Estuary coastal wetland multi-source land and sea environment information integration method

Abstract

The invention discloses an estuary coastal wetland multi-source land-sea environment information integration method. The method comprises the following steps: step 1, acquiring multi-data-source remote sensing image data, sea chart data, ecological data and climate data of an estuary coastal wetland; 2, calculating to obtain the land utilization condition, the vegetation coverage dynamic change characteristic and the leaf surface spectrum characteristic of the estuary coastal wetland by utilizing the multi-data-source remote sensing image data and the sea chart data of the estuary coastal wetland; and step 3, fusing ecological data, climate data, land utilization conditions, vegetation coverage dynamic change characteristics and leaf surface spectrum characteristics by adopting a big data technology to obtain estuary coastal wetland multi-source land-sea environment information.

Description

Technical Field

[0001] The invention belongs to the technical field of wetland monitoring, and in particular relates to a method for integrating multi-source land and sea environmental information of estuarine coastal wetlands. Background Art

[0002] Estuarine coastal wetlands are the interlaced transition zone between terrestrial and marine ecosystems, and play an important role in regulating climate, conserving water resources, degrading pollutants, responding to climate change, maintaining the global carbon cycle, and protecting biodiversity. Due to the impact of high-intensity human activities and climate change, the estuarine coastal wetland ecosystem has undergone significant changes over the past few decades. Coastal wetlands are under the influence of dual stresses, resulting in a continuous reduction in wetland area, degradation of wetland habitats, a decline in ecological environmental carrying capacity, and an imbalance in ecosystem functions. Maintaining wetlands in optimal condition is an urgent issue, requiring research on wetland ecosystem assessment, regulation, and biomass management, as well as extensive monitoring and assessment of wetlands.

[0003] Existing wetland monitoring systems generally detect changes in ecology, meteorology and water quality, and use remote sensing technology to observe wetland changes. However, the current impact on wetland degradation is not only caused by some local climate changes. Global climate change and human activities are also important factors affecting wetland transformation. There is an urgent need for a comprehensive data integration method to provide data support for wetland management and protection. Summary of the invention

[0004] Purpose of the invention: To solve the problem that the existing wetland monitoring system cannot fully acquire data related to wetland monitoring, the present invention proposes a method for integrating multi-source land and sea environmental information of estuarine coastal wetlands.

[0005] Technical solution: A method for integrating multi-source land and sea environmental information of estuarine coastal wetlands, comprising the following steps:

[0006] Step 1: Obtain remote sensing image data, nautical chart data, ecological data and climate data from multiple sources of estuarine coastal wetlands;

[0007] Step 2: Using the multi-data source remote sensing image data and nautical chart data of the estuarine coastal wetlands, calculate the land use status, dynamic change characteristics of vegetation coverage and leaf area index of the estuarine coastal wetlands;

[0008] Step 3: Integrate ecological data, climate data, land use conditions, dynamic characteristics of vegetation cover, and leaf area index to obtain multi-source land and sea environmental information of estuarine coastal wetlands.

[0009] Furthermore, the ecological data include: greenhouse gases, water quality detection data, water level data, leaf spectrum data and leaf area index of plant types at sampling points;

[0010] The climate data include: climate data of estuarine coastal wetland ecosystems and global climate data.

[0011] Furthermore, the greenhouse gases include methane, carbon dioxide and nitrous oxide; the water quality detection data include carbon content data, phosphorus content data, nitrogen content data, biological content data, chemical oxygen demand, pH value, total oxygen demand, conductivity, turbidity, suspended solids and dissolved oxygen; the biological content data include: algae content data and bacterial content data.

[0012] Furthermore, the land use conditions of the estuarine coastal wetlands include: the area of ​​the reclaimed area and the conditions of erosion and siltation;

[0013] The land use status of the estuarine coastal wetlands is calculated by using the multi-data source remote sensing image data and nautical chart data of the estuarine coastal wetlands. The specific operations include:

[0014] Perform radiation correction and false color synthesis processing on remote sensing image data from multiple data sources to obtain processed remote sensing images;

[0015] Performing image registration on the processed remote sensing image to obtain a registered remote sensing image;

[0016] Combined with the field exploration data, the registered remote sensing image is divided into regions, and the remote sensing image is digitized into raster images to obtain a remote sensing image with digitized raster images;

[0017] By using GIS software to overlay the digitized images of raster images from different periods, we can analyze the increased areas between the shorelines in different periods and calculate the areas of reclamation in different periods.

[0018] The isobaths of the chart data are digitized and converted into different formats to obtain the processed chart data; the processed chart data are imported into the GIS software for difference calculation and comparative analysis to extract the isobaths of previous years; based on the isobaths of previous years, the erosion and siltation conditions of the estuarine coastal wetlands are obtained.

[0019] Furthermore, the multi-data source remote sensing image data and nautical chart data of the estuarine coastal wetland are used to calculate the land use status, dynamic change characteristics of vegetation coverage and leaf area index of the estuarine coastal wetland. The specific operations include:

[0020] The vegetation index is calculated using multi-data source remote sensing image data and nautical chart data of estuarine coastal wetlands;

[0021] By analyzing the changes in vegetation coverage, the dynamic change characteristics of vegetation coverage and its leaf area index can be obtained.

[0022] Furthermore, the vegetation index is calculated by using the remote sensing image data and nautical chart data from multiple data sources of estuarine coastal wetlands, specifically including:

[0023] The vegetation index includes the normalized difference vegetation index and the enhanced vegetation index;

[0024] By calculating the reflectance difference between the near-infrared band and the red light band in the multi-data source remote sensing image data of the estuarine coastal wetland, the normalized difference vegetation index reflecting the luxuriance of vegetation is obtained.

[0025] By calculating the reflectance differences among near-infrared bands, red bands, and blue bands in multi-data source remote sensing image data of estuarine coastal wetlands, the enhanced vegetation index for assessing the greenness of vegetation is obtained.

[0026] Furthermore, the dynamic change characteristics of vegetation coverage and its leaf area index are obtained by analyzing the change of vegetation coverage, and the specific operations include:

[0027] Using professional remote sensing knowledge and geographic information system technology, we classified and interpreted the multi-source remote sensing image data of estuarine coastal wetlands to obtain vegetation coverage information and then the changes in vegetation coverage;

[0028] By fusing and analyzing remote sensing image data from multiple sources of estuarine coastal wetlands, the leaf area index, which is used to measure the ratio of plant leaf area to land area, was estimated.

[0029] Furthermore, the fusing of the remote sensing image data from multiple data sources of the estuarine coastal wetland includes: fusing the remote sensing image data from multiple data sources of the estuarine coastal wetland at the pixel level, at the feature level or at the decision level.

[0030] Furthermore, in step 3, statistical methods, information theory methods or cognitive model methods are used to integrate ecological data, climate data, land use conditions, dynamic change characteristics of vegetation cover and leaf area index to obtain multi-source land and sea environmental information of estuarine coastal wetlands.

[0031] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0032] (1) The present invention establishes a multi-data source remote sensing image database, a wetland monitoring database, and a climate monitoring database, which not only monitors the change data of the wetland itself, but also monitors the human activity data and global climate data, and comprehensively monitors the ecological data, climate change, and human activity data that affect wetland changes;

[0033] (2) The present invention uses data fusion technology to clarify the relationship between the correlation and regularity of internal data of wetland ecosystems and the expressiveness and influence of external data, which facilitates the subsequent analysis of wetland data and provides data support for subsequent wetland protection and management. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 This is a flow chart of a method for integrating multi-source land and sea environmental information of estuarine coastal wetlands proposed by the present invention;

[0035] Figure 2 This is a block diagram of the application system of the present invention;

[0036] Figure 3 It is the principle diagram of step 2 of the present invention;

[0037] Figure 4 This is a schematic diagram of an application scenario of the present invention. DETAILED DESCRIPTION

[0038] The technical solution of the present invention is now further described in conjunction with the accompanying drawings and embodiments.

[0039] Example 1

[0040] like Figure 1 As shown, the present invention discloses a method for integrating multi-source land and sea environmental information of estuarine coastal wetlands, which mainly includes the following steps:

[0041] Step 1: Construct a multi-data source remote sensing image database, a wetland monitoring database, and a climate monitoring database; the specific operations include:

[0042] Acquire historical multi-data source remote sensing image data and nautical chart data of estuarine coastal wetland ecosystems, acquire real-time updated multi-data source remote sensing image data and nautical chart data of estuarine coastal wetland ecosystems, and store the multi-data source remote sensing image data and nautical chart data in a multi-data source remote sensing image database.

[0043] Obtain historical and real-time ecological data of the wetland. In this embodiment, the ecological data includes but is not limited to greenhouse gas data, water quality detection data, water level data, spectral data of plant types at sampling points, and leaf area index; wherein greenhouse gases include but are not limited to methane data, carbon dioxide data, and nitrous oxide data; water quality detection data include but are not limited to parameters: carbon, phosphorus, nitrogen, biological and chemical oxygen demand; physical parameters: pH, redox, conductivity, turbidity, suspended solids, and dissolved oxygen; biological parameters: algae and bacteria; store the historical and real-time ecological data of the wetland in a wetland monitoring database.

[0044] Historical and real-time climate data are obtained. In this embodiment, the climate data includes but is not limited to wetland local climate data and global climate data. The climate data is stored in a climate monitoring database.

[0045] Step 2: Calculate the land use status, dynamic change characteristics of vegetation coverage and leaf area index of estuarine coastal wetlands through data calculations from multi-data source remote sensing image database; specific operations include:

[0046] Among them, the land use status of estuarine coastal wetlands is calculated by using multi-data source remote sensing image data and nautical chart data, including:

[0047] Perform radiation correction and false color synthesis preprocessing on remote sensing image data from multiple data sources to obtain preprocessed images. The preprocessing can greatly improve the recognizability of remote sensing image data from multiple data sources.

[0048] Perform image registration on the preprocessed images;

[0049] Combined with the field survey data, the registered images are divided into regions and the images are digitized into raster images;

[0050] By using GIS software to overlay the digitized raster images from different periods, we analyzed the increased areas between the coastlines in different periods, the reclaimed areas between the two periods, and calculated the areas of beach reclamation in different periods.

[0051] The chart data is digitized and converted to the same projection and coordinate system as the multi-data source remote sensing image data, and then imported into the GIS software for corresponding calculation and analysis, extracting the historical change information of the isobaths, and calculating the erosion and siltation of the beach wetlands in the estuary ecological monitoring area. In accordance with the requirements of maintaining ecological balance, the limit of beach reclamation should not exceed the speed of beach siltation, and reclamation should be based on the dynamic balance of continuous siltation of low- and medium-tide beaches and continuous replenishment of high- and medium-tide beaches.

[0052] Among them, the dynamic change characteristics of vegetation coverage and its leaf area index are calculated by using multi-data source remote sensing image data and nautical chart data of estuarine coastal wetlands, including:

[0053] The vegetation index is calculated using multi-data source remote sensing image data and nautical chart data of estuarine coastal wetlands;

[0054] By analyzing the changes in vegetation coverage, the dynamic change characteristics of vegetation coverage and its leaf area index can be obtained.

[0055] Remote sensing image data and nautical chart data from multiple data sources provide rich geographical and ecological information. These data can be processed and analyzed using remote sensing technology to obtain the dynamic characteristics of vegetation coverage. Specifically, the following steps can be taken:

[0056] 1. Vegetation index calculation: By calculating vegetation indices, such as the Normalized Difference Vegetation Index (NDVI) and the Enhanced Vegetation Index (EVI), the coverage and growth of vegetation can be quantitatively evaluated. These indices reflect the luxuriance of vegetation by calculating the reflectance differences of different spectral bands in remote sensing data. For example, NDVI evaluates the greenness of vegetation by the reflectance differences between the near-infrared band and the red band, while EVI adds a blue band to NDVI to enhance vegetation signals and correct the effects of soil background and aerosol scattering.

[0057] 2. Vegetation coverage analysis: By analyzing the changes in vegetation coverage, we can understand the growth, fall, succession, etc. of vegetation. The calculation of vegetation coverage can be achieved through the interpretation and classification of remote sensing images. Using professional remote sensing knowledge and geographic information system (GIS) technology, the images are classified and interpreted to obtain vegetation coverage information.

[0058] 3. Leaf Area Index (LAI) estimation: LAI is an important parameter to measure the ratio of plant leaf area to land area. It reflects the photosynthetic capacity of plant populations and the productivity of ecosystems. Through the fusion and analysis of multi-source remote sensing data, LAI can be estimated to understand the growth status of vegetation and the health status of ecosystems. By comprehensively utilizing multi-source remote sensing data and surveying and mapping data, more comprehensive and accurate vegetation information can be obtained. Data fusion methods include pixel-level fusion, feature-level fusion, and decision-level fusion, which can improve the accuracy and reliability of evaluation and monitoring results.

[0059] By utilizing remote sensing image data and nautical chart data from multiple data sources, combined with appropriate analysis methods and models, the dynamic change characteristics of vegetation coverage and its leaf area index can be effectively calculated, providing a scientific basis for ecological environmental protection and resource management.

[0060] Step 3: Integrate ecological data, climate data, dynamic change characteristics of estuarine wetland land use and coverage, and spectral feature information to clarify the impact of ecological changes, human activities and climate change on wetland changes.

[0061] Data fusion technology is an information processing technology that uses computers to automatically analyze and synthesize a number of observation information obtained in time series under certain criteria. This technology can enhance the survivability of the system, improve the reliability and robustness of the entire system, enhance the credibility of data, improve accuracy, expand the time and space coverage of the system, and increase the real-time and information utilization of the system. In big data applications, data fusion and integration technology is a crucial link, which can integrate data from different sources and different formats to provide strong support for data analysis, mining and application. After data fusion is completed, the fused data can be analyzed and mined to obtain valuable information and insights, including statistical analysis, machine learning, data visualization and other technologies.

[0062] Data fusion methods include statistical methods, information theory methods, and cognitive model methods. Statistical methods are based on statistical algorithms, such as Bayes estimation, maximum likelihood estimation, Kalman filtering, DS evidence theory, etc.; information theory methods include cluster analysis, voting method, neural network, entropy method, etc.; cognitive model methods involve fuzzy logic, knowledge system, logical template, etc. The selection and application of these methods depends on the specific data fusion needs and scenarios.

[0063] This embodiment establishes a multi-data source remote sensing image database, a wetland monitoring database, and a climate monitoring database, which not only monitors the change data of the wetland itself, but also monitors human activity data and global climate data, and comprehensively monitors the ecological data, climate change, and human activity data that affect wetland changes.

[0064] Embodiment 2:

[0065] like Figure 2 As shown, this embodiment proposes a multi-source land and sea environment information integration system for estuarine coastal wetlands, which mainly includes:

[0066] Multi-data source remote sensing image database, used to store historical and real-time updated multi-data source remote sensing image data and nautical chart data of estuarine coastal wetland ecosystems;

[0067] Wetland monitoring database, used to store ecological data from collection devices;

[0068] Climate monitoring database, used to store climate data;

[0069] The collection device is used to collect ecological data, specifically including a drone equipped with a greenhouse gas detection sensor, a water quality detection sensor, a water level sensor, a spectrometer and a plant canopy analyzer. The drone is provided with a positioning module and a wireless communication module for drone positioning and data transmission respectively. The wireless communication module obtains real-time data from the climate detection data. When the real-time climate data meets the detection requirements, the drone is started to perform ecological data detection according to the GIS planning route. The number of drones increases as the monitoring area increases.

[0070] A communication device for acquiring remote sensing image data, nautical chart data and climate data from multiple data sources. The communication device includes a plurality of communication interfaces for receiving and inputting remote sensing image data, nautical chart data and climate data from multiple data sources.

[0071] The big data technology fusion calculation module, based on GIS database management technology, combines the characteristics of the indicator system and existing data to build a comprehensive database including wetland GIS data and wetland attribute data. Based on the comprehensive database, it uses raster, interpolation, raster calculation, vector data overlay and other methods to extract the geographical factors in the indicator system for integration of various databases and calculated intermediate data. Based on cloud services, a comprehensive analysis of the above data is carried out, and the artificial intelligence algorithm model is used to calculate and output the relationship between the correlation and regularity of the internal data of the wetland ecosystem and the expressiveness and influence of the external data, which facilitates the subsequent analysis of wetland data and provides data support for subsequent wetland protection and management.

[0072] Embodiment 3:

[0073] On the basis of Example 1 and Example 2, the comprehensive management of Spartina alterniflora is achieved through the multi-source land and sea environmental information integration method / system of estuarine coastal wetlands.

[0074] Using remote sensing, geographic information system and MAXENT maximum entropy model1, the spatiotemporal dynamic monitoring and simulation of the distribution evolution of Spartina alterniflora in the coastal wetlands of the Yellow River Estuary were carried out.

[0075] The MAXENT maximum entropy model is a spatial distribution model of species at geographic scales based on the theory of maximum entropy. The MAXENT maximum entropy model was constructed by SJPhillips et al. in 2004. It is currently the best performing and most widely used ecological niche model ENM (ecological niche model). Since it is mainly used to predict species distribution areas, it is also classified as a species distribution model (SDM). The specific steps of using the MAXENT model are: first, obtain the known distribution points of the species (Spartina alterniflora) through field investigations or literature, specimens, and database retrieval. And input the longitude and latitude of the distribution points into the "Samples" module of the MAXENT model in the form of a CSV format file; then, according to the biological characteristics of the species, the environmental layers closely related to the species distribution (climate, meteorology, elevation, vegetation type, slope, soil and other factors affecting the growth and expansion of Spartina alterniflora) are input into the "Environmental layers" module; select 25% of the species data as test data, 75% of the data as training data, and the other parameters are the default settings of the model. Then, the number of repeated operations or the proportion of verification data, whether to draw a response curve to evaluate the accuracy of the model, whether to use the jackknife method (a non-parametric statistical method) to select the dominant environmental factors, etc. are set; finally, the output results are interpreted. Using the above-mentioned "3S" technology and MAXENT maximum entropy model, the spread and invasion forms and laws of Spartina alterniflora in different time and space are studied, the mechanism of Spartina alterniflora invasion of the salt marsh wetland at the mouth of the Yellow River is revealed, and comprehensive management measures for Spartina alterniflora are explored. A technical system for the management of Spartina alterniflora is constructed to provide theoretical and technical support for the ecological protection and restoration of coastal wetlands and the improvement and utilization of saline-alkali land. Thereby achieving the purpose of controlling the spread or removal of Spartina alterniflora.

[0076] Specific steps, such as Figure 4 As shown in the figure: SPOT / VEGETATION NDVI satellite remote sensing data based on continuous time series is collected and collated, and the spatial distribution data of China's annual vegetation index (NDVI) since 1998 is generated by the maximum synthesis method. This dataset effectively reflects the distribution and change of vegetation coverage in various regions across the country on a spatial and temporal scale, and has a very important reference significance for the monitoring of vegetation change, the rational use of vegetation resources and other research in the field of ecological environment.

[0077] The leaf area index and photosynthetically active radiation component data and field measured leaf area index data from 2000 to 2022 were collected and organized (data from the National Earth Observation Data Center). A variety of optimized machine learning algorithms (support vector regression, random forest regression and extreme gradient boosting regression) were used to construct a LAI prediction model for Spartina alterniflora. Remote sensing, geographic information system and other technologies and the MAXENT maximum entropy model (Maximumentropy model1) were used to conduct spatiotemporal dynamic monitoring and simulation studies on the distribution evolution of Spartina alterniflora in estuarine coastal wetlands.

[0078] Embodiment 4:

[0079] Based on the multi-source land and sea environment information integration method for estuarine coastal wetlands proposed in Example 1, further extension is made to integrate different land feature information of representative areas of estuarine coastal wetlands. The main operations include:

[0080] Obtain field spectra of various ground features including Spartina alterniflora;

[0081] Through analysis, the characteristic spectra of different types of land features in representative areas of estuarine coastal wetlands are obtained, which serve as the basis for extracting representative land feature information.

[0082] The present embodiment uses the Field Spec Pro spectrometer of ASD (Analytical Spectra Devices) to obtain the on-site spectra of the main plant types such as Spartina alterniflora; the spectral curve of the sample point objects in the visible light-infrared band range (350nm-2500nm) is collected, and the number of channels is 2151, and the spectral interval is 1nm. The sampling point is located by using the GPS positioning system. In areas where the surveyors cannot reach, the drone technology is used for sampling and measurement.

[0083] Spectral measurement is highly dependent on weather and solar altitude angle. Choose sunny, windless and cloudless weather, and choose the time period of 10:00-14:00 for spectral measurement. During the measurement, the operator should point the probe vertically to the ground object, generally about 1 meter away from the ground object. When the ground object pixels are relatively pure, the probe can be lowered appropriately. If there are many mixed ground types of pixels, the sensor probe can be raised appropriately. Each sample plot selects 6 locations for spectral measurement (i.e. east, south, west, north and their diagonal directions), and 15 spectral curves are taken at each location in the order of whiteboard-ground object-whiteboard and the average is calculated. After the experiment is completed, the average of the spectral curves of these 6 locations is taken as the spectral curve of the sample plot.

[0084] The data measured by the ASD spectrometer cannot be used directly. It needs to be processed by the data processing software View-Spec-Pro that comes with the spectrometer to output a binary file of a valid spectral curve.

[0085] Embodiment 5:

[0086] Based on the multi-source land and sea environmental information integration method for estuarine coastal wetlands proposed in Example 1, a further extension is made to integrate the Leaf Area Index (LAI) of representative areas of coastal wetlands. The main operations include:

[0087] Leaf Area Index (LAI) is an important evaluation index for studying vegetation and is widely used in many fields. Vegetation LAI not only affects the ability of canopy photosynthesis, thus playing a vital role in the carbon cycle, but also plays a key role in the water cycle, affecting biological and physical processes such as soil and surface water evaporation, canopy interception, accumulation and transpiration. Therefore, the LAI of wetland vegetation is a key indicator for measuring the primary productivity, ecological and environmental characteristics and health status of wetland ecosystems.

[0088] This embodiment adopts the plant canopy analyzer LAI-2000. The principle of LAI-2000 is to deduce the number of leaves in the vegetation canopy by measuring the degree of attenuation of light when it passes through the canopy. By measuring this attenuation from different zenith angles, the inclination of the leaves can also be obtained. LAI-2000 simultaneously measures the attenuation of diffuse sky rays at 5 angles. The LAI-2000 optical probe provides the scene of the near hemisphere to 5 detectors located on concentric rings. The sampling point is located using a GPS positioning system.

[0089] According to the vegetation type, growth state and distribution, this embodiment selects typical representative sample points (about 30 are planned) at the estuary and coastal areas. According to the vegetation coverage, the sample plot is specified as 1m1m (average coverage>20%) and 3m3m (average coverage<20%). The sample plot selection follows the principle of regional dominant species. The principle of LAI-2000 is to deduce the number of leaves in the vegetation canopy by measuring the degree of attenuation of light when it passes through the canopy, measure this attenuation value from different zenith angles, and obtain the inclination of the leaves at the same time. LAI-2000 measures at least 10 data: 5 values ​​when the probe is above the canopy, and 5 values ​​when the probe is below the canopy. In these two measurements, the probe is always pointed to the sky. The 5 values ​​of the canopy transmittance are obtained by dividing these corresponding readings by two by two. According to the transmittance of these 5 zenith angles, LAI-2000 can calculate the number of leaves and the inclination of the leaves.

[0090] Embodiment 6:

[0091] Based on the multi-source land and sea environment information integration method for estuarine coastal wetlands proposed in Example 1, further extension is made to organize the integrated information and analyze and invert it, specifically including:

[0092] Collect relevant data and information required for the research, such as relevant literature, time series data of seawater entering the sea, water depth data of estuary coastal areas, remote sensing image data, on-site hydrological observation data, tidal data, wind data, etc., and systematically classify, process and analyze the above data and information;

[0093] The measurement experimental data of the estuary / nearshore, tidal wetland, non-tidal wetland and saline-alkali farmland surface observation fields established by the Estuarine and Coastal Wetland Ecological Experimental Station and the coastal wetland and climate change field control experimental platform are collected to build a conventional monitoring database and basic geographic information database based on estuary and coastal wetlands.

[0094] Collect multi-temporal, multi-resolution, and multi-data source remote sensing images and long-term monitoring data series related to estuarine coastal wetlands, use remote sensing to monitor wetlands, and obtain long-term series of dynamic change characteristics of land use and coverage and spectral characteristic information of estuarine wetlands and surrounding environments. At the same time, use geographic information systems (GIS) to map, monitor, and evaluate estuarine wetlands, study the integration methods of multi-source land and sea environmental information of estuarine coastal wetlands under the influence of high-intensity human activities and global climate change, and develop a wetland ecological information system based on big data fusion.

[0095] Remote sensing technology can provide a large amount of data on wetland resources, which greatly facilitates the investigation of large-scale estuarine and coastal wetland resources. After processing and analysis, its images can also reflect the internal environmental conditions of the wetland. It can conduct in-depth research on the relationship between the estuarine and coastal wetlands and the surrounding climate, topography, land use, vegetation changes, wetland biodiversity, and social and economic development; it can estimate the distribution and area of ​​wetland resources and the distribution, growth status, and spatial pattern evolution of wetland biological resources.

[0096] Dynamic remote sensing monitoring of wetlands uses multi-temporal remote sensing data as the main information source, supplemented by necessary thematic information (geological maps, topographic maps, DEM data, etc.), combined with appropriate field surveys, and comprehensive use of RS, GIS, GPS, Beidou global positioning system and other technologies to monitor the status of wetlands in different periods and the changing characteristics of wetlands within a certain period of time. The organic fusion of radar images and optical images is used to achieve better estimation results. The development and application of spatial information technologies such as RS and GIS provide a variety of data sources and favorable analysis methods for wetland ecological health assessment. As an emerging discipline, GIS has unparalleled advantages over other information systems in dealing with space-related issues. The combination of data management, spatial analysis, output expression functions and evaluation systems based on GIS technology can effectively improve the efficiency and accuracy of the estuarine coastal wetland evaluation system and data visualization effects.

[0097] The collection and processing of the above-mentioned data are now further explained.

[0098] Cloud-free land satellite images acquired by multispectral scanners (MSS), thematic mappers (TM), enhanced thematic mappers (ETM), and operational land imagers (OLI) since the 1990s, hyperspectral remote sensing technology, airborne LiDAR technology, spaceborne LiDAR technology, and microwave radar technology were collected. Topographic maps and wetland use maps of the study area were collected for classification reference. In addition, vegetation maps, tidal channels, roads, shorelines, dikes, and other data, as well as survey point data recorded in field surveys were used as auxiliary data to achieve improved wetland cover classification.

[0099] Separation of water and land, coastline extraction: There is a muddy tidal flat near the estuary, the water is highly turbid and the boundary between water and land is blurred, making it difficult to capture the coastline. In this embodiment, the collected remote sensing images are first radiometrically calibrated, and then the normalized difference water index (NDWI) method is used to calculate the bands, and a suitable threshold is selected for binarization processing to separate the water and land and extract the coastline.

[0100] In order to quickly discover the information of surface changes, high-resolution orthophotos acquired at different times are used for change detection. In view of the fact that high resolution has a great impact on the details of the objects and the spectrum is easily confused, the satellite remote sensing image patch comparison platform software system - point cloud processing software P3C is used to introduce image segmentation algorithms to extract homogeneous objects and use them as the minimum detection unit, and then judge the change area based on the spectral and texture characteristics of the objects. The satellite remote sensing image patch comparison platform software system - point cloud processing software P3C has the following processing functions: point cloud display, point cloud filtering, point cloud clustering, classification by height above the ground, classification by clustering, classification by point class, elevation normalization, correction of misclassified ground points, filling of hill holes, calculation and export of fracture lines, boundary tracking and export, classification expansion based on boundaries, building extraction, external wall point extraction, power line extraction, slender object extraction, road extraction, and batch processing. Software advantages: ① Supports the extraction of vegetation, buildings, power lines, slender features, exterior walls, etc.; ② Complete classification of buildings, extraction rate of more than 90%, and misclassification rate of less than 0.1%; ③ Complete separation of the front and back ends of the software to achieve service-oriented; ④ Flexible processing, support for different pre-classified point clouds; ⑤ With batch processing mode, it can read V7 format DGN files, automatically associate hard disk files, and process multiple blocks at a time; ⑥ High degree of automation and high processing efficiency, the processing speed of 3 million points is less than 1 minute; ⑦ Flexible design, batch processing process can be flexibly handled; ⑧ Use multi-threading technology to ensure that the interface does not freeze during complex processing, and other operations can be performed at the same time; ⑨ Use thread locks to ensure data security.

[0101] The estuarine coastal wetland cover classification system was established with reference to the Chinese wetland cover classification system and taking into account the cover types of the study area. The classification of wetland use types was carried out through the relevant interpretation of land satellite images. In order to accurately extract the wetland reclamation area, the land use boundary in the 1990s was used as the baseline data, and the area changes of estuarine coastal wetlands at each stage compared with the basic area were divided into reclamation areas and natural wetlands.

[0102] Taking the different vegetations of the Yellow River coastal wetlands as the research object, from the perspective of the practical application of remote sensing technology in ecology, starting from three aspects: the extraction of wetland characteristic parameters from multi-source remote sensing data, the extraction of different wetland vegetation, and the construction of inversion models for different wetland vegetation, the airborne LiDAR data and hyperspectral images are integrated to extract the characteristics of wetland vegetation in a more detailed and three-dimensional way, which effectively improves the accuracy of the inversion of wetland-related vegetation indices.

[0103] The field survey point data were used, part of which was used as auxiliary data useful for interpretation, and the remaining point data was used to verify the accuracy of the classification results. The Kappa coefficient (consistency test index, which can also be used to measure classification accuracy) and overall accuracy of the confusion matrix were used to evaluate the consistency between the results and the basic facts. The extracted wetland cover types were compared with the field survey points to evaluate the accuracy of the wetland cover classification results.

[0104] Embodiment 7:

[0105] Based on the method for integrating multi-source land and sea environmental information of estuarine coastal wetlands proposed in Example 1, further extension is made to perform data analysis and inversion on the integrated information based on big data mining and machine learning technology, specifically including:

[0106] This embodiment uses the big data mining technology of comprehensive statistics, database technology and artificial intelligence technology to find the rules from the large amount of data in the above-mentioned database and information system. Automatically extract patterns, associations, changes, anomalies and meaningful structures from the data, and use big data mining technology to mine and improve the key parameters of inversion and establish corresponding mathematical models. The specific processing method and process are as follows:

[0107] Classification and cluster analysis: Use classification and clustering big data mining and inversion techniques such as object-oriented random forest classification algorithms to achieve high-precision classification of estuarine coastal wetland ecological vegetation systems. Based on the morphological characteristics, structural characteristics, and combination characteristics of estuarine wetland ecosystems, a multi-scale description index system for estuarine coastal wetland systems is constructed to systematically analyze the structural characteristics and spatial differentiation characteristics of wetlands at different scales. Based on multi-source observation data at different temporal and spatial scales, the key elements of necessary ecological processes are selected.

[0108] Analyzing the main ecological and environmental problems in the region and their spatial heterogeneity and delineating wetland ecological functional zones are of great guiding significance for ensuring regional ecological security and optimizing wetland development. Based on ecosystem services and ecological sensitivity, a regional ecological functional zoning index system is constructed, and the self-organizing feature map (SOFM) neural network and support vector machine (SVM) machine learning algorithm are coupled to delineate the ecological functional zones of estuarine coastal wetlands.

[0109] The purpose of ecological function zoning of estuarine coastal wetlands is to identify important areas of regional ecological security and clarify the characteristics of various types of ecosystems in the region. Zoning needs to reflect both the ability of ecosystems to provide services to humans directly or indirectly and the possibility of ecological and environmental problems in the region. Taking into account the natural ecological characteristics of the study area, the rationality of indicators and the timeliness of data, this embodiment study intends to select two categories of indicators, regional key ecosystem services and ecological sensitivity, to construct an ecological function zoning indicator system.

[0110] This embodiment uses a self-organizing feature mapping network to divide ecological function types based on pixel clustering results. The SOFM network was proposed by Finnish scholar Kohonen. It is a network model that maps high-dimensional data sets to low-dimensional space and obtains similarity relationships between data. The SOFM network consists of two parts: an input layer and a competition layer. The dimension of the input vector is the same as the dimension of each node in the competition layer. After the data is input, the nodes in the competition layer compete with each other, and the weights of the winning nodes and neighboring nodes are adjusted. The data clustering results can be obtained while keeping the topological structure of the data set unchanged.

[0111] This embodiment uses a support vector machine to extract the boundaries of ecological functional zones. Support vector machine is a supervised learning algorithm based on machine learning for data analysis in classification, regression and outlier detection. Intuitively, the farther the classification boundary is from the nearest training data point, the greater the confidence of the classification, and the support vector machine achieves classification by constructing an optimal hyperplane or set of hyperplanes that separate the two classes with the maximum interval in a high-dimensional or infinite-dimensional space. In theory, the support vector machine can achieve the optimal classification of data and obtain the global optimal solution.

[0112] The research steps of this embodiment can be roughly divided into three steps: ① Calculate the zoning indicator data set through ArcMap and other software and normalize the data to eliminate the influence caused by unit and measurement differences between indicators; ② Use a self-organizing feature mapping network to cluster the zoning units based on the similarity of indicator characteristics, and construct a classification effect index and determine the optimal clustering scheme based on the consistency principle of geographical background elements; ③ Cross-select the optimal parameters of the support vector machine and perform model training, thereby finding the best ecological functional zone zoning boundaries between broken patches to obtain the final zoning results.

[0113] Different vegetation index inversion processes are constructed using different algorithms of machine learning regression models (random forest regression model and support vector regression model, etc.), and the fitting effect of the regression model is quantitatively analyzed and compared by calculating the model accuracy evaluation index. The inversion is analyzed from three perspectives: different data sources, different vegetation, and different modeling methods. The distribution of vegetation and other characteristic quantities of estuarine coastal wetland ecosystems is analyzed from different levels such as remote sensing, ecology, and economy.

[0114] Using the above-mentioned association rules, neural networks and other big data analysis methods, we will develop a wetland ecological information system based on big data fusion to clarify the relationship between the correlation and regularity of the internal data of the ecosystem and the expressiveness and influence of the external data. We will focus on studying the composition, structure, functional evolution process and its regulatory mechanism of estuarine coastal wetland ecosystems; data analysis will study the relationship between wetland ecological processes and material fluxes at the land-sea interface.

[0115] By using land cover transfer matrix, area weight centroid analysis and remote sensing images to invert and reveal the spatiotemporal pattern of estuarine coastal wetlands and reclamation, the movement trend of reclamation and coastal wetlands is identified. The spatiotemporal dynamics of estuarine coastal wetlands and reclamation are quantified, and the driving factors of these dynamics are determined using big data mining and analysis techniques. Based on the above research, a scientific basis is provided for the formulation of proactive protection strategies to ensure ecosystem structure and services and regional ecological security.

[0116] In order to more accurately simulate the wave characteristics of plant areas, based on the above data measurement, information analysis and remote sensing inversion, combined with the research results of relevant literature, the main vegetation types, plant density, diameter and other parameters of estuarine wetland distribution are calibrated and inverted to study the distribution law of wetland vegetation. This embodiment uses Landsat-8OLI and TM remote sensing image data, and pre-processes them through the ERDASIMAGE software platform, including geometric correction and atmospheric correction. The pre-processed images are classified and inverted to provide input data for the coupled ecological geomorphological model, mainly including surface landscape and vegetation types interpreted by remote sensing and vegetation biomass data obtained by remote sensing inversion.

[0117] Combined with ground survey data, remote sensing classification training samples were constructed, and supervised classification methods were used to classify wetland features. The classification data was used as the input and verification data of the model. The bottom friction coefficient of the study area was highly correlated with the feature coverage type, and empirical parameters were used to obtain the bottom friction coefficient of different feature coverage types.

[0118] In the vegetation-covered area, the growth state of vegetation has an important influence on the bottom friction of the model. The study calculated the NDVI information of the salt marsh vegetation area and calculated the biomass information of the vegetation through remote sensing inversion method. By calculating the range of vegetation biomass in different seasons, a linear regression model was used to determine the spatial distribution of the bottom friction coefficient of the salt marsh wetland. The data such as the coverage range, vegetation species and vegetation density of the wetland vegetation were extracted and then substituted into the vegetation energy reduction module. The wave energy dissipation term of the original SCHISM model was improved, and the drag attenuation dissipation and turbulence attenuation dissipation caused by plants were inverted and calculated.

[0119] Embodiment 8:

[0120] Based on the multi-source land and sea environmental information integration method for estuarine coastal wetlands proposed in Example 1, a further extension is made to study the law of estuarine sediment transport and the response mechanism of estuarine wetland topography and geomorphic evolution under the background of sea level rise. The specific operations include:

[0121] (1) Main data processing and analysis methods

[0122] ①Main analysis methods of data time series

[0123] The nonparametric Mann-Kendall trend mutation test method was used to test the trend and mutation of water and sediment. Trend analysis, wavelet analysis, cross-spectral analysis and other methods were used to diagnose the spectrum analysis and its internal response relationship and construct the model.

[0124] ②Underwater topography calculation data

[0125] The Dem analysis system built by the research team is used to analyze topographic data. The Dem analysis system uses the WebGIS system as a carrier to analyze and calculate water depth data. The graphic data generated by the analysis process and results are managed and displayed by WebGIS. It mainly provides the following functions: measurement section and arbitrary section analysis; isobath analysis, overlay, isobath enclosed area calculation, dynamic evolution of isobaths of multi-period data; scouring and silting analysis, scouring and silting comparison, regional analysis and calculation of waterway dredging engineering quantities, etc.

[0126] ③ Empirical orthogonal function analysis method

[0127] Empirical orthogonal function decomposition (EOF), also known as empirical orthogonal function decomposition or eigenvector analysis, can be used to analyze the structural characteristics of matrix data and is a method for extracting the main characteristic quantities of the matrix. The EOF analysis method can decompose the variable field that changes with time into a spatial function that does not change with time and a time function that does not change with space to achieve the purpose of dimensionality reduction, so as to analyze the main influencing factors of the field and the main period of its change. The study of the law of change of the main components with time can replace the study of the law of change of the field with time, thus facilitating the analysis of the mechanism of the spatiotemporal change of the field quantity.

[0128] (2) Computational and analytical research

[0129] Based on the above-mentioned water, sand, topography and landform measurement and collection data, combined with remote sensing inversion and numerical simulation calculation results, using the above-mentioned methods, we carried out:

[0130] ① Study the interaction mechanism and response process between estuarine wetland geomorphology, vegetation and other elements and tidal waves, currents, and wind-wave dynamics.

[0131] ② Analyze the changing patterns and development trends of water and sediment in the upper reaches of the estuary; establish the relationship between river water and sediment and erosion and siltation of estuarine shallow wetlands; analyze and study the role of changes in sediment entering the sea in the evolution of erosion and siltation of estuarine shallow wetlands.

[0132] ③ Based on the relevant theories of estuarine water and sediment transport and estuarine coastal evolution dynamics, the adjustment and development process and key spatiotemporal conversion nodes of the equilibrium and non-equilibrium states of the estuarine coastal wetland evolution dynamic system are systematically studied, and the research reveals the interdecadal adjustment and conversion process of the estuarine coastal wetland system state and the controlling dynamic coupling mechanism.

[0133] Embodiment 9:

[0134] Based on the multi-source land and sea environmental information integration method for estuarine coastal wetlands proposed in Example 1, a data information processing system is further extended to construct the data information processing system. The data information processing system has the functions shown in Table 1, and is mainly divided into two parts: GIS and DEM analysis. The two modules are organically combined, wherein the GIS part plays the role of a graphic carrier, providing graphic display, roaming, positioning, data editing, data query, graphic printing and other functions for basic geographic data and water depth measurement data, and the DEM part completes the analysis and calculation functions of water depth data.

[0135] Table 1 System function list

[0136]

[0137]

[0138] The above embodiments take estuarine coastal wetlands as the research objects, and focus on the simulation expression of the physical processes of estuarine coastal wetland ecosystems in terms of research ideas. They are carried out in sequence and step by step around "information integration and big data mining analysis and inversion of estuarine coastal wetland ecosystems", "research on spatiotemporal differentiation and evolution process of multi-scale estuarine coastal wetland morphology", "research on state adjustment and conversion characteristics of estuarine coastal wetland systems", "research on interactive feedback mechanisms of coastal wetland interaction coupling systems", and "theory of ecological conservation and restoration of estuarine coastal wetlands and protection and regulation technology systems". In the organization and implementation of the research work, remote sensing analysis, field observation, numerical simulation and mathematical modeling are organically combined. The previous research foundation in multi-source remote sensing image fusion inversion technology, comprehensive statistics, database technology and artificial intelligence technology, big data mining technology, the construction of atmosphere-ocean-wave coupled assimilation model, and the "basin-estuary wetland-nearshore waters" large-scale comprehensive numerical model are fully utilized. With the help of the research team's accumulation in massive remote sensing data processing and technical advantages in big data mining, the key links of the research plan and the technical difficulties that may be encountered in the research process are fully considered, and efforts are made to clarify the goals and specific work plans, so as to ensure that the overall research plan and technical route are reasonable and feasible.

Claims

1. A method for integrating multi-source land and sea environmental information of estuarine coastal wetlands, characterized by: The following steps are involved: Step 1: Obtain remote sensing image data, nautical chart data, ecological data and climate data from multiple sources of estuarine coastal wetlands; Step 2: Using the multi-data source remote sensing image data and nautical chart data of the estuarine coastal wetlands, calculate the land use status, dynamic change characteristics of vegetation coverage and leaf area index of the estuarine coastal wetlands; Step 3: Integrate ecological data, climate data, land use conditions, dynamic characteristics of vegetation cover, and leaf area index to obtain multi-source land and sea environmental information of estuarine coastal wetlands.

2. The method for integrating multi-source land and sea environmental information of estuarine coastal wetlands according to claim 1 is characterized by: The ecological data include: greenhouse gas, water quality detection data, water level data, leaf spectrum data and leaf area index of plant types at sampling points; The climate data include: climate data of estuarine coastal wetland ecosystems and global climate data.

3. The method for integrating multi-source land and sea environmental information of estuarine coastal wetlands according to claim 2 is characterized by: The greenhouse gases include methane, carbon dioxide and nitrous oxide; the water quality detection data include carbon content data, phosphorus content data, nitrogen content data, biological content data, chemical oxygen demand, pH value, total oxygen demand, conductivity, turbidity, suspended solids and dissolved oxygen; the biological content data include: algae content data and bacterial content data.

4. The method for integrating multi-source land and sea environmental information of estuarine coastal wetlands according to claim 1 is characterized by: The land use conditions of the estuarine coastal wetlands include: the area of ​​the reclaimed area and the conditions of erosion and siltation; The land use status of the estuarine coastal wetlands is calculated by using the multi-data source remote sensing image data and nautical chart data of the estuarine coastal wetlands. The specific operations include: Perform radiation correction and false color synthesis processing on remote sensing image data from multiple data sources to obtain processed remote sensing images; Performing image registration on the processed remote sensing image to obtain a registered remote sensing image; Combined with the field exploration data, the registered remote sensing image is divided into regions, and the remote sensing image is digitized into raster images to obtain a remote sensing image with digitized raster images; By using GIS software to overlay the digitized images of raster images from different periods, we can analyze the increased areas between the shorelines in different periods and calculate the areas of reclamation in different periods. The isobaths of the chart data are digitized and converted into different formats to obtain the processed chart data; the processed chart data are imported into the GIS software for difference calculation and comparative analysis to extract the isobaths of previous years; based on the isobaths of previous years, the erosion and siltation conditions of the estuarine coastal wetlands are obtained.

5. The method for integrating multi-source land and sea environmental information of estuarine coastal wetlands according to claim 1 is characterized by: The above-mentioned use of multi-data source remote sensing image data and nautical chart data of estuarine coastal wetlands to calculate the dynamic change characteristics of vegetation coverage and its leaf area index includes: The vegetation index is calculated using multi-data source remote sensing image data and nautical chart data of estuarine coastal wetlands; By analyzing the changes in vegetation coverage, the dynamic change characteristics of vegetation coverage and its leaf area index can be obtained.

6. The method for integrating multi-source land and sea environmental information of estuarine coastal wetlands according to claim 5 is characterized by: The above-mentioned use of multi-data source remote sensing image data and nautical chart data of estuarine coastal wetlands to calculate the vegetation index specifically includes: The vegetation index includes the normalized difference vegetation index and the enhanced vegetation index; By calculating the reflectance difference between the near-infrared band and the red light band in the multi-data source remote sensing image data of the estuarine coastal wetland, the normalized difference vegetation index reflecting the luxuriance of vegetation is obtained. By calculating the reflectance differences among near-infrared bands, red bands, and blue bands in multi-data source remote sensing image data of estuarine coastal wetlands, the enhanced vegetation index for assessing the greenness of vegetation is obtained.

7. The method for integrating multi-source land and sea environmental information of estuarine coastal wetlands according to claim 5 is characterized by: The dynamic change characteristics of vegetation coverage and its leaf area index are obtained by analyzing the change of vegetation coverage, and the specific operations include: Using remote sensing knowledge and geographic information system technology, we classified and interpreted the multi-source remote sensing image data of estuarine coastal wetlands to obtain vegetation coverage information and then the changes in vegetation coverage. The leaf area index, which is used to measure the ratio of plant leaf area to land area, was estimated by fusing and analyzing remote sensing image data from multiple sources of estuarine coastal wetlands.

8. The method for integrating multi-source land and sea environmental information of estuarine coastal wetlands according to claim 7 is characterized by: The fusing of the multi-data source remote sensing image data of the estuarine coastal wetland includes: fusing the multi-data source remote sensing image data of the estuarine coastal wetland at the pixel level, at the feature level or at the decision level.

9. The method for integrating multi-source land and sea environmental information of estuarine coastal wetlands according to claim 1, characterized in that: In step 3, statistical methods, information theory methods or cognitive model methods are used to integrate ecological data, climate data, land use conditions, dynamic changes in vegetation cover and leaf area index to obtain multi-source land and sea environmental information of estuarine coastal wetlands.

Images