the downward source function at water depth z and wavelength λ;

[0111] S422: When the medium is uniform, simplify the two-stream approximate radiation transfer equation group to obtain a simplified two-stream approximate radiation transfer equation group:

[0112]

[0113] Where a is the total absorption coefficient, a(λ)=a water (λ)+a sed (λ), a water (λ) is the absorption coefficient of pure water at the preset wavelength λ, a sed (λ) is the absorption coefficient of suspended sediment in water at wavelength λ, b b (λ) is the total backscattering coefficient, b b(λ) =b b,water (λ)+b b,sed (λ), b b,water (λ) is the backscattering coefficient of pure water at the preset wavelength λ, b b,sed (λ) is the backscattering coefficient of suspended sediment in water at wavelength λ;

[0114] S423: The equations of the simplified two-stream approximate radiation transfer equations are combined, and based on the integral factor method and the preset boundary conditions, the sediment sensitive reflectivity-optical parameter coupling model is established, and its expression is:

[0115]

[0116] Where R(λ) is the sediment sensitive reflectivity at wavelength λ, sed (λ) = k a ×λ -α , b b,sed (λ) = k b ×λ -β , α is the wavelength attenuation index of the preset sediment absorption coefficient, β is the wavelength attenuation index of the preset sediment backscattering coefficient, k a is the proportional constant of the sediment absorption coefficient, k b is the proportionality constant of the sediment backscattering coefficient.

[0117] For steps S421 and S422, in this embodiment, in the two-stream approximation, assuming that the medium is uniform, optical parameters such as the total absorption coefficient and the total backscattering coefficient do not change with depth. In this case, Ku(z,λ)=Kd(z,λ)=a(λ)+b b For the upward source function, considering the backscattering contribution from the downward light, Su(z,λ)=b b(λ)×Ld(z, λ). For the downward source function, the forward scattering contribution from the upward light is usually ignored in the two-stream approximation, and at this time Sd(z, λ) ≈ 0. Thus, the two-stream approximation radiative transfer equations are simplified to obtain the simplified two-stream approximation radiative transfer equations.

[0118] For step S423, the pure water absorption coefficient at the preset wavelength λ can be obtained from the standard water body optical parameters published by the International Ocean Color Coordination Group (IOCCG) according to the specific wavelength λ. The pure water backscattering coefficient at the wavelength λ can be calculated and obtained based on the following calculation formula:

[0119]

[0120] In the formula, b w (λ) is the total scattering coefficient of pure water, which can be obtained by retrieving the pure water optical parameters built in the radiative transfer simulation software HydroLight or from the standard water body optical parameters published by the International Ocean Color Coordination Group (IOCCG). β(θ = 180°) is the backscattering probability. In the Rayleigh scattering phase function of pure water, β(θ = 180°) ≈ 0.5.

[0121] Define the sediment-sensitive reflectance as the ratio of the upward light intensity to the downward light intensity at the water surface (z = 0):

[0122] R(λ) = Lu(0, λ) / Ld(0, λ)

[0123] The boundary conditions are that the depth of the water body tends to infinity and the depth of the water body is 0. According to the ratio of the upward light intensity to the downward light intensity at the water surface and the boundary conditions, based on the integral factor method, the sediment-sensitive reflectance-optical parameter coupling model is obtained.

[0124] The wavelength attenuation exponent of the sediment absorption coefficient indicates that the sediment absorption coefficient decreases as the wavelength increases. The wavelength attenuation exponent of the sediment backscattering coefficient indicates that the sediment backscattering coefficient decreases as the wavelength increases. For different types of suspended sediment in water bodies, such as clay, silt, and sand, etc., the corresponding wavelength attenuation exponents of the sediment absorption coefficient and the sediment backscattering coefficient are all different. The wavelength attenuation exponents of the sediment absorption coefficient and the sediment backscattering coefficient can be preset based on the empirical method according to the type of suspended sediment in the water body.

[0125] The sediment-sensitive reflectance includes the remote sensing reflectance in the broad near-infrared band and the blue light band. Substitute the remote sensing reflectance in the broad near-infrared band and the blue light band into the sediment-sensitive reflectance-optical parameter coupling model, and solve the equations to obtain the proportionality constant of the sediment absorption coefficient and the proportionality constant of the sediment backscattering coefficient, so as to obtain the inversion values of the water body suspended sediment absorption coefficient and the water body suspended sediment backscattering coefficient.

[0126] For step S43, the calculation formula of the water body suspended sediment index is:

[0127] NDBI = (ρ8 - ρ2) / (ρ8 + ρ2)

[0128] In the formula, NDBI is the water body suspended sediment index, ρ8 is the remote sensing reflectance in the broad near-infrared band, and ρ2 is the remote sensing reflectance in the blue light band.

[0129] For step S44, combine the measured concentration of the water body suspended sediment, the inversion values of the water body suspended sediment absorption coefficient and the water body suspended sediment backscattering coefficient, and the water body suspended sediment index to construct a multiple linear regression model to obtain the sediment content inversion model. The weight coefficients of the water body suspended sediment index, the water body suspended sediment absorption coefficient, and the water body suspended sediment backscattering coefficient, as well as the first empirical coefficient, can be calculated by minimizing the sum of the squared residuals between the predicted value and the measured value.

[0130] For step S5, in one embodiment, the spatial statistical index includes the Pearson correlation coefficient and the spatial autocorrelation index of the leaf area index and the inversion result of the water body suspended sediment concentration, and is used to divide the stress degree level of the target monitoring area.

[0131] Please refer to Figure 7 , Figure 7 is the flowchart of the method for spatial statistical analysis of the inversion results of the leaf area index and the water body suspended sediment concentration in a mangrove growth monitoring method of the present application. The spatial statistical analysis of the inversion results of the leaf area index and the water body suspended sediment concentration to obtain the spatial statistical index includes:

[0132] S51: Randomly sample based on the mangrove multi-temporal remote sensing image to obtain an observation data set at different sampling points. Among them, the observation data set at any sampling point includes the inversion results of the leaf area index and the water body suspended sediment concentration at the same position and the same time point;

[0133] S52: According to the observation data set, based on the Pearson correlation coefficient calculation formula, obtain the Pearson correlation coefficient of the inversion results of the leaf area index and the water body suspended sediment concentration:

[0134]

[0135] In the formula, r is the Pearson correlation coefficient between the leaf area index and the inversion result of the water body suspended sediment concentration, and x i is the leaf area index in the i-th observation dataset, and y i is the water body suspended sediment content in the i-th observation dataset, is the average value of the leaf area index of each observation dataset, is the average value of the water body suspended sediment content of each observation dataset, and n is the number of the observation datasets;

[0136] S53: According to the observation dataset, based on the inverse distance weighting interpolation method, generate a continuous spatial distribution surface of the inversion results of the leaf area index and the water body suspended sediment concentration, and obtain the interpolation of the inversion results of the leaf area index and the water body suspended sediment concentration in a number of spatial units. The specific formula is:

[0137]

[0138] In the formula, Z(P) is the interpolation of the leaf area index or the inversion result of the water body suspended sediment concentration in the spatial unit P, and Z i is the leaf area index or the water body suspended sediment content in the i-th observation dataset, is the distance from the collection point corresponding to the i-th observation dataset to the spatial unit P, x i and y i are the coordinates of the collection point corresponding to the i-th observation dataset, and x P and y P are the coordinates of the spatial unit P;

[0139] S54: According to the spatial unit interpolation corresponding to the inversion results of the leaf area index and the water body suspended sediment concentration, perform a bivariate spatial autocorrelation analysis on the inversion results of the leaf area index and the water body suspended sediment concentration, and obtain the spatial autocorrelation index of the inversion results of the leaf area index and the water body suspended sediment concentration. The specific formula is:

[0140]

[0141] In the formula, I is the spatial autocorrelation index, and g i is the interpolation of the leaf area index in the i-th spatial unit, and h j is the interpolation of the water body suspended sediment content in the j-th spatial unit, is the average value of the interpolation of the leaf area index in each spatial unit, is the average value of the interpolation of the water body suspended sediment content in each spatial unit, and m is the number of the spatial units, wij is a preset spatial weight matrix.

[0142] According to steps S51 - S54, the Pearson correlation coefficient between the retrieved results of the leaf area index and the water - body suspended sediment concentration is a statistical indicator measuring the degree of linear correlation between the retrieved results of the leaf area index and the water - body suspended sediment concentration. Its value ranges between - 1 and 1, where 1 indicates a perfect positive correlation, - 1 indicates a perfect negative correlation, and 0 indicates no correlation.

[0143] Based on the inverse - distance - weighted interpolation method, the retrieved results of the leaf area index and the water - body suspended sediment concentration at discrete sampling points are converted into a continuous spatial distribution surface, and a bivariate spatial autocorrelation analysis is performed to analyze the co - variation pattern of the retrieved results of the leaf area index and the water - body suspended sediment concentration on this spatial distribution surface.

[0144] Exemplarily, if the retrieved results of the leaf area index and the water - body suspended sediment concentration show a significant negative correlation in space (the spatial autocorrelation index is negative and the Pearson correlation coefficient r < 0), it indicates that there is an inverse association pattern in their spatial distribution. This inverse association pattern may stem from: in areas with a high leaf area index, the sediment - carrying capacity of the water flow is reduced through the blocking effect of mangroves, resulting in a decrease in the suspended sediment concentration; or in areas with a retrieved result of the water - body suspended sediment concentration, the growth of mangroves is inhibited due to a decrease in light transmittance or sediment coverage, leading to a low leaf area index.

[0145] For step S6, in this embodiment, the change parameter of the mangrove growth status can also be calculated based on the following formula:

[0146]

[0147] In the formula, ΔNm is the change parameter of the mangrove growth status, and Nm1 and Nm2 are the mangrove growth status parameters corresponding to any two seasons respectively.

[0148] Please refer to Figure 8 , Figure 8 is the flowchart of the method for constructing a mangrove - water - body suspended sediment stress coupling model in a mangrove growth monitoring method of the present application. Constructing the mangrove - water - body suspended sediment stress coupling model includes:

[0149] S61: Obtain the retrieved results of the leaf area index and the water - body suspended sediment concentration at multiple historical time points according to the historical multi - temporal remote - sensing images of mangroves;

[0150] S62: Obtain the mangrove canopy width, soil humidity, and temperature at each of the historical time points in the mangrove area;

[0151] S63: Based on the retrieved results of the mangrove canopy width, leaf area index, water body suspended sediment concentration, soil humidity, and temperature, a mangrove-water body suspended sediment stress coupling model is constructed through linear regression, and its expression is:

[0152] Nm = Ns·[(ω1·LAI)+(ω2·θ)+(ω3·T)] + b3

[0153] In the formula, Nm can be the mangrove canopy width, the mangrove growth condition change parameter, or the mangrove growth condition parameter; Ns is the retrieved result of the water body suspended sediment concentration; LAI is the leaf area index; θ is the soil humidity; T is the temperature; b3 is the third empirical coefficient; ω1, ω2, and ω3 are the weight coefficients of the leaf area index, soil humidity, and temperature, respectively.

[0154] For step S61, specifically, based on the historical multi-temporal remote sensing images of the mangroves, for each pixel or quadrat area, the leaf area index of the later temporal phase is subtracted from the leaf area index of the previous temporal phase, and the change amount of the leaf area index in the area between the two historical time points can be obtained. This calculation process is also applicable to calculating the change amount of the retrieved result of the water body suspended sediment concentration and the change amounts of the following mangrove canopy width, soil humidity, and temperature.

[0155] For step S62, the mangrove canopy width, soil humidity, and temperature can be obtained from the local environmental monitoring agency. Among them, obtaining the leaf area index, the retrieved result of the water body suspended sediment concentration, the mangrove canopy width, soil humidity, and temperature should be based on the same historical time point.

[0156] For step S63, based on the mangrove canopy width, leaf area index, the retrieved result of the water body suspended sediment concentration, soil humidity, and temperature, a multiple linear regression model is constructed as the mangrove-water body suspended sediment stress coupling model. Among them, the weight coefficients of the leaf area index, soil humidity, and temperature can be calculated by minimizing the sum of the squares of the residuals between the predicted value and the measured value.

[0157] Nm can be the mangrove canopy width, the mangrove growth condition change parameter, or the mangrove growth condition parameter. When the change amounts of the leaf area index, the retrieved result of the water body suspended sediment concentration, and the change amounts of the mangrove canopy width, soil humidity, and temperature are input into the mangrove-water body suspended sediment stress coupling model, the output parameter is the mangrove growth condition change parameter.

[0158] For step S7, in this embodiment, please refer to Figure 9 , Figure 9This is a method flow chart for dividing the stress degree levels of the target monitoring area in a mangrove growth monitoring method of this application. Based on the spatial statistical index and the mangrove growth status change parameters, and based on the preset stress level division rules, dividing the stress degree levels of the target monitoring area includes:

[0159] S71: Divide the spatial statistical index and the mangrove growth status change parameters into several change amplitude ranges based on the analytic hierarchy process;

[0160] S72: Divide the stress degree levels of the target monitoring area according to the several change amplitude ranges.

[0161] For steps S71 and S72, exemplarily, divide the spatial statistical index and the mangrove growth status change parameters into four change amplitude ranges based on the analytic hierarchy process, and divide the stress degree levels of the target monitoring area into four levels, which can be specifically shown in Table 1 below:

[0162] Table 1:

[0163]

[0164]

[0165] When r ≤ -0.7, I < -0.3, and Nm ≤ -30%, it indicates that the water body suspended sediment concentration has a strong negative correlation with the mangrove growth status, the sediment shows significant spatial suppression on the mangrove growth, and the mangrove growth status has a severe decline. At this time, divide the target monitoring area into the severe stress level.

[0166] When -0.7 < r ≤ -0.4, -0.3 ≤ I < -0.1, and -30% < Nm ≤ -10%, it indicates that the water body suspended sediment concentration has a medium negative correlation with the mangrove growth status, the sediment shows weak spatial suppression on the mangrove growth, and the mangrove growth status has a moderate decline. At this time, divide the target monitoring area into the moderate stress level.

[0167] When -0.4 < r ≤ -0.2, -0.1 ≤ I < 0, and -10% < Nm ≤ 0%, it indicates that the water body suspended sediment concentration has a weak negative correlation with the mangrove growth status, the sediment has no spatial suppression on the mangrove growth, and the mangrove growth status has a slight decline. At this time, divide the target monitoring area into the mild stress level.

[0168] When r > -0.2, I ≥ 0, and Nm > 0%, it indicates that the water body suspended sediment concentration has a relatively weak negative correlation with the mangrove growth status, the sediment has no spatial association with the mangrove growth, and the mangrove growth status is normal. At this time, divide the target monitoring area into the non-stress level.

[0169] In other embodiments, according to actual requirements, the number of the change amplitude ranges can also be adaptively modified, and the thresholds of the Pearson correlation coefficient r, the spatial autocorrelation index I, and the mangrove growth condition change parameter Nm in the change amplitude ranges can be adaptively modified. For example, in a strong tidal area where sediment is easily washed away, the threshold of the spatial autocorrelation index I can be relaxed, and I corresponding to the severe stress level can be set to < -0.5, etc. According to the change amplitude ranges, the stress level classification of the target monitoring area is adaptively modified.

[0170] For step S8, in this embodiment, the warning instruction includes the generation time, the geographical coordinates of the target monitoring area, the stress level, the key parameters, and the warning suggestions, etc. Among them, the stress level includes the severe stress, moderate stress, mild stress, and no stress. The key parameters include the Pearson correlation coefficient r, the spatial autocorrelation index I, and the mangrove growth condition change parameter Nm.

[0171] When the stress level classification result of the target monitoring area is no stress, the warning suggestion can be to maintain routine monitoring, record the mangrove growth data, and there is no need to initiate an emergency response. When the stress level classification result of the target monitoring area is mild stress, the warning suggestion can be to increase the frequency of inspection and monitoring and submit a monitoring report to the management department. When the stress level classification result of the target monitoring area is moderate stress, the warning suggestion can be to initiate ecological restoration measures, manually clean the silt on the surface layer (the area with a thickness > 15 cm) of the mangrove root area to promote root ventilation.

[0172] When the stress level classification result of the target monitoring area is severe stress, the warning suggestion can be to initiate emergency intervention measures, manually clean the silt on the surface layer (the area with a thickness > 30 cm) of the mangrove root area, replant mangrove species resistant to high sedimentation (such as Avicennia marina, Sonneratia caseolaris), use container seedlings to improve the survival rate, and implement a watershed sediment interception project (such as an ecological sediment dam) to reduce the sediment input from the upstream, etc.

[0173] The mangrove growth monitoring server can push the warning instruction to the mangrove growth warning device through a wireless network. The mangrove growth warning device can be a display screen, a mobile terminal, etc. Of course, the mangrove growth monitoring server can also be docked with the warning device system. In other embodiments, the warning instruction can be adaptively modified according to actual monitoring requirements.

[0174] Embodiment 2

[0175] Please refer to Figure 10 , Figure 10 which is a schematic diagram of a mangrove growth monitoring system of the present application.

[0176] This application also provides a mangrove growth monitoring system, including:

[0177] Multi-temporal remote sensing image acquisition module 1: It is used to acquire multi-temporal remote sensing images of the target monitoring area and extract the spectral feature indices of the multi-temporal remote sensing images. Among them, the target monitoring area includes mangrove areas and non-mangrove areas;

[0178] Mangrove classification module 2: It is used to classify the multi-temporal remote sensing images based on the spectral feature indices and a preset mangrove classification model, and generate mangrove multi-temporal remote sensing images that only include the mangrove areas;

[0179] Leaf area index calculation module 3: It is used to analyze the canopy coverage degree of mangroves based on the spectral feature indices of the mangrove multi-temporal remote sensing images and a preset leaf area index model to obtain the leaf area index;

[0180] Water body suspended sediment concentration inversion module 4: It is used to extract the sediment-sensitive reflectance in the mangrove multi-temporal remote sensing images, and based on a preset sediment content inversion model, invert the water body suspended sediment content in the water body to obtain the water body suspended sediment concentration inversion result. Among them, the sediment-sensitive reflectance is the remote sensing reflectance of the band in the mangrove multi-temporal remote sensing images that is sensitive to water body suspended sediment;

[0181] Spatial statistical index calculation module 5: It is used to perform spatial statistical analysis on the leaf area index and the water body suspended sediment concentration inversion result to obtain the spatial statistical index;

[0182] Mangrove growth condition change parameter calculation module 6: It is used to analyze the stress degree of water body suspended sediment on mangrove growth based on the leaf area index and the water body suspended sediment concentration inversion result and a preset mangrove-water body suspended sediment stress coupling model to obtain the mangrove growth condition change parameter;

[0183] Stress degree level division module 7: It is used to divide the stress degree level of the target monitoring area based on the spatial statistical index and the mangrove growth condition change parameter and a preset stress level division rule;

[0184] Early warning instruction transmission module 8: It is used to generate an early warning instruction corresponding to the stress degree level according to the stress degree level division result of the target monitoring area, and transmit the early warning instruction to the mangrove growth early warning device.

[0185] It should be noted that the data obtained when a mangrove growth monitoring system provided by the present application implements a mangrove growth monitoring method are stored in the storage of the system one by one. When relevant calculations need to be performed, the data required for the calculations can be directly obtained from the storage correspondingly.

[0186] It should also be noted that when a mangrove growth monitoring system provided in the above embodiment implements a mangrove growth monitoring method, only the division of the above functional modules is used as an example for illustration. In actual applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, a mangrove growth monitoring system provided in the above embodiment and a mangrove growth monitoring method in Embodiment 1 belong to the same concept. The implementation process thereof is detailed in the method embodiment and will not be elaborated here.

[0187] The present application is not limited to the above embodiments. If various changes or deformations to the present application do not depart from the spirit and scope of the present application, and if these changes and deformations are within the scope of the claims of the present application and equivalent technical scope, then the present application also intends to include these changes and deformations.

Claims

1. A method for monitoring the growth of mangroves, characterized in that, It includes the following steps: Obtain multi-temporal remote sensing images of the target monitoring area, and extract the spectral characteristic indices of the multi-temporal remote sensing images, where the target monitoring area includes a mangrove area and a non-mangrove area; According to the spectral characteristic indices, based on a preset mangrove classification model, classify the multi-temporal remote sensing images to generate multi-temporal remote sensing images of mangroves that only include the mangrove area; According to the spectral characteristic indices of the multi-temporal remote sensing images of mangroves, based on a preset leaf area index model, analyze the canopy coverage degree of mangroves to obtain the leaf area index; Extract the sediment-sensitive reflectance in the multi-temporal remote sensing images of mangroves, and based on a preset sediment content inversion model, invert the suspended sediment content in the water body to obtain an inversion result of the suspended sediment concentration in the water body, where the sediment-sensitive reflectance is the remote sensing reflectance of the band sensitive to the suspended sediment in the water body in the multi-temporal remote sensing images of mangroves; Conduct spatial statistical analysis on the leaf area index and the inversion result of the suspended sediment concentration in the water body to obtain a spatial statistical index; According to the leaf area index and the inversion result of the suspended sediment concentration in the water body, based on a preset mangrove-suspended sediment stress coupling model, analyze the stress degree of the suspended sediment in the water body on the growth of mangroves to obtain the change parameters of the growth status of mangroves; According to the spatial statistical index and the change parameters of the growth status of mangroves, based on a preset stress level division rule, divide the stress degree level of the target monitoring area; According to the division result of the stress degree level of the target monitoring area, generate a warning instruction corresponding to the stress degree level, and transmit the warning instruction to the mangrove growth warning device.

2. The mangrove growth monitoring method according to claim 1, characterized in that, Constructing the sediment content inversion model includes: Obtain historical multi-temporal remote sensing images of the mangrove area and the measured concentration of suspended sediment in the water body in the same period; According to the sediment-sensitive reflectance in the historical multi-temporal remote sensing images of mangroves, based on a preset sediment-sensitive reflectance-optical parameter coupling model, obtain the absorption coefficient of suspended sediment in the water body and the backscattering coefficient of suspended sediment in the water body; According to the sediment-sensitive reflectance in the historical multi-temporal remote sensing images of mangroves, based on the calculation formula of the suspended sediment index in the water body, obtain the suspended sediment index in the water body; According to the measured concentration of suspended sediment in the water body, the absorption coefficient of suspended sediment in the water body, the backscattering coefficient of suspended sediment in the water body, and the suspended sediment index in the water body, construct the sediment content inversion model through linear regression: CSS = k1·NDBI + k1·a sed (λ) + k3·b b,sed (λ) + b2 Wherein, CSS is the measured concentration of the suspended sediment in the water body or the inversion result of the suspended sediment concentration in the water body, NDBI is the suspended sediment index of the water body, a sed (λ) is the absorption coefficient of the suspended sediment in the water body at the wavelength of λ, b b,sed (λ) is the backscattering coefficient of the suspended sediment in the water body at the wavelength of λ, k1, k2, and k3 are the weighting coefficients of the suspended sediment index, the absorption coefficient of the suspended sediment in the water body, and the backscattering coefficient of the suspended sediment in the water body respectively, and b2 is the second empirical coefficient.

3. The mangrove growth monitoring method according to claim 2, characterized in that, Constructing the sediment-sensitive reflectance-optical parameter coupling model includes: Simplify the water body radiation field into an upward radiation flux and a downward radiation flux, and establish a two-stream approximate radiation transfer equation set: Where Lu(z,λ) is the upward radiation flow with a water depth of z and a wavelength of λ, Ld(z,λ) is the downward radiation flow with a water depth of z and a wavelength of λ, Ku(z,λ) is the upward diffuse attenuation coefficient with a water depth of z and a wavelength of λ, Kd(z,λ) is the downward diffuse attenuation coefficient with a water depth of z and a wavelength of λ, Su(z,λ) is the upward source function with a water depth of z and a wavelength of λ, and Sd(z,λ) is the downward source function with a water depth of z and a wavelength of λ; When the medium is uniform, the two-stream approximate radiation transfer equations are simplified to obtain the simplified two-stream approximate radiation transfer equations: Wherein, a is the total absorption coefficient, and a(λ) = a water (λ) + a sed (λ), where a water (λ) is the absorption coefficient of pure water at a preset wavelength of λ, and a sed (λ) is the absorption coefficient of suspended sediment in water at a wavelength of λ, and b b (λ) is the total backscattering coefficient, and b b(λ) = b b,water (λ) + b b,sed (λ), where b b,water (λ) is the backscattering coefficient of pure water at a preset wavelength of λ, and b b,sed (λ) is the backscattering coefficient of suspended sediment in water at a wavelength of λ; The equations of the simplified two-stream approximate radiation transfer equations are combined, and based on the integral factor method and the preset boundary conditions, the sediment sensitive reflectivity-optical parameter coupling model is established, and its expression is: where R(λ) is the sediment-sensitive reflectance at wavelength λ, a sed (λ) = k a ×λ -ɑ , b b,sed (λ) = k b ×λ -β , α is the wavelength attenuation exponent of the preset sediment absorption coefficient, β is the wavelength attenuation exponent of the preset sediment backscattering coefficient, k a is the proportionality constant of the sediment absorption coefficient, k b is the proportionality constant of the sediment backscattering coefficient.

4. A method for monitoring the growth of mangroves according to claim 1, characterized in that, The spatial statistical index includes the Pearson correlation coefficient of the leaf area index and the inversion result of the suspended sediment concentration in the water body; The spatial statistical analysis of the leaf area index and the water body suspended sediment concentration inversion results is performed to obtain a spatial statistical index, including: Random sampling is performed based on the multi-temporal remote sensing images of the mangrove forest to obtain observation data sets at different sampling points, wherein the observation data set at any sampling point contains the leaf area index and water body suspended sediment concentration inversion results at the same location and the same time point; According to the observation data set, based on the Pearson correlation coefficient calculation formula, the Pearson correlation coefficient of the leaf area index and the water body suspended sediment concentration inversion result is obtained: Wherein, r is the Pearson correlation coefficient between the leaf area index and the inversion result of the water body suspended sediment concentration, and x i is the leaf area index in the i-th observation dataset, and y i is the water body suspended sediment content in the i-th observation dataset, is the average value of the leaf area index of each observation dataset, is the average value of the water body suspended sediment content of each observation dataset, and n is the number of the observation datasets.

5. A method for monitoring the growth of mangroves according to claim 4, characterized in that, The spatial statistical index also includes the spatial autocorrelation index of the leaf area index and the inversion result of the water body suspended sediment concentration; The spatial statistical analysis of the leaf area index and the water body suspended sediment concentration inversion results to obtain the spatial statistical index also includes: According to the observation data set, based on the inverse distance weighted interpolation method, a continuous spatial distribution surface of the leaf area index and the water body suspended sediment concentration inversion results is generated, and the interpolation of the leaf area index and the water body suspended sediment concentration inversion results in several spatial units is obtained. The specific formula is: Wherein, Z(P) is the interpolation of the leaf area index or the inversion result of the water body suspended sediment concentration in the spatial unit P, and Z i is the leaf area index or the water body suspended sediment content in the i-th observation dataset, is the distance from the collection point corresponding to the i-th observation dataset to the spatial unit P, x i and y i are the coordinates of the collection point corresponding to the i-th observation dataset, x P and y P are the coordinates of the spatial unit P; According to the spatial unit interpolation corresponding to the inversion results of the leaf area index and the suspended sediment concentration in water body, a bivariate spatial autocorrelation analysis is performed on the inversion results of the leaf area index and the suspended sediment concentration in water body to obtain the spatial autocorrelation index of the inversion results of the leaf area index and the suspended sediment concentration in water body. The specific formula is: Wherein, I is the spatial autocorrelation index, g i is the interpolation of the leaf area index in the i-th spatial unit, h j is the interpolation of the water suspended sediment content in the j-th spatial unit, is the average value of the interpolations of the leaf area index in each spatial unit, is the average value of the interpolations of the water suspended sediment content in each spatial unit, m is the number of the spatial units, w ij is the preset spatial weight matrix.

6. The mangrove growth monitoring method according to claim 4, characterized in that, The method of classifying the stress level of the target monitoring area according to the spatial statistical index and the mangrove growth condition change parameter based on a preset stress level classification rule includes: Based on the hierarchical analysis method, the spatial statistical index and the mangrove growth condition change parameters are divided into several ranges of change; The target monitoring area is divided into stress degree levels according to the several ranges of variation.

7. A method for monitoring the growth of mangroves according to claim 2, characterized in that, The mangrove-water suspended sediment stress coupling model is constructed, including: Obtaining the leaf area index and water body suspended sediment concentration inversion results at multiple historical time points based on the historical mangrove multi-temporal remote sensing images; Obtain the mangrove canopy width, soil humidity, and temperature at each of the historical time points in the mangrove area; Based on the mangrove canopy width, leaf area index, inversion result of water body suspended sediment concentration, soil humidity, and temperature, construct the mangrove-water body suspended sediment stress coupling model through linear regression. Its expression is: Nm = Ns·[(ω1·LAI)+(ω2·θ)+(ω3·T)] + b3 In the formula, Nm is the mangrove canopy width or the change parameter of mangrove growth status, Ns is the inversion result of the water body suspended sediment concentration, LAI is the leaf area index, Δθ is the soil humidity, ΔT is the temperature, ω1, ω2, and ω3 are the weight coefficients of the leaf area index, soil humidity, and temperature respectively, and b3 is the third empirical coefficient.

8. A method for monitoring the growth of mangroves according to claim 2, characterized in that, The spectral feature index at least includes: normalized difference vegetation index; Construct the leaf area index model, including: Obtain several historical multi-temporal remote sensing images of the mangroves and the measured values of the leaf area index in the same period. Among them, the historical multi-temporal remote sensing images of the mangroves include a wide near-infrared band and a red band; Based on the remote sensing reflectance of the near-infrared band and the red band, obtain the normalized difference vegetation index according to the normalized difference vegetation index calculation formula; Based on the measured values of the leaf area index and the normalized difference vegetation index, establish an empirical inversion model through linear regression to obtain the leaf area index model. Its expression is: LAI = A×NDVI + b1 In the formula, LAI is the leaf area index, NDVI is the normalized difference vegetation index, A is the weight coefficient of the normalized difference vegetation index, and b1 is the first empirical coefficient.

9. A method for monitoring the growth of mangroves according to claim 7, characterized in that, The spectral feature index also includes: water body index and mangrove extraction index; Construct the mangrove classification model, including: Obtain several historical multi-temporal remote sensing images of the target monitoring area. Among them, the historical multi-temporal remote sensing images of the mangroves include: green band, short-wave infrared band, red edge band, wide near-infrared band, and narrow near-infrared band; Label the historical multi-temporal remote sensing images into mangrove areas and non-mangrove areas to obtain standard historical multi-temporal remote sensing images; Based on the remote sensing reflectance of the green band, short-wave infrared band, red edge band, wide near-infrared band, and narrow near-infrared band, calculate the water body index and the mangrove extraction index according to the preset water body index calculation formula and mangrove extraction index calculation formula; Take the normalized difference vegetation index, water body index, and mangrove extraction index as input features, and the standard historical multi-temporal remote sensing images as output features to construct sample data, and divide the sample data into a training sample data set and a test sample data set according to the ratio of 8:2; Based on the input features and output features of the training sample data set, construct the mangrove classification model based on the machine learning algorithm; Based on the input features of the test sample data set, obtain the test results of the mangrove multi-temporal remote sensing images corresponding to the test sample data set based on the mangrove classification model; The accuracy of the mangrove classification model is evaluated by counting the number of correctly classified mangrove areas and non-mangrove areas in the test results of the multi-temporal remote sensing images of the mangroves, and the accuracy evaluation result of the mangrove classification model is obtained. When the accuracy evaluation result of the mangrove classification model is lower than the preset accuracy threshold of the mangrove classification model, the hyperparameters of the mangrove classification model are adjusted using an optimization algorithm.

10. A mangrove growth monitoring system, characterized in that, It includes: Multi-temporal remote sensing image acquisition module: used to acquire multi-temporal remote sensing images of the target monitoring area and extract the spectral feature indices of the multi-temporal remote sensing images, where the target monitoring area includes mangrove areas and non-mangrove areas. Mangrove classification module: used to classify the multi-temporal remote sensing images based on the spectral feature indices and a preset mangrove classification model to generate multi-temporal remote sensing images of mangroves that only include the mangrove areas. Leaf area index calculation module: used to analyze the canopy coverage degree of mangroves based on the spectral feature indices of the multi-temporal remote sensing images of mangroves and a preset leaf area index model to obtain the leaf area index. Water body suspended sediment concentration inversion module: used to extract the sediment-sensitive reflectance in the multi-temporal remote sensing images of mangroves and invert the water body suspended sediment content in the water body based on a preset sediment content inversion model to obtain the water body suspended sediment concentration inversion result, where the sediment-sensitive reflectance is the remote sensing reflectance of the band in the multi-temporal remote sensing images of mangroves that is sensitive to water body suspended sediment. Spatial statistic index calculation module: used to perform spatial statistical analysis on the leaf area index and the water body suspended sediment concentration inversion result to obtain the spatial statistic index. Mangrove growth status change parameter calculation module: used to analyze the stress degree of water body suspended sediment on mangrove growth based on the leaf area index and the water body suspended sediment concentration inversion result and a preset mangrove-water body suspended sediment stress coupling model to obtain the mangrove growth status change parameters. Stress degree level division module: used to divide the stress degree level of the target monitoring area based on the spatial statistic index and the mangrove growth status change parameters and a preset stress level division rule. Early warning instruction transmission module: used to generate an early warning instruction corresponding to the stress degree level according to the stress degree level division result of the target monitoring area and transmit the early warning instruction to the mangrove growth early warning device.

Images