Ecological environment monitoring method and system based on geographic information technology

Abstract

The invention discloses an ecological environment monitoring method and system based on a geographic information technology, and relates to the technical field of ecological environment monitoring, and the method comprises the steps: carrying out the multi-source heterogeneous data fusion of a multispectral image and point-like time series data of a water ecological protection target water area, and obtaining a three-dimensional water quality dynamic situation field; future trajectory prediction and water ecological health analysis are carried out on the three-dimensional water quality dynamic situation field to obtain a global water ecological health index thermodynamic cloud picture, pollution diffusion is carried out on the global water ecological health index thermodynamic cloud picture to obtain a pollutant trajectory matrix, ecological risk prediction is carried out based on the pollutant trajectory matrix, and the ecological risk prediction result is obtained. According to the method, the spatial-temporal dynamic risk early warning grade map is obtained, the restoration scheme is optimized based on the spatial-temporal dynamic risk early warning grade map, the restoration scheme optimal solution set is obtained, the timeliness and accuracy of ecological risk early warning are improved, and the economy of restoration decision is enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of ecological environment monitoring, in particular to an ecological environment monitoring method and system based on geographic information technology. BACKGROUND

[0002] Water ecosystems maintain ecological balance, biodiversity and water resource circulation. At present, pollution, overdevelopment and the like cause water ecological degradation, often leading to water quality deterioration, flooding and the like, while the attention to the protection and restoration of water ecology is also rising, so the monitoring of water ecological environment is crucial. However, the existing ecological environment monitoring has the following defects: The existing ecological environment monitoring is difficult to construct an accurate three-dimensional water quality dynamic diagram, and the details of water quality at different water depths and different time periods cannot be considered; it is mostly limited to static analysis or short-term trend judgment, and the long-term trajectory of water quality evolution and the spatiotemporal characteristics of pollution diffusion cannot be considered, so it is difficult to identify mutation risks in time; it can only focus on rough simulation of macro-basins, and the simulation accuracy of pollution migration trajectory is limited; at the same time, the index setting is single and the threshold is fixed, so it is difficult to dynamically consider key factors such as pollution concentration and residence time, and the matching degree of early warning level and actual ecological damage is not high; it mostly relies on experience, lacks comprehensive consideration of risk reduction effect and cost investment, and cannot screen out the optimal solution balancing ecological benefits and economic costs. SUMMARY

[0003] In order to solve the above technical problems in the prior art, the present application provides, on the one hand, an ecological environment monitoring method based on geographic information technology, and on the other hand, an ecological environment monitoring system based on geographic information technology.

[0004] The technical scheme adopted by the present application to solve the above technical problems is as follows: In a first aspect, the present application provides an ecological environment monitoring method based on geographic information technology, comprising the following steps: Step S1: multi-source heterogeneous data fusion is performed on the multi-spectral images and point-like time series data of the water ecological protection target water area to obtain a three-dimensional water quality dynamic situation field; Step S2: future trajectory prediction and water ecological health analysis are performed on the three-dimensional water quality dynamic situation field to obtain a global water ecological health index thermal cloud map; Step S3: pollution diffusion is performed on the global water ecological health index thermal cloud map to obtain a pollution trajectory matrix; Step S4: ecological risk prediction is performed based on the pollution trajectory matrix to obtain a spatiotemporal dynamic risk early warning level atlas; Step S5: repair scheme optimization is performed based on the spatiotemporal dynamic risk early warning level atlas to obtain an optimal solution set of the repair scheme.

[0005] Further, the three-dimensional water quality dynamic trend field analysis step is as follows: The water ecological protection target water area is acquired by a satellite sensor to obtain multi-spectral images, radiation calibration and atmospheric correction are performed, standard reflectivity images are generated, the standard reflectivity images are processed by a multi-band collaborative inversion model to obtain spectral feature data, spectral decision cluster generation is performed on the spectral feature data, and a water quality parameter spatial distribution atlas is output; Based on the water quality parameter spatial distribution atlas, point-like time series data are fused by a spatio-temporal Kriging interpolation method to construct a three-dimensional water quality dynamic trend field.

[0006] Further, the global water ecological health index thermal cloud map analysis step is as follows: The voxel attribute of the three-dimensional water quality dynamic trend field is calculated to obtain local parameter differences, and an abnormal parameter distribution area is identified based on the local parameter differences to obtain pollution spatial features of the three-dimensional water quality dynamic trend field; time evolution mutation features of a hydrological evolution trajectory matrix are predicted by a gated recurrent unit, and the pollution spatial features of the three-dimensional water quality dynamic trend field and the time evolution mutation features of the predicted hydrological evolution trajectory matrix are weighted and fused to obtain a weighted fusion feature vector; The weighted fusion feature vector is decoded by a fully connected network to generate a basic data matrix of an evaluation index, the evaluation index weight is calculated by an entropy weight method based on the basic data matrix of the evaluation index, and the water ecological health index is calculated by combining TOPSIS ideal solution sorting; The direction of a predicted parameter value spatial gradient and the intensity of a predicted parameter value time gradient of the hydrological evolution trajectory matrix are calculated to synthesize a pollution migration vector of each spatio-temporal point, and a global water ecological health index thermal cloud map is generated based on the water ecological health index and the pollution migration vector of each spatio-temporal point.

[0007] Further, the predicted hydrological evolution trajectory matrix analysis step is as follows: The three-dimensional water quality dynamic trend field is divided into a plurality of water body units, water quality parameters and corresponding real-time meteorological data of the water body units are extracted, and the water quality parameters and the real-time meteorological data are processed to obtain a hydrological and meteorological spatio-temporal feature tensor; The hydrological and meteorological spatio-temporal feature tensor is subjected to spatio-temporal nonlinear correlation and future trajectory prediction to obtain a predicted hydrological evolution trajectory matrix.

[0008] Further, the pollutant trajectory matrix analysis step is as follows: The SWMM model is established based on the hydrological evolution trajectory matrix and the pollution source diffusion path, wherein the hydrological evolution trajectory matrix is used as a dynamic boundary condition of the SWMM model, the SWMM model comprises Saint-Venant equation set and a pollutant convection-diffusion equation, and the river section flow velocity and the pollutant concentration field are obtained by solving the SWMM model; and the river section flow velocity and the pollutant concentration field are used as an inlet boundary of the CFD model, and the CFD model comprises a Navier-Stokes equation constrained by a k-ε turbulent flow model and a species transport equation; The SWMM model and the CFD model are coupled through a real-time data exchange interface, the CFD model outputs a three-dimensional flow velocity field and a concentration space-time distribution, the three-dimensional flow velocity field is used to obtain a residence time spectrum through a Lagrangian tracking method, and the concentration space-time distribution and the residence time spectrum form a pollutant trajectory matrix.

[0009] Further, the pollution source diffusion path analysis step is as follows: The water body units of the global water ecological health index thermal cloud map are mapped into graph nodes, the node feature vectors are assigned as coupled values of the health indexes and evolution confidence probabilities of the corresponding water body units, if the pollution migration vector direction of the global water ecological health index thermal cloud map is consistent with the hydrological connectivity determined based on the digital elevation model or the included angle is less than 45 degrees, a directed edge is established in this direction, otherwise, a directed edge is established based on the water flow direction determined by the hydrological connectivity, a directed edge structure between the nodes is obtained, the edge weight is defined as the product of the modulus of the pollution migration vector and the reciprocal of the Euclidean distance between the center points of adjacent water body units, the obtained weighted directed graph is input into the graph convolution network to aggregate and weight the information of the neighbor nodes, an embedding vector is output, and the output embedding vectors of all nodes are analyzed through a clustering algorithm to obtain the pollution source diffusion path.

[0010] Further, the spatiotemporal dynamic risk early warning grade atlas analysis step is as follows: Based on the ecological risk entropy index and the residence time spectrum, a dynamic threshold triggering mechanism is used to generate a spatiotemporal dynamic risk early warning grade atlas.

[0011] Further, the ecological risk entropy index analysis step is as follows: A dynamic stress response matrix is constructed through the pollutant trajectory matrix, wherein the rows represent different combinations of spatial positions and time points, and the columns represent evaluation indexes derived from chlorophyll concentration; objective weights of the evaluation indexes are calculated based on the dynamic stress response matrix through information entropy; and relative closeness degrees of each sample to positive and negative ideal solutions are calculated based on the objective weights of the evaluation indexes through a TOPSIS algorithm, and the relative closeness degrees of each sample to the positive and negative ideal solutions are mapped into ecological risk entropy indexes.

[0012] Further, the optimal solution set of the repair scheme analysis step is as follows: The final remediation treatment pollutant concentration distribution is obtained by the calculation mode of step S4, the ecological risk entropy index after the final remediation treatment is obtained, the risk reduction efficiency is calculated based on the ecological risk entropy index after the final remediation treatment, the cost-benefit ratio of the multi-scenario remediation scheme is obtained by analyzing the cost-benefit ratio of the multi-scenario remediation scheme and the ecological risk entropy index after the final remediation treatment, and the remediation scheme performance matrix is output. The risk reduction efficiency and the cost-benefit ratio in the remediation scheme performance matrix are obtained by setting a screening rule to obtain the optimal solution set of the remediation scheme.

[0013] Further, the final remediation treatment pollutant concentration distribution analysis step is as follows: Based on the time-space dynamic risk early warning grade atlas, the dynamic intervention parameters of different remediation strategies are set to obtain a multi-scenario remediation scheme, a coupled hydrodynamic model is established for the multi-scenario remediation scheme, the model includes a depth-averaged shallow water equation, a depth-averaged flow field is obtained by simultaneous solution and numerical algorithm, the depth-averaged flow field is substituted into the pollutant degradation Monod equation, and the time-space discrete iterative solution of the pollutant degradation Monod equation is performed by using a numerical algorithm to obtain the pollutant concentration distribution after the final remediation treatment.

[0014] In a second aspect, the present application provides an ecological environment monitoring system based on geographic information technology, comprising a heterogeneous data fusion module, an ecological health analysis module, a pollution diffusion module, an ecological risk prediction module and a remediation scheme optimization module. The heterogeneous data fusion module is used for obtaining multi-source heterogeneous data fusion of multi-spectral images and point-like time series data in the water ecological protection target water area, obtaining a three-dimensional water quality dynamic situation field, and sending it to the ecological health analysis module.

[0015] The ecological health analysis module is used for future trajectory prediction and water ecological health analysis of the three-dimensional water quality dynamic situation field to obtain a global water ecological health index thermal cloud map, and send it to the pollution diffusion module.

[0016] The pollution diffusion module is used for pollution diffusion of the global water ecological health index thermal cloud map to obtain a pollutant trajectory matrix, and send it to the ecological risk prediction module.

[0017] The ecological risk prediction module is used for ecological risk prediction based on the pollutant trajectory matrix to obtain a time-space dynamic risk early warning grade atlas, and send it to the remediation scheme optimization module.

[0018] The remediation scheme optimization module is used for remediation scheme optimization based on the time-space dynamic risk early warning grade atlas to obtain the optimal solution set of the remediation scheme.

[0019] Compared with the prior art, the present application has the following beneficial effects: 1、The present application carries out multi-source heterogeneous data fusion on the multi-spectral image and point-like time series data of the water ecological protection target water area to obtain a three-dimensional water quality dynamic situation field, carries out future trajectory prediction and water ecological health analysis on the three-dimensional water quality dynamic situation field to obtain a global water ecological health index thermal cloud map, carries out pollution diffusion on the global water ecological health index thermal cloud map to obtain a pollutant trajectory matrix, carries out ecological risk prediction based on the pollutant trajectory matrix to obtain a time-space dynamic risk early warning grade atlas; through the consideration of multi-source data, an accurate three-dimensional water quality dynamic graph is constructed, the water quality details of different water depths and different time periods are considered, the long-term trajectory of water quality evolution and the time-space characteristics of pollution diffusion are considered, and then the mutation risk is identified in time; the simulation of the pollutant migration trajectory is accurate; the key factors such as pollution concentration and residence time are dynamically considered, and the matching degree of the early warning grade and the actual ecological damage situation is high.

[0020] 2、The present application carries out repair scheme optimization based on the time-space dynamic risk early warning grade atlas to obtain an optimal solution set of the repair scheme, can comprehensively consider the risk reduction effect and the cost investment, and can screen out an optimal scheme balanced between ecological benefits and economic costs. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 The present application is a method flow chart; Figure 2 The present application is a pollutant trajectory matrix analysis flow chart; Figure 3 The present application is a step S1 analysis flow chart; Figure 4 The present application is a schematic diagram of the connection of each module. DETAILED DESCRIPTION

[0022] The technical solutions of the present application will be described clearly and completely in combination with the drawings and specific embodiments.

[0023] Embodiment 1: As shown in the figure, the ecological environment monitoring method based on geographic information technology comprises the following steps: Figure 1 Step S1: multi-source heterogeneous data fusion is carried out on the multi-spectral image and point-like time series data of the water ecological protection target water area to obtain a three-dimensional water quality dynamic situation field; Step S2: future trajectory prediction and water ecological health analysis are carried out on the three-dimensional water quality dynamic situation field to obtain a global water ecological health index thermal cloud map; Step S3: pollution diffusion is carried out on the global water ecological health index thermal cloud map to obtain a pollutant trajectory matrix; ​Step S4: Based on the pollutant trajectory matrix, perform ecological risk prediction to obtain a spatiotemporal dynamic risk warning level map; Step S5: Optimize the repair scheme based on the spatiotemporal dynamic risk warning level map to obtain the optimal solution set of the repair scheme.

[0024] like Figure 3 As shown in the embodiment, the specific steps of step S1 are as follows: Multispectral images of target water bodies for water ecological protection are acquired using geographic information technology and satellite sensors. Radiometric calibration and atmospheric correction are performed to generate standard reflectance images. These standard reflectance images are then processed using a multi-band collaborative inversion model to obtain spectral feature data. Spectral decision cluster generation is then performed on the spectral feature data to output a spatial distribution map of water quality parameters. Geographic information technology includes remote sensing (RS), global positioning system (GPS), and geographic information system (GIS) to facilitate the acquisition of relevant data for target water bodies for water ecological protection.

[0025] Based on the spatial distribution map of water quality parameters, a three-dimensional dynamic situation field of water quality is constructed by fusing point time series data through spatiotemporal kriging interpolation. In the embodiment, the water ecological protection target water area (such as a lake, a river or a wetland protection area, and the specific protection area is not limited in the present application) receives the surface reflected solar radiation energy through a satellite sensor, records the radiation intensity of a specific wave band (blue, green, red, near-infrared, etc.), forms an original multispectral image, and generates a standard reflectivity image after radiation calibration (eliminating the sensor itself error, converting the original signal into absolute radiation) and atmospheric correction (removing atmospheric scattering, absorption and other interference). The specific radiation calibration is to convert the original DN value (digital quantization value) into physical quantities such as radiance and reflectivity by using the sensor response model + calibration coefficient (onboard calibration), so as to eliminate the sensor itself response error. The atmospheric correction is to further process the standard reflectivity image based on the water body spectral response mechanism by using a physical method (such as 6S, MODTRAN). Specifically, the standard reflectivity image is converted into spectral feature data that can be used for water quality analysis by using a multi-band collaborative inversion model. The multi-band collaborative inversion model is a prior art, which is not described herein. The spectral feature data is subjected to a spectral decision cluster generation operation by using a fuzzy clustering algorithm (FCM), and a water quality parameter spatial distribution map is output. Each pixel contains a parameter type (chlorophyll / turbidity, etc.), a concentration level, and a membership confidence value. The specific fuzzy clustering algorithm defines an objective function, establishes a clustering rationality measurement rule, and alternately optimizes the iterative membership reassignment and cluster center spectral attraction calibration, so as to constantly update the membership and cluster center vectors, realize the decomposition of mixed pixels to chlorophyll / turbidity and other parameter spectral decision clusters, and obtain the membership of each pixel belonging to each water quality cluster (for example, the possibility of belonging to the chlorophyll decision cluster is 0.8, and the possibility of belonging to the turbidity decision cluster is 0.1) and the optimized water quality cluster center spectrum when the membership and cluster center change is less than a threshold value. Based on the water quality parameter spatial distribution map (containing chlorophyll, turbidity and other rasterized data), the point-shaped time series data of pH, dissolved oxygen and turbidity collected in real time by the Internet of Things sensor are fused, and a three-dimensional water quality dynamic situation field is constructed by using a spatiotemporal Kriging interpolation method. Specifically, the autocorrelation of the water quality parameter under the spatial distance and time lag is quantified by using a spatiotemporal variogram function model, the parameter values at unmonitored positions are calculated based on the spatiotemporal variogram function by using a dynamic weight tensor, the water quality parameters at unmonitored positions are calculated by using the above weight in combination with the known point data, and finally a three-dimensional water quality dynamic situation field containing spatial coordinates (longitude / latitude), water depth profile (0-10m layering) and time dimension (minute-level update) is generated. The voxel attribute of the three-dimensional water quality dynamic situation field contains multi-parameter concentration values (chlorophyll, turbidity, pH, dissolved oxygen, etc.) and interpolation confidence entropy (measuring the uncertainty of the interpolation result). The spatiotemporal variogram function model is a prior art, which is not described herein.

[0026] In the embodiment, the specific steps of step S2 are as follows: The three-dimensional water quality dynamic trend field is divided into a plurality of water body units, water quality parameters of each water body unit and corresponding real-time meteorological data are extracted, the water quality parameters and the real-time meteorological data are processed, and a hydro-meteorological spatio-temporal feature tensor is obtained; The hydro-meteorological spatio-temporal feature tensor is subjected to spatio-temporal nonlinear correlation and future trajectory prediction, and a predicted hydrological evolution trajectory matrix is obtained; A local parameter difference is calculated based on voxel attributes of the three-dimensional water quality dynamic trend field, an abnormal parameter distribution area is identified based on the local parameter difference, and a pollution space feature of the three-dimensional water quality dynamic trend field is obtained; a time evolution mutation feature of the hydrological evolution trajectory matrix is predicted through a gated recurrent unit, and a weighted fusion feature vector is obtained by weighting and fusing the pollution space feature of the three-dimensional water quality dynamic trend field and the time evolution mutation feature of the predicted hydrological evolution trajectory matrix; The weighted fusion feature vector is decoded through a fully connected network to generate a basic data matrix of an evaluation index, the evaluation index weight is calculated based on the basic data matrix of the evaluation index through an entropy weight method, the water ecological health index is calculated in combination with TOPSIS ideal solution sorting; The direction of a predicted parameter value space gradient and the intensity of a time gradient of a predicted parameter value of the hydrological evolution trajectory matrix are calculated, a pollution migration vector of each spatio-temporal point is synthesized, and a global water ecological health index heat cloud map is generated based on the water ecological health index and the pollution migration vector of each spatio-temporal point; In the embodiment, the three-dimensional water quality dynamic state field is divided into 10m*10m grid, 1m water depth layer, and a plurality of water body units, the chlorophyll concentration, dissolved oxygen, turbidity parameters and other water quality parameters of each water body unit (longitude x, latitude y, water depth z) in the three-dimensional water quality dynamic state field are extracted, the real-time weather data (specifically rainfall, flow rate) of the region where each water body unit is located is synchronously collected, the water quality parameters and real-time weather data are stacked along the time axis to form a high-dimensional input sequence, which is marked as hydro-meteorological spatiotemporal feature tensor = [water quality parameters (x, y, z, t), meteorological parameters (t)], wherein t represents time, when the LSTM neural network is used to predict the hydrological evolution trajectory, the hydro-meteorological spatiotemporal feature tensor is input into the forget gate and the historical information to be retained is screened according to the current hydro-meteorological feature and the historical water quality state; the input gate synchronously identifies the value of new information and decides the retention weight, and the two cooperate to update the cell state and dynamically integrate the effective information of long time series; the output gate extracts key features to generate prediction results based on the updated cell state, outputs the predicted hydrological evolution trajectory matrix, through the learning of the LSTM gated memory chain, two kinds of mappings are realized: 1) nonlinear relationship modeling, learning the spatiotemporal nonlinear correlation between hydrological parameters (such as dissolved oxygen decay rate, turbidity diffusion coefficient) and meteorological driving (rainfall, flow rate), 2) future trajectory prediction: outputting the hydrological evolution trajectory matrix in the next 24 hours (with an interval of 15 minutes), which includes the predicted parameter values (chlorophyll, dissolved oxygen, turbidity, rainfall, flow rate, etc.) and evolution confidence probability (calculated based on the Monte Carlo Dropout layer, quantifying the uncertainty of the prediction result, the closer the probability value is to 1, the higher the confidence) of each water body unit. Through CNN (Convolutional Neural Network), the pollution spatial features (such as pollution patch morphology) of the three-dimensional water quality state field are extracted. Specifically, the voxel attributes (such as chlorophyll, turbidity, and other multi-parameter concentration values) of the three-dimensional water quality dynamic state field are input into CNN in the form of a three-dimensional matrix, the convolution kernel slides in the matrix, the local parameter difference (such as the gradient between the high-concentration area of chlorophyll and the surrounding area) is calculated, the continuous abnormal parameter distribution area (determined by threshold judgment method) is identified, and these areas are the pollution patch morphology, realizing the extraction of the shape features such as the outline and range of the pollution patch. Through GRU (Gated Recurrent Unit), the time evolution mutation features (such as mutation inflection points) of the hydrological trajectory are analyzed. Specifically, based on the predicted hydrological evolution trajectory matrix, the future 24-hour time series of chlorophyll, dissolved oxygen, turbidity and other multi-parameter prediction values of each water body unit are extracted, together with the evolution confidence probability, input into the gated recurrent unit (GRU), and the GRU uses the gating mechanism to mine the long-term correlation between parameters, calculates the dynamic change rate, and when the parameter change rate breaks through the historical statistical threshold (such as a large fluctuation of dissolved oxygen in a short time), combined with the evolution confidence probability, the low-confidence noise interference is screened out, if the preset mutation judgment condition is met, the corresponding time is identified as a hydrological mutation inflection point. Through the federal feature fusion gate and time evolution mutation characteristics Weighted fusion features, get weighted fusion feature vector , wherein denotes element-wise multiplication, denotes a dynamically generated gating weight matrix, , wherein σ is a Sigmoid function, output [0, 1] weight, respectively represent the weight matrix and the bias, and the weighted fusion feature vector is decoded through a fully connected network to generate a basic data matrix of evaluation indicators (such as dissolved oxygen decay rate, pH abnormal fluctuation degree, turbidity mutation rate, pollution influence range (pollution patch covered water area), pollution group diffusion rate (pollution patch diffusion speed with time), positive / negative deviation degree of chlorophyll burst risk , k represents the index number, and the maximum value is m, m=6), based on the basic data matrix of the evaluation indicators, the index weight is calculated by the entropy weight method (The entropy weight method is prior art, which is not described here), combined with TOPSIS ideal solution sorting, the water ecological health index is finally calculated , the space gradient direction (unit vector) of the prediction parameter value of the hydrological evolution trajectory matrix and the intensity (scalar size) of the time gradient are combined to synthesize the pollution migration vector of each space-time point (latitude and longitude x, y+ time t) , taking the chlorophyll concentration C as the representative of the pollution migration vector, respectively represent the spatial gradient and the time gradient of the chlorophyll concentration, wherein the spatial gradient can adopt Sobel operator or central difference algorithm, and the negative sign means that the pollution diffuses from the high concentration area to the low concentration area (opposite to the direction of the spatial gradient), and the pollution migration vector is integrated with the water ecological health index to generate a global water ecological health index thermal cloud map with geographical coordinates as the base and time axis as the third dimension, the pixel value of which contains the real-time water ecological health index of each water body unit, the pollution migration vector and the evolution confidence probability (from the predicted hydrological evolution trajectory matrix).

[0027] In the embodiment, the specific steps of step S3 are as follows: The water body unit of the global water ecological health index thermal cloud map is mapped to a graph node, and the node feature vector is assigned to the coupling value of the health index and the evolution confidence probability of the corresponding water body unit. If the pollution migration vector direction of the global water ecological health index thermal cloud map is consistent with the hydrological connectivity determined based on the digital elevation model or the included angle is less than 45 degrees, a directed edge is established in this direction. Otherwise, a directed edge is established based on the water flow direction determined by the hydrological connectivity. The directed edge structure between nodes is obtained. The edge weight is defined as the product of the modulus of the pollution migration vector and the reciprocal of the Euclidean distance between the centers of adjacent water body units. The obtained weighted directed graph is input into the graph convolution network to aggregate the information of the weighted neighbor nodes, and the embedding vector is output. The output embedding vectors of all nodes are analyzed by a clustering algorithm to obtain the pollution source diffusion path. A SWMM model is established based on the hydrological evolution trajectory matrix and the pollution source diffusion path, wherein the hydrological evolution trajectory matrix is used as a dynamic boundary condition of the SWMM model. The SWMM model includes Saint-Venant equation set and pollutant convection and diffusion equation. The river section flow velocity and pollutant concentration field are obtained by solving the SWMM model. The river section flow velocity and pollutant concentration field are used as the inlet boundary of the CFD model. The CFD model includes the Navier-Stokes equation constrained by the k-ε turbulence model and the species transport equation. The SWMM model and the CFD model are coupled through a real-time data exchange interface. The CFD model outputs a three-dimensional flow velocity field and a concentration spatiotemporal distribution. Based on the three-dimensional flow velocity field, the residence time spectrum is obtained by the Lagrangian tracking method. The concentration spatiotemporal distribution and the residence time spectrum constitute the pollutant trajectory matrix. As Figure 2As shown, in the embodiment, the water body units of the global water ecological health index thermal cloud map are mapped to graph nodes, and the node feature vector is assigned to the coupled value of the health index and the evolution confidence probability of the corresponding water body unit (corresponding to HEI x evolution confidence probability), which can enhance the feature signal of the high confidence area and attenuate the feature of the low confidence area. The directed edge structure between nodes is jointly determined by fusing the pollution migration vector and hydrological connectivity: for any two water body units (nodes), if the direction of the pollution migration vector is consistent with the water flow direction (hydrological connectivity) determined based on the D8 algorithm of the digital elevation model (DEM) or the included angle is less than 45 degrees, a directed edge is established in this direction, otherwise, the water flow direction determined based on the hydrological connectivity is used as the basis to establish a directed edge, so as to construct a directed graph network topology that conforms to the physical diffusion law. The edge weight is specifically defined as the product of the length of the pollution migration vector (representing the diffusion intensity) and the inverse of the Euclidean distance between the center points of adjacent units, so as to quantify the potential intensity of pollution transmission between nodes. The weighted directed graph constructed above is input into the graph convolution network (GCN), and through the multi-layer message passing mechanism of the graph convolution network, each node continuously aggregates the feature information of its upstream neighbors (according to the direction of the directed edge). This process enables the pollution features to be propagated and enhanced in the reverse direction along the diffusion path: nodes located at the end of the pollution path have their embedding vector values increased by aggregating a large amount of high pollution features of downstream nodes, while the feature values of clean areas or isolated nodes remain low. Finally, by analyzing the output embedding vectors of all nodes through a clustering algorithm, the core aggregation point with extremely high embedding vector value can be automatically identified as the location of the pollution source, and the complete directed path connecting the pollution source and the high pollution node, i.e., the pollution diffusion path topology, is restored, thereby accurately realizing the pollution source positioning and diffusion trajectory tracing from the global pollution state. where Q represents the flow (unit time water discharge), the initial value is set by the hydrological evolution trajectory matrix, and the subsequent value is iterated by the model, A represents the water area (the cross-sectional area of the river / channel), the area of the submerged part in the river / channel cross section, which is calculated in combination with the water depth h and the cross-sectional shape (such as rectangular, trapezoidal), the initial value of the water depth h is obtained from the hydrological evolution trajectory matrix, and the subsequent value is iterated by the model, x represents the spatial coordinate (one-dimensional length of the river / channel), t represents time, the time series (future 24 hours) of the hydrological evolution trajectory matrix, and q represents the side inflow (such as pipe network access flow). Based on the pollution source diffusion path, the pollution convection and diffusion equation is established where C denotes the pollutant concentration (referring to chlorophyll), the initial value is set by the pollutant source diffusion path (pollution starting concentration), and the subsequent value is iterated by the model, A and Q are the same as the Saint-Venant equations, the Saint-Venant equations dynamically solve the flow Q and the water area A derived from the water depth h combined with the cross-section shape at different times and spaces, the pollutant convection-diffusion equation directly uses the dynamic results, realizes the mutual simulation of water flow movement and pollutant transport process, D denotes the diffusion coefficient (pollutant diffusion in water flow), the initial value is set by experience / literature, and the subsequent value is corrected by the CFD model (after CFD calculates turbulent mixing, it is fed back to SWMM), S denotes the source and sink term, the pollutant source diffusion path is directly given, the amount of pollutant input per unit time C, the river cross-section flow rate (Q / A) and the pollutant concentration field C are solved, and the river cross-section flow rate and the pollutant concentration field are taken as the inlet boundary of the CFD model (including Navier-Stokes equation and species transport equation) where denotes the three-dimensional flow velocity vector, the initial value is set by the river cross-section flow rate (one-dimensional, along the river longitudinal direction x, obtained by the SWMM model) and the set vertical / lateral flow rate (initially set to 0), and the subsequent value is iterated by the model, ρ denotes the fluid density (such as the density of water), p denotes the pressure, the initial value is the hydrostatic pressure p=ρgh, h denotes the water depth of the hydrological evolution trajectory matrix, μ denotes the dynamic viscosity (viscosity of water), denotes the gravity acceleration vector, and the species transport equation where C and are from the SWMM model and the Navier-Stokes equation respectively, denotes the effective diffusion coefficient (considering turbulent diffusion), according to the direction information provided by the pollutant source diffusion path topology, in the preferential diffusion direction of the pollutant, the effective diffusion coefficient is increased, so that the pollutant is more easily diffused in this direction, for example, the component corresponding to the preferential diffusion direction is multiplied by a coefficient α greater than 1, such as α=1.5, and in the direction perpendicular or opposite to the preferential diffusion direction, the component value of the diffusion coefficient is reduced, which can be multiplied by a coefficient β less than 1, such as β=0.5, denotes the divergence, denotes the gradient, denotes the Laplace operator, the SWMM model and the CFD model are coupled through a real-time data exchange interface (SWMM provides updated river hydraulic parameters to CFD every 15 minutes, CFD feeds back the concentration gradient caused by local turbulent mixing to SWMM to correct the diffusion coefficient D, by solving the Navier-Stokes equation and the species transport equation, the pollutant concentration distribution at each point in the three-dimensional space is obtained, the concentration gradient in the local area is calculated, and the specific adjustment rule can be determined based on the empirical formula or through further model calibration, for example, a linear relationship Dnew =D old +k x concentration gradient, where k is a calibrated coefficient, D new , D old respectively represent the updated and pre-updated diffusion coefficients), finally generating a pollutant trajectory matrix integrating the migration path and the turbulent diffusion mechanism, containing the concentration spatiotemporal distribution C) x, y, z, t) and the residence time spectrum T(x, y), wherein the three-dimensional flow field output from the CFD model k-ε turbulent model constrained Navier-Stokes equation is used to obtain the residence time spectrum by the Lagrangian tracking method, and the residence time spectrum T(x, y) refers to the time distribution of pollutants at different positions (represented by two-dimensional coordinates x, y) in the water body, which reflects the time experienced by the pollutants from entering to leaving at a specific position. The Lagrangian tracking method is a prior art and will not be described in detail here. The SWMM model is used to simulate the surface runoff and pipe network confluence process of the watershed to obtain the macroscopic water flow movement and the one-dimensional spatial migration trend of pollutants; the CFD model focuses on accurately simulating the turbulent diffusion, vertical mixing and chemical reaction process of pollutants in the water body. The two are coupled through a real-time data exchange interface, and the concentration spatiotemporal distribution output by the CFD model, combined with the residence time spectrum obtained based on the Lagrangian tracking method, together constitute a pollutant trajectory matrix that can accurately present the whole process of pollutant migration.

[0028] In the embodiment, the specific steps of step S4 are as follows: A dynamic stress response matrix is constructed through the pollutant trajectory matrix, wherein the rows represent different combinations of spatial positions and time points, and the columns represent evaluation indexes derived from the chlorophyll concentration. The objective weights of the evaluation indexes are calculated based on the dynamic stress response matrix through information entropy, and the relative closeness of each sample to the positive and negative ideal solutions is calculated based on the objective weights of the evaluation indexes through the TOPSIS algorithm. The relative closeness of each sample to the positive and negative ideal solutions is mapped to an ecological risk entropy index. Based on the ecological risk entropy index and the residence time spectrum, a spatiotemporal dynamic risk warning grade atlas is generated using a dynamic threshold triggering mechanism.

[0029] In the embodiment, a dynamic stress response matrix is constructed by a pollutant trajectory matrix, wherein rows represent different spatial positions and time points combined samples, and columns represent multiple evaluation indexes derived from chlorophyll concentration (specifically: ① current concentration and critical concentration ratio, reflecting the immediate pressure of pollution exceeding threshold ② concentration duration and residence time ratio, measuring the proportion of pollution stress ③ concentration change trend slope, the larger the positive slope, the faster the pollution accumulation and the faster the risk rises), objective weights of each evaluation index are calculated based on the dynamic stress response matrix by information entropy (specifically: sample data of each index in the matrix is normalized to convert into a probability distribution (the relative proportion of a sample under a certain index), the data dispersion of the index is quantified by the information entropy formula: the higher the dispersion, the stronger the index's ability to distinguish sample differences, the higher the weight, and the objective weights of the three types of indexes are output), and the relative closeness of each sample to the positive and negative ideal solutions is calculated based on the objective weights of each evaluation index by the TOPSIS algorithm, specifically: the three types of indexes are normalized to standard values between 0 and 1 by the maximum-minimum normalization method, the sample standard values are weighted according to the objective weights of each evaluation index, and then the Euclidean distance of the weighted sample to the positive ideal solution (the optimal value combination of each index) and the negative ideal solution (the worst value combination of each index) is calculated by the TOPSIS algorithm. Finally, the distance from the sample to the negative ideal solution is divided by the sum of the distances to the positive and negative ideal solutions to obtain the relative closeness, which reflects the closeness of the sample to the ideal state. The relative closeness of each sample to the positive and negative ideal solutions is mapped to the ecological risk entropy index ERI∈[0,1], and the relative closeness is subtracted by 1 to achieve it, that is, the closer the relative closeness to 1 (the closer to the positive ideal solution, the safer the ecological state), the closer the ecological risk entropy index to 0; the closer the relative closeness to 0 (the closer to the negative ideal solution, the more dangerous the ecological state), the closer the ecological risk entropy index to 1. Thus, the quantitative mapping from closeness to risk index is completed. The dynamic threshold triggering mechanism includes setting multi-level linkage decision rules, for example, when the ecological risk entropy index is greater than 0.7 and the residence time spectrum is greater than 24 hours, a red warning is triggered (indicating ecosystem collapse), when 0.4 < ecological risk entropy index ≤ 0.7 and 12 < residence time spectrum ≤ 24 hours, an orange warning is triggered (indicating serious degradation of ecological function), when 0.3 < ecological risk entropy index ≤ 0.4 and residence time spectrum ≤ 12 hours, a yellow warning is triggered (indicating mild ecological stress), and other conditions are green (indicating normal ecology), and then the affected specific ecological functions under each risk level are matched according to the ecological toxicology database (such as fish mass mortality and algal bloom for red warning, and benthic community decline for orange warning); a spatiotemporal dynamic risk warning level atlas is generated, each grid cell of which contains two attributes: risk level (red / orange / yellow / green four-color coding) and list of expected damaged ecological functions.

[0030] In the embodiment, the specific steps of step S5 are as follows: Based on the space-time dynamic risk warning level atlas, dynamic intervention parameters of different repair strategies are set to obtain multi-scenario repair schemes. A coupled hydrodynamic model is established for the multi-scenario repair schemes. The model includes a depth-averaged shallow water equation. The depth-averaged flow velocity field is obtained by simultaneous solution and numerical algorithm. The depth-averaged flow velocity field is substituted into the pollutant degradation Monod equation. The pollutant degradation Monod equation is discretized and iteratively solved in space-time by numerical algorithm to obtain the final pollutant concentration distribution after repair treatment. The final pollutant concentration distribution after repair treatment is calculated by the calculation method of step S4 to obtain the final ecological risk entropy index after repair treatment. The risk reduction efficiency is calculated based on the final ecological risk entropy index after repair treatment. The cost-benefit ratio of the multi-scenario repair scheme is obtained by cost-benefit analysis of the multi-scenario repair scheme based on the ecological risk entropy index before and after repair treatment. The repair scheme performance matrix is output based on the cost-benefit ratio of the multi-scenario repair scheme and the final ecological risk entropy index after repair treatment. The risk reduction efficiency and cost-benefit ratio in the repair scheme performance matrix are obtained by setting screening rules to obtain the optimal solution set of the repair scheme.

[0031] In the embodiment, through the space-time dynamic risk warning level atlas, dynamic intervention parameters (aeration intensity, bacteria dosage, and floating island coverage) of different repair strategies (such as fixed-point aeration, upstream bacteria injection, and ecological floating island deployment) in the red / orange risk level display area are set to form multi-scenario repair schemes. There are two dimensions: one is the repair strategy type (such as fixed-point aeration, upstream bacteria injection, and ecological floating island), and the other is the dynamic intervention parameter under the same strategy (such as different aeration intensities and bacteria injection amounts). A coupled hydrodynamic model is established for the multi-scenario repair schemes. The model includes a depth-averaged shallow water equation, including a continuity equation and a momentum equation wherein represents the depth-averaged flow velocity vector, and v represents the viscosity coefficient of water flow movement, represents the bottom friction resistance coefficient, which is calculated by the Manning formula. The water depth field and the depth-averaged flow velocity field are obtained by simultaneous solution and numerical algorithm (such as finite volume method and finite element method). The depth-averaged flow velocity field is substituted into the pollutant degradation Monod equation: wherein represents the pollutant concentration after repair treatment, and chlorophyll content, represents the reaction term, wherein represents the maximum specific growth rate of microorganisms, respectively represent the half-saturation constant and the microbial concentration, used to quantify the influence of biochemical reactions on the concentration of pollutants, the above partial differential equations are discretized in space and time by numerical algorithms (such as finite volume method), iteratively solved to obtain the final pollutant concentration distribution after remediation treatment, the final pollutant concentration distribution after remediation treatment is obtained through the calculation method of step S4, and the ecological risk entropy index after remediation is obtained. Before remediation, the ecological risk entropy index calculated according to the original pollution condition is integrated in the whole region and the whole period to obtain the total risk before remediation. After remediation, based on the new pollutant concentration simulated (i.e. the pollutant concentration after remediation treatment), the same calculation method as before remediation is used to calculate the ecological risk entropy index after remediation, which is also integrated in the whole region and the whole period to obtain the total risk after remediation. The total risk before remediation is subtracted from the total risk after remediation, and then divided by the total risk before remediation to obtain the risk reduction efficiency, which quantifies the effect of remediation measures in the whole region and the whole period. Obtain the cost of the remediation measures, including the capital cost of purchasing equipment, construction engineering, and the operating cost of subsequent energy consumption, chemicals, labor, etc. Add the two parts of the cost to obtain the total cost. Calculate the benefit of remediation by calculating the reduction amount of the ecological risk entropy index before and after remediation, and then multiplying it by the area of the affected area to quantify the benefit brought by the reduction of ecological risk. Divide the total cost by the benefit brought by the reduction of ecological risk to obtain the cost-benefit ratio, which can intuitively reflect the cost of each unit of ecological risk reduction benefit. Finally, output the remediation scheme performance matrix (each row corresponds to a scheme, and the column vector is the risk reduction efficiency and cost-benefit ratio), select the scheme set with the maximum target function value in the remediation scheme performance matrix by setting the screening rule (the weights of risk reduction efficiency and cost-benefit ratio are a1 and a2 respectively, the specific weights are not limited in the present application, and are set according to the designer's consideration), and obtain the optimal solution set of the remediation scheme.

[0032] Specifically, by fusing multi-source data to construct a three-dimensional water quality dynamic trend field, accurate calculation of water quality changes and effective prediction of future trajectory are realized, which can timely identify pollution diffusion path and ecological risk area, and provide accurate basis for risk warning. At the same time, it couples hydraulic and biochemical models to simulate the performance of different remediation schemes, quantifies the risk reduction efficiency and cost-benefit ratio, and finally selects the optimal remediation scheme, which not only improves the timeliness and accuracy of ecological risk warning, but also enhances the economy of remediation decision.

[0033] Embodiment 2: As Figure 4As shown, the ecological environment monitoring system based on geographic information technology comprises a heterogeneous data fusion module, an ecological health analysis module, a pollution diffusion module, an ecological risk prediction module and a remediation scheme optimization module.

[0034] The heterogeneous data fusion module is connected with the ecological health analysis module, the ecological health analysis module is connected with the pollution diffusion module, the pollution diffusion module is connected with the ecological risk prediction module, and the ecological risk prediction module is connected with the remediation scheme optimization module.

[0035] The heterogeneous data fusion module is used for obtaining multi-spectral images and point-shaped time series data of water ecological protection target water area to perform multi-source heterogeneous data fusion, obtaining a three-dimensional water quality dynamic situation field, and sending the three-dimensional water quality dynamic situation field to the ecological health analysis module.

[0036] The ecological health analysis module is used for performing future trajectory prediction and water ecological health analysis on the three-dimensional water quality dynamic situation field, obtaining a global water ecological health index thermal cloud map, and sending the global water ecological health index thermal cloud map to the pollution diffusion module.

[0037] The pollution diffusion module is used for performing pollution diffusion on the global water ecological health index thermal cloud map, obtaining a pollutant trajectory matrix, and sending the pollutant trajectory matrix to the ecological risk prediction module.

[0038] The ecological risk prediction module performs ecological risk prediction based on the pollutant trajectory matrix, obtains a time-space dynamic risk early warning grade atlas, and sends the time-space dynamic risk early warning grade atlas to the remediation scheme optimization module.

[0039] The remediation scheme optimization module performs remediation scheme optimization based on the time-space dynamic risk early warning grade atlas, and obtains an optimal solution set of the remediation scheme.

[0040] The above is a description of the present application and should not be considered as a limitation thereof. Those skilled in the art can make many modifications to the exemplary embodiments without departing from the novel teachings and advantages of the present application, and all such modifications are intended to be included within the scope of the present application as defined by the claims. The present application is defined by the claims and their equivalents.

Claims

1. An ecological environment monitoring method based on geographic information technology, characterized in that, Includes the following steps: Step S1: Multispectral images and point-based time-series data of the target water area for water ecological protection are fused from multiple sources to obtain a three-dimensional dynamic water quality situation field; Step S2: Perform future trajectory prediction and water ecological health analysis on the three-dimensional water quality dynamic situation field to obtain a heat map of the water ecological health index of the whole region. Step S3: Perform pollution diffusion analysis on the heat map of the overall water ecological health index to obtain the pollutant trajectory matrix; Step S4: Based on the pollutant trajectory matrix, perform ecological risk prediction to obtain a spatiotemporal dynamic risk warning level map; Step S5: Optimize the repair scheme based on the spatiotemporal dynamic risk warning level map to obtain the optimal solution set of the repair scheme.

2. The ecological environment monitoring method based on geographic information technology according to claim 1, characterized in that, The steps for the three-dimensional water quality dynamic situation field analysis are as follows: Multispectral images of target water areas for water ecological protection are acquired, radiometric calibration and atmospheric correction are performed, and standard reflectance images are generated. The standard reflectance images are then processed through a multi-band collaborative inversion model to obtain spectral feature data. Spectral decision cluster generation is performed on the spectral feature data to output a spatial distribution map of water quality parameters. Based on the spatial distribution map of water quality parameters, a three-dimensional dynamic situation field of water quality is constructed by fusing point-based time-series data using spatiotemporal kriging interpolation.

3. The ecological environment monitoring method based on geographic information technology according to claim 1, characterized in that, The steps for analyzing the heat map of the overall water ecological health index are as follows: The local parameter differences of the voxel properties of the three-dimensional water quality dynamic situation field are calculated, and the distribution areas of abnormal parameters are identified based on the local parameter differences to obtain the pollution spatial characteristics of the three-dimensional water quality dynamic situation field. The temporal evolution abrupt change characteristics of the hydrological evolution trajectory matrix are predicted by the gated cycle unit. The pollution spatial characteristics of the three-dimensional water quality dynamic situation field and the temporal evolution abrupt change characteristics of the predicted hydrological evolution trajectory matrix are weighted and fused to obtain a weighted fused feature vector. The weighted fusion feature vector is decoded through a fully connected network to generate the basic data matrix of evaluation indicators. The weights of the evaluation indicators are calculated using the entropy weight method based on the basic data matrix of evaluation indicators. The water ecological health index is calculated by combining the TOPSIS ideal solution ranking. The predicted parameter values ​​of the hydrological evolution trajectory matrix are calculated to determine the direction of the spatial gradient and the intensity of the temporal gradient. The pollution migration vector of each spatiotemporal point is synthesized, and a heat map of the whole-domain water ecological health index is generated based on the water ecological health index and the pollution migration vector of each spatiotemporal point.

4. The ecological environment monitoring method based on geographic information technology according to claim 3, characterized in that, The steps for analyzing the predicted hydrological evolution trajectory matrix are as follows: The three-dimensional dynamic field of water quality is divided into several water body units. The water quality parameters and corresponding real-time meteorological data of each water body unit are extracted. The water quality parameters and real-time meteorological data are processed to obtain the hydrological and meteorological spatiotemporal feature tensor. Spatiotemporal nonlinear correlation and future trajectory prediction are performed on the spatiotemporal characteristic tensors of hydrology and meteorology to obtain the predicted hydrological evolution trajectory matrix.

5. The ecological environment monitoring method based on geographic information technology according to claim 1, characterized in that, The steps for analyzing the pollutant trajectory matrix are as follows: A Southwest Swing Model (SWMM) is established based on the hydrological evolution trajectory matrix and the pollution source diffusion path. The hydrological evolution trajectory matrix serves as the dynamic boundary condition of the SWMM model. The SWMM model includes the Saint-Venant equations and the pollutant convection-diffusion equations. Solving the SWMM model yields the river cross-section velocity and pollutant concentration fields. The river cross-section velocity and pollutant concentration fields are used as the inlet boundary of the CFD model. The CFD model includes the Navier-Stokes equations constrained by the k−ε turbulence model and the species transport equations. The SWMM model and the CFD model are coupled through a real-time data exchange interface. The CFD model outputs a three-dimensional velocity field and a spatiotemporal distribution of concentration. Based on the three-dimensional velocity field, the residence time spectrum is obtained through the Lagrange tracking method. The spatiotemporal distribution of concentration and the residence time spectrum form a pollutant trajectory matrix.

6. The ecological environment monitoring method based on geographic information technology according to claim 5, characterized in that, The steps for analyzing the pollution source diffusion path are as follows: The water body units in the heat map of the overall water ecological health index are mapped as graph nodes. The node feature vector is assigned the coupling value of the health index and evolution confidence probability of the corresponding water body unit. If the direction of the pollution migration vector in the heat map of the overall water ecological health index is consistent with the hydrological connectivity determined based on the digital elevation model or the angle is less than 45 degrees, a directed edge is established in this direction. In other cases, a directed edge is established based on the water flow direction determined by the hydrological connectivity, thus obtaining the directed edge structure between nodes. The edge weight is defined as the product of the magnitude of the pollution migration vector and the reciprocal of the Euclidean distance between the center points of adjacent water bodies. The obtained weighted directed graph is input into a graph convolutional network to aggregate the information of weighted neighbor nodes and output the embedding vector. The output embedding vector of all nodes is analyzed by a clustering algorithm to obtain the pollution source diffusion path.

7. The ecological environment monitoring method based on geographic information technology according to claim 1, characterized in that, The analysis steps for the spatiotemporal dynamic risk warning level map are as follows: Based on the ecological risk entropy index and residence time spectrum, a spatiotemporal dynamic risk warning level map is generated using a dynamic threshold triggering mechanism.

8. The ecological environment monitoring method based on geographic information technology according to claim 7, characterized in that, The steps for analyzing the ecological risk entropy index are as follows: A dynamic stress response matrix is ​​constructed using a pollutant trajectory matrix, where rows represent combined samples at different spatial locations and time points, and columns represent evaluation indicators derived from chlorophyll concentration. Based on the dynamic stress response matrix, the objective weights of each evaluation indicator are calculated using information entropy. Based on the objective weights of each evaluation indicator, the relative closeness of each sample to the positive and negative ideal solutions is calculated using the TOPSIS algorithm. The relative closeness of each sample to the positive and negative ideal solutions is then mapped to an ecological risk entropy index.

9. The ecological environment monitoring method based on geographic information technology according to claim 1, characterized in that, The optimal solution set analysis steps for the repair scheme are as follows: The pollutant concentration distribution after final remediation is calculated using the method in step S4 to obtain the ecological risk entropy index after final remediation. The risk reduction efficiency is calculated based on the ecological risk entropy index after final remediation. A cost-benefit analysis of multiple remediation schemes is conducted based on the ecological risk entropy indices before and after remediation to obtain the cost-benefit ratio of multiple remediation schemes. Based on the cost-benefit ratio of multiple remediation schemes and the ecological risk entropy index after final remediation, the remediation scheme effectiveness matrix is ​​output. By setting screening rules, the optimal solution set of the repair schemes is obtained by considering the risk reduction efficiency and cost-effectiveness ratio in the repair scheme effectiveness matrix. The steps for analyzing the pollutant concentration distribution after the final remediation treatment are as follows: Based on the spatiotemporal dynamic risk warning level map, dynamic intervention parameters are set for different remediation strategies to obtain multi-scenario remediation schemes. Coupled hydraulic dynamic models are established for the multi-scenario remediation schemes. The models include depth-averaged shallow water equations. The depth-averaged flow velocity field is obtained by solving the equations simultaneously using a numerical algorithm. The depth-averaged flow velocity field is substituted into the pollutant degradation Monod equation. The pollutant degradation Monod equation is then solved spatiotemporally discretely using a numerical algorithm to obtain the final pollutant concentration distribution after remediation.

10. An ecological environment monitoring system based on geographic information technology, characterized in that... The method for monitoring the ecological environment based on geographic information technology as described in any one of claims 1-9 includes a heterogeneous data fusion module, an ecological health analysis module, a pollution diffusion module, an ecological risk prediction module, and a remediation scheme optimization module. The heterogeneous data fusion module is used to acquire multispectral images and point-based time-series data of water ecological protection target water areas, perform multi-source heterogeneous data fusion to obtain a three-dimensional water quality dynamic situation field, and send it to the ecological health analysis module. The ecological health analysis module is used to predict the future trajectory and analyze the water ecological health of the three-dimensional water quality dynamic field, obtain a heat map of the water ecological health index of the whole area, and send it to the pollution diffusion module. The pollution diffusion module is used to diffuse pollution from the heat map of the whole-area water ecological health index, obtain the pollutant trajectory matrix, and send it to the ecological risk prediction module. The ecological risk prediction module performs ecological risk prediction based on the pollutant trajectory matrix, obtains a spatiotemporal dynamic risk warning level map, and sends it to the remediation scheme optimization module. The repair scheme optimization module optimizes the repair scheme based on the spatiotemporal dynamic risk warning level map to obtain the optimal solution set of the repair scheme.

Images