A wetland partitioning method based on surface water-groundwater connecting space heterogeneity

Abstract

The application discloses a wetland partitioning method based on surface water-groundwater connecting space heterogeneity, and comprises the following steps: surface water connecting partitioning based on hydrological barrier identification; wetland groundwater connecting partitioning based on stable isotopes; and groundwater connecting partitioning verification based on eDNA. The application creatively proposes a wetland partitioning method considering surface water-groundwater exchange, combines remote sensing images, eDNA technology, stable isotope analysis and other technologies, comprehensively analyzes the hydrological connectivity of the wetland, fully considers the interaction between the groundwater and the surface water, identifies the hydrological connectivity area in the wetland, and makes up for the limitation that the existing method only relies on the surface water to divide the wetland partitioning. Through the accurate wetland partitioning, water resource allocation can be optimized, water resource waste can be avoided, and the effect and sustainability of the ecological water supplement of the wetland can be improved.

Description

Technical Field

[0001] This invention relates to the field of wetland zoning technology, and in particular to a wetland zoning method based on the spatial heterogeneity of surface water-groundwater connectivity. Background Technology

[0002] Wetlands, as ecological barriers in the water-land transition zone, play an irreplaceable role in maintaining regional hydrological balance, protecting biodiversity, and regulating carbon and nitrogen cycles. Since 1970, due to water scarcity and disruption of hydrological connectivity, the global wetland area has shrunk by 35%, and more than half of wetlands face the threat of ecological degradation. Global wetland degradation over the past 70 years has directly resulted in greenhouse gas emissions equivalent to 276.4 Gt of carbon dioxide, equivalent to the emissions from burning 110 billion tons of standard coal. To restore wetland hydrological rhythms and ecological functions, ecological water replenishment has become a hot topic in wetland scientific research and is widely applied in practice.

[0003] Current research on wetland ecological water replenishment focuses on ensuring wetland ecological water demand, restoring hydrological connectivity, and refining the spatial and temporal allocation of water resources. Globally, scholars have explored the benefit assessment, ecological response mechanisms, and water replenishment management strategies for different types of wetlands (such as riverine wetlands, lacustrine wetlands, and marsh wetlands). As research deepens, more and more studies are focusing on the regional differences and precise regulation of wetland ecological water replenishment, proposing zoning-based water replenishment strategies. Subsequently, the selection of water sources, water volume regulation, and ecological threshold identification for zoning water replenishment have become research priorities. These studies generally quantify the key role of zoning water replenishment in improving management efficiency and ecological benefits by evaluating the ecological water replenishment responses of different functional zones.

[0004] However, current wetland ecological water replenishment zoning methods mainly rely on surface water processes. Based on hydrological processes, hydrochemical characteristics, remote sensing monitoring, and hydrodynamic simulation, scholars have constructed various surface water zoning models to guide precise water replenishment. Due to the complex internal topography of wetlands, there are significant non-permanent hydrological barriers between different areas. During the dry season, these areas are distributed as independent "point-like" water bodies, gradually connecting to form large water surfaces during the wet season. However, some areas have formed permanent hydrological barriers due to topography or human activities, making it difficult to achieve water surface connectivity through water replenishment. This provides a theoretical basis for wetland water replenishment zoning. However, current wetland zoning research only focuses on surface connectivity, neglecting the role of groundwater connectivity in regulating wetland water levels. In fact, seemingly isolated water bodies often achieve hydrological connectivity through groundwater-surface water exchange. Groundwater is not only an important hidden source of wetland replenishment but also a key factor in regulating wetland hydrological connectivity. The lack of in-depth understanding of groundwater replenishment mechanisms and their coupling relationship with surface water has become a significant problem restricting the scientific management of wetland ecological water replenishment. Therefore, it is urgent to establish a zoning method that covers the synergistic effects of surface water and groundwater, accurately identify the water replenishment relationships of wetland bodies, and provide a scientific basis for ecological water replenishment.

[0005] In summary, it is necessary to design a wetland zoning method based on the spatial heterogeneity of surface water-groundwater connectivity to solve the above problems. Summary of the Invention

[0006] Existing wetland zoning technologies mainly focus on the direct connectivity of surface water, dividing wetlands based on whether the water surfaces are connected. In reality, there are complex interactions between surface water and groundwater within wetland systems, and isolated ponds and marshes can also replenish each other through groundwater flow. Therefore, when zoning wetlands, the connectivity of both surface water and groundwater must be taken into account.

[0007] To address the above problems, this invention proposes a novel wetland zoning method based on the spatial heterogeneity of surface water-groundwater connectivity, comprising the following steps:

[0008] S1. Surface water connectivity zones based on hydrological barrier identification, including:

[0009] S1.1 Remote Sensing Image Acquisition and Preprocessing:

[0010] Multi-temporal remote sensing images were acquired through a remote sensing information platform. Geometric and radiometric corrections were performed on the remote sensing images, and NDWI and random forest algorithms were used to improve the accuracy of water body extraction.

[0011] S1.2 Hydrological barrier identification:

[0012] By combining remote sensing inversion and GIS analysis, we can identify changes in wetland water bodies and the distribution of artificial hydrological barriers.

[0013] S1.3, Construction of Surface Water Connectivity Zones:

[0014] Based on the identification of hydrological barriers, the wetland is divided into multiple patches with clear internal connectivity, forming surface water connectivity zones;

[0015] S2. Wetland groundwater connectivity zones based on stable isotopes, including:

[0016] S2.1 Identification of Groundwater Connectivity Mechanism:

[0017] Based on the surface water connectivity zone, using hydrogen and oxygen stable isotopes (δ²H, δ¹) 8 O) Determine whether groundwater is connected; if the isotope values ​​of adjacent patches that are blocked by the surface are close to each other, it is determined that there is common recharge or groundwater exchange, and they are classified as the same underground connected unit; if the difference is significant, it is considered that the underground connection is weak or non-existent.

[0018] S2.2 Isotope Cluster Analysis:

[0019] Multiple clustering methods were used to perform cluster analysis on hydrogen and oxygen stable isotope data from multiple surface water sampling points in wetlands, and the optimal clustering method was determined by combining evaluation indicators.

[0020] S2.3, Construction of groundwater connectivity zones:

[0021] Based on the clustering results, patches with similar isotopic characteristics and spatial proximity are merged into the same partition, thereby integrating a functional zoning framework that reflects groundwater connectivity.

[0022] S3. Validation of groundwater connectivity zones based on eDNA:

[0023] The Fauna Similarity Index (AFR) was used to quantify the biological connectivity between sampling points in different groundwater connectivity zones.

[0024] In a preferred embodiment of the present invention, the remote sensing image acquisition method in S1.1 is as follows: the water inundation range is extracted based on Landsat-8 satellite remote sensing image data, and annual image data of the past five years of high water season are collected, covering the visible light, near infrared and thermal infrared bands, with a spatial resolution of 30 meters.

[0025] Furthermore, S1.1 uses ENVI remote sensing image processing software for radiometric calibration and atmospheric correction.

[0026] Furthermore, in S1.2, the hydrological barrier identification method uses a random forest supervised classification algorithm to construct a training set and optimize the classification model using manually labeled wetland sample data from the most recent year.

[0027] Based on the identification of long-term hydrological barriers, artificial facilities that still maintain their barrier function during the high-water season were further screened by overlaying multiple remote sensing images and field survey data. These artificial facilities include flood control dikes and main roads.

[0028] Based on the spatial distribution of hydrological barrier zones and artificial barrier facilities, areas that still maintain significant water body barrier function during the high-water season were selected, and wetlands were divided into multiple independent patches based on their spatial distribution characteristics.

[0029] In a preferred embodiment of the present invention, the cluster analysis in S2.2 includes:

[0030] S2.2.1 Obtaining the optimal number of clusters:

[0031] The optimal number of clusters is determined by calculating the sum of squared errors and the changes in silhouette coefficients under different numbers of clusters.

[0032] S2.2.2 Selecting the optimal clustering method:

[0033] The algorithm performance was evaluated by combining the silhouette coefficient, Calinski-Harabasz index, Davies-Bouldin index, and BIC / AIC, and the optimal clustering method was selected for clustering.

[0034] Furthermore, S2.3 specifically involves: spatially aggregating the already divided surface connectivity patches based on the hydrogen and oxygen isotope clustering results; prioritizing the merging of patches with similar isotope characteristics and spatial proximity into the same partition during the integration process; and forming a functional zoning framework with multiple partitions that reflects groundwater connectivity after integration.

[0035] In a preferred embodiment of the present invention, the quantification standard in S3 is as follows: when the AFR value of adjacent patches is greater than 60%, it indicates that there is a hydraulic connection between the two. If surface hydrological observation shows that there is a barrier between the patches, it can be inferred that the groundwater system is the key channel for maintaining its hydrological connectivity. On this basis, the dynamic changes in water level of the two patches are further compared. If there is a significant synchronicity, it can be used to prove that cross-patch water replenishment is achieved through groundwater exchange.

[0036] Implementing this invention has the following beneficial effects:

[0037] This invention innovatively proposes a wetland zoning method that considers surface water-groundwater exchange. Combining remote sensing imagery, eDNA technology, stable isotope analysis, and other technologies, it comprehensively analyzes the hydrological connectivity of wetlands, fully considers the interaction between groundwater and surface water, and identifies hydrological connectivity areas within wetlands. This overcomes the limitations of existing methods that rely solely on surface water for wetland zoning. Through precise wetland zoning, water resource allocation can be optimized, water waste can be avoided, and the effectiveness and sustainability of wetland ecological water replenishment can be improved. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 Map showing the changes in wetland inundation area from 2019 to 2024;

[0040] Figure 2 A wetland patch zoning map based on hydrological barriers;

[0041] Figure 3 This is a diagram illustrating the sum of errors and changes in the contours of the washing process.

[0042] Figure 4 This is a diagram showing the locations of the nine partitions;

[0043] Figure 5 The degree of species sharing in each region. Detailed Implementation

[0044] The technical solutions of this invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0045] Taking Momoge Wetland as an example, Momoge Wetland (45°42′25″-46°18′0″N, 123°27′00″-124°04′33.7″E) is located at the confluence of the Nenjiang River and the Taoer River in the Songnen Plain of Northeast China (Li. et al, 2025), with a total area of ​​1,440 km².

[0046] This invention provides a wetland zoning method based on the spatial heterogeneity of surface water-groundwater connectivity, comprising the following steps:

[0047] S1. Remote sensing image interpretation and hydrological barrier identification, including:

[0048] S1.1 Remote sensing image acquisition:

[0049] The extent of water inundation was extracted using Landsat-8 satellite remote sensing imagery. The data was obtained from the U.S. Geological Survey (USGS) EarthExplorer platform (https: / / earthexplorer.usgs.gov / ). Annual imagery data for the high-water season from 2019 to 2024 was collected, covering the visible, near-infrared, and thermal infrared bands, with a spatial resolution of 30 meters. Please see [link / reference]. Figure 1 , Figure 1 Map showing the changes in wetland inundation extent from 2019 to 2024.

[0050] S1.2 Data Preprocessing:

[0051] Radiometric calibration and atmospheric correction were performed using ENVI remote sensing image processing software.

[0052] The specific steps are as follows: Open the Landsat 8 header file to display the unprocessed true-color image. Open the Radiometric Calibration tool in RadiometricCorrection and select multispectral data. Select Radiance as the calibration type, BIL as the output format, Float as the output data type, and 0.1 as the coefficient. Open the RadiometricCorrection—Atmospheric Correction Module—QUAC-Quick AtmosphericCorrection tool, import the radiometrically calibrated data, and obtain the atmospherically corrected image.

[0053] S1.3 Hydrological barrier identification:

[0054] Traditional NDWI thresholding methods struggle to effectively distinguish between open water bodies and vegetated water bodies (such as reed marshes), where vegetation significantly reduces surface reflectivity. Therefore, this invention utilizes a Random Forest supervised classification algorithm, constructing a training set using manually labeled wetland sample data from 2024 to optimize the classification model.

[0055] Based on the identification of long-term hydrological barriers, by overlaying multiple remote sensing images with field survey data, artificial facilities that still maintain their barrier function during the high-water season, such as flood control dikes and main roads, were further screened out.

[0056] Based on the spatial distribution of both hydrological barriers and artificial barriers, areas that maintain significant water barrier function during the high-water season were selected. These areas were then divided into 22 independent patches based on their spatial distribution characteristics. (See also: [link to relevant documentation]). Figure 2 , Figure 2 This is a wetland patch zoning map based on hydrological barriers.

[0057] The delineation results show that: the Nenjiang floodplain on the east side of the wetland has no long-term barriers and is divided into one patch except for the Yue Liang Pao Reservoir; the western and southern areas have a higher density of permanent barriers and obvious patch fragmentation characteristics; the core area of ​​Baihe Lake consists of six ponds and marshes and is decomposed into six patches.

[0058] Analysis of hydrological barriers from 2017 to 2023 reveals significant changes in the distribution and connectivity of wetland water bodies, particularly during the high-water season when the water bodies become contiguous. Man-made structures (such as dams, sluices, and roads) have created permanent barriers in wetlands, severely impacting hydrological connectivity. The hydrological connectivity of wetlands exhibits dynamic changes under different hydrological conditions, especially during seasons with abundant rainfall, when wetland water bodies gradually become contiguous.

[0059] S2. Wetland groundwater connectivity zone based on stable isotopes:

[0060] Stabilizing isotopes with hydrogen and oxygen (δ²H, δ¹) 8 O) Determine whether groundwater is connected; if the isotope values ​​of adjacent patches that are blocked by the surface are close to each other, it is determined that there is common recharge or groundwater exchange, and they are classified as the same underground connected unit; if the difference is significant, it is considered that the underground connection is weak or non-existent.

[0061] Using multiple clustering methods, including K-means clustering, Mean Shift density clustering, spectral clustering (SC), and Gaussian mixture model (GMM), the stable hydrogen and oxygen isotopes (δ²H, δ¹H, δ¹⁰H) of 25 wetland surface water sampling points (S1–S25) were analyzed. 8 O) Perform cluster analysis on the data.

[0062] S2.1 Obtaining the optimal number of clusters:

[0063] The optimal number of clusters was determined by calculating the sum of squared errors and the variation in silhouette coefficients under different numbers of clusters, and was found to be 4. (See also...) Figure 3 , Figure 3 This is a diagram showing the sum of errors and the changes in the contours of the face during washing.

[0064] S2.2 Selecting the optimal clustering method:

[0065] The algorithm performance was evaluated by combining the silhouette coefficient (intra-class compactness), Calinski-Harabasz index (between-class separation), Davies-Bouldin index, and BIC / AIC (GMM only), and the optimal clustering method was selected for clustering. Considering all indicators, K-means and Spectral Clustering algorithms can be considered the best performing clustering methods overall when k=4, as shown in Table 1.

[0066] Table 1 shows the performance of each algorithm.

[0067] K-means 4 0.666 145.57 0.417 / Good performance K-means 5 0.609 138.58 0.503 Mean Shift 5 0.504 98.29 0.418 / Spectral Clustering 4 0.666 145.57 0.417 / Good performance Spectral Clustering 5 0.595 215.52 0.46 / / GMM 4 0.518 81.76 0.518 2.07 / GMM 5 0.609 138.58 0.503 -120.8

[0068] S2.3, Integrate functional partitions:

[0069] Based on hydrogen and oxygen isotope clustering results, spatial aggregation was performed on the already divided surface connectivity patches. During the integration process, patches with similar isotopic characteristics and spatial adjacency were preferentially merged into the same partition. The resulting integration formed a functional zoning framework with nine partitions that reflects groundwater connectivity. Please refer to [link to relevant documentation]. Figure 4 , Figure 4 This is a schematic diagram of the nine partition locations.

[0070] The wetland zoning optimization method based on stable isotope analysis significantly improves the accuracy of water resource management. By accurately identifying the sources of water replenishment within wetlands, the zoning water replenishment strategy effectively avoids excessive or insufficient water replenishment, further enhancing the stability of wetland ecosystems.

[0071] S3. Validation of groundwater connectivity zones based on eDNA:

[0072] The faunal similarity index (AFR) was used to quantify the biological connectivity between two sampling points: when the AFR value of adjacent patches was >60%, it indicated that there was a hydraulic connection between them. If surface hydrological observations showed barriers between the patches (i.e., a lack of surface runoff connectivity), it could be inferred that the groundwater system was the key channel maintaining their hydrological connectivity. Based on this, further comparison of the water level dynamics of the two patches showed significant synchronicity, which corroborated the fact that cross-patch water replenishment was achieved through groundwater exchange. Please see [link to relevant documentation]. Figure 5 , Figure 5 The degree of species sharing in each region.

[0073] Analysis shows that areas with similar water levels generally exhibit strong biological connectivity, but short-term hydrological changes (such as precipitation and climate fluctuations) can lead to differences in species distribution. For example, in areas with similar water levels (such as some rivers and marshes), there is a high degree of species overlap, indicating that hydrological connectivity between water bodies provides a physical basis for species exchange. However, short-term fluctuations in hydrological conditions can still cause changes in biological connectivity between water bodies, especially when occult hydrological connectivity (such as groundwater runoff and lateral seepage) is present, in which species exchange remains active.

[0074] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A wetland zoning method based on the spatial heterogeneity of surface water-groundwater connectivity, characterized in that, Includes the following steps: S1. Surface water connectivity zones based on hydrological barrier identification, including: S1.1 Remote Sensing Image Acquisition and Preprocessing: Multi-temporal remote sensing images were acquired through a remote sensing information platform. Geometric and radiometric corrections were performed on the remote sensing images, and NDWI and random forest algorithms were used to improve the accuracy of water body extraction. S1.2 Hydrological barrier identification: By combining remote sensing inversion and GIS analysis, we can identify changes in wetland water bodies and the distribution of artificial hydrological barriers. S1.3, Construction of Surface Water Connectivity Zones: Based on the identification of hydrological barriers, the wetland is divided into multiple patches with clear internal connectivity, forming surface water connectivity zones; S2. Wetland groundwater connectivity zones based on stable isotopes, including: S2.1 Identification of Groundwater Connectivity Mechanism: Based on the surface water connectivity zoning, hydrogen and oxygen stable isotopes are used to determine whether groundwater is connected; if the isotope values ​​of adjacent patches that are blocked by the surface are close to each other, it is determined that there is common recharge or groundwater exchange, and they are classified as the same underground connectivity unit; if the difference is significant, it is considered that the underground connectivity is weak or non-existent. S2.2 Isotope Cluster Analysis: Multiple clustering methods were used to perform cluster analysis on hydrogen and oxygen stable isotope data from multiple surface water sampling points in wetlands, and the optimal clustering method was determined by combining evaluation indicators. S2.3, Construction of groundwater connectivity zones: Based on the clustering results, patches with similar isotopic characteristics and spatial proximity are merged into the same partition, thereby integrating a functional zoning framework that reflects groundwater connectivity. S3. Validation of groundwater connectivity zones based on eDNA: The Fauna Similarity Index (AFR) was used to quantify the biological connectivity between sampling points in different groundwater connectivity zones.

2. The wetland zoning method based on the spatial heterogeneity of surface water-groundwater connectivity according to claim 1, characterized in that, The remote sensing image acquisition method in S1.1 is as follows: the water inundation range is extracted based on Landsat-8 satellite remote sensing image data, and annual image data of the past five years of high water season are collected, covering the visible light, near infrared and thermal infrared bands, with a spatial resolution of 30 meters.

3. The wetland zoning method based on the spatial heterogeneity of surface water-groundwater connectivity according to claim 2, characterized in that, In step S1.1, radiometric calibration and atmospheric correction are performed using ENVI remote sensing image processing software.

4. The wetland zoning method based on the spatial heterogeneity of surface water-groundwater connectivity according to claim 3, characterized in that, The hydrological barrier identification method in S1.2 is as follows: the random forest supervised classification algorithm is used to construct a training set and optimize the classification model using manually labeled wetland sample data from the most recent year. Based on the identification of long-term hydrological barriers, artificial facilities that still maintain their barrier function during the high-water season were further screened by overlaying multiple remote sensing images and field survey data. These artificial facilities include flood control dikes and main roads. Based on the spatial distribution of hydrological barrier zones and artificial barrier facilities, areas that still maintain significant water body barrier function during the high-water season were selected, and wetlands were divided into multiple independent patches based on their spatial distribution characteristics.

5. The wetland zoning method based on the spatial heterogeneity of surface water-groundwater connectivity according to claim 1, characterized in that, The cluster analysis in S2.2 includes the following steps: S2.2.1 Obtaining the optimal number of clusters: The optimal number of clusters is determined by calculating the sum of squared errors and changes in the silhouette coefficient under different numbers of clusters. S2.2.2 Selecting the optimal clustering method: The algorithm performance was evaluated by combining the silhouette coefficient, Calinski-Harabasz index, Davies-Bouldin index, and BIC / AIC, and the optimal clustering method was selected for clustering.

6. The wetland zoning method based on the spatial heterogeneity of surface water-groundwater connectivity according to claim 5, characterized in that, Specifically, S2.3 involves: spatially aggregating the already divided surface connectivity patches based on the hydrogen and oxygen isotope clustering results; during the integration process, patches with similar isotope characteristics and spatial proximity are preferentially merged into the same partition; and after integration, a functional partitioning framework with multiple partitions and reflecting groundwater connectivity is formed.

7. The wetland zoning method based on the spatial heterogeneity of surface water-groundwater connectivity according to claim 1, characterized in that, The quantitative standard in S3 is as follows: when the AFR value of adjacent patches is greater than 60%, it indicates that there is a hydraulic connection between the two. If surface hydrological observation shows that there is a barrier between the patches, it can be inferred that the groundwater system is the key channel to maintain its hydrological connectivity. On this basis, the dynamic changes of water level in the two patches are further compared. If there is a significant synchronicity, it can be used to prove that cross-patch water replenishment is achieved through groundwater exchange.

Images