A differential risk early warning method for flash flood disasters in small and medium-sized river basins with grading and zoning

Abstract

The present invention discloses a method for differential risk early warning of mountain flood disasters in small and medium-sized river basins. First, the risk areas of the river basin are identified through data analysis of mountain flood disasters; the basin lag time is calculated, and a simplified hydrodynamic model is used to simulate the cross-section water level and the basin inundation range, and the model is quantitatively analyzed; combined with unstructured grid division and risk source identification, a high-precision grid database is formed, and then the Gaussian mixture model (GMM) is used to cluster and classify the multi-dimensional features of the grid to obtain the risk levels of each grid in the basin. Finally, differential early warning thresholds and response strategies are formulated according to the grid partition and classification results, and dynamic update and linkage release are realized on the GIS platform in combination with measured data. This method takes into account the realistic conditions of limited data in small and medium-sized river basins and the accuracy requirements of model simulation, not only realizes the refined setting of early warning thresholds on the basis of partition and classification, but also significantly improves the accuracy of mountain flood risk identification.

Description

Technical Field

[0001] The present invention relates to the technical field of mountain flood disaster warning and prevention, and particularly relates to a differential risk warning method for grading and zoning mountain flood disasters in small and medium-sized river basins. Background Art

[0002] Mountain flood disasters are one of the most common and serious natural disasters in mountainous and hilly areas of our country. They are characterized by rapid onset, short duration, and uneven influence scope, posing a huge threat to villages, transportation arteries, infrastructure, etc. within small and medium-sized river basins. The suddenness and local nature of mountain flood disasters make the prevention work very challenging, especially in areas with complex terrain and rapidly changing rainfall intensity. Traditional mountain flood disaster warning methods often have certain limitations in terms of refined management and rapid response.

[0003] Currently, mountain flood disaster warning mainly relies on two types of methods: flood forecasting based on hydrological models and index determination based on mathematical statistics. Flood forecasting based on hydrological models usually simulates the rainstorm-flood evolution process within the river basin through a high-precision hydrological model, and triggers a warning with the simulated flow or water level threshold. However, due to the lack of observational data in small and medium-sized river basins, it is difficult to calibrate and verify model parameters, resulting in limited accuracy and reliability of the forecasting results in many cases. The index determination method based on mathematical statistics establishes empirical thresholds such as critical rainfall and critical water level, and combines the real-time monitored rainfall or water level changes to determine whether to initiate a warning. Although this method is easy to operate, it cannot fully consider the spatial differences within the river basin.

[0004] In addition, mountain flood disasters have significant spatial and temporal differences. Different regions within the same river basin vary significantly in terms of flood propagation speed, overflow range, and risk impact degree. For example, the upstream region is usually steep, with sudden rises and falls of floods and strong impact force, but the foresight period is short; the middle reaches have gentle river channels, a wider flood overflow range, and a more complex spatial distribution of risks; the downstream region has a longer flood propagation time, but a wide affected area and serious impact. Traditional warning methods use unified thresholds or single models, making it difficult to accurately match the risk characteristics of different regions, unable to scientifically distinguish key defense areas from ordinary areas, and also difficult to respond promptly to sudden disasters in different scenarios. Summary of the Invention

[0005] In view of the deficiencies of the existing technologies, the present invention aims to solve the problems of insufficient accuracy in regional risk assessment, inflexible response mechanisms, and unreasonable allocation of prevention and control resources existing in traditional mountain flood disaster warning methods, and proposes a differential warning method for mountain flood disasters with classification and zoning applicable to small and medium-sized river basins. By integrating basin lag time calculation, constructing a hydrodynamic model, grid-based risk assessment, and intelligent clustering algorithms, a dynamic warning system with classification and zoning is established. For different risk regions, differential warning thresholds are set, and the warning levels are dynamically adjusted in combination with meteorological forecasts and real-time monitoring data to ensure that high-risk regions can respond preferentially.

[0006] The present invention is realized through the following technical solutions: A differential risk warning method for mountain flood disasters with classification and zoning in small and medium-sized river basins, the method comprising the following steps:

[0007] Step S1, taking a small and medium-sized river basin as a unit, collecting underlying surface characteristics: topographic data, satellite images, and meteorological and hydrological data, analyzing the confluence path, basin slope, and roughness; combining the existing basic investigation results of mountain flood disasters, clarifying the preliminary distribution of risk regions within the basin, and obtaining design rainstorms and flood information through data analysis and derivation.

[0008] Step S2, after clarifying the basin area and underlying surface characteristics, calculating the basin lag time for the determined risk areas, approximating the time interval between the centroid of the flow process at the outlet section of the rainfall basin and the centroid of the corresponding net rainfall process as the average basin confluence time, and calculating the basin lag time using different formulas according to the basin area.

[0009] Step S3, calculating the water levels at different levels and the inundation range under the occurrence scenarios of mountain flood disasters at the dangerous sections of key villages for different flood frequencies by constructing a two-dimensional or three-dimensional hydrodynamic model.

[0010] Step S4, dividing the mountain flood risk area and identifying risk sources based on unstructured grids, dividing land areas and water areas on the basis of unstructured grids, clarifying the grid scale, carrying out flow direction analysis and terrain depression filling treatment on the study area based on DEM data and river system vectors, determining the basin boundary, and superimposing water conservancy facilities and potential hazard sources on the grid database, thereby forming a grid cell system and a risk source information matrix to support risk analysis.

[0011] Step S5, based on the Gaussian mixture model (GMM) clustering method, combining the flood simulation results, basin lag time, and regional risk source analysis, dynamically classifying and zoning the mountain flood hazard areas. The zoning results are displayed on a GIS platform after manual verification and optimization of administrative boundaries, and are dynamically updated in linkage with real-time water regime and rainfall data.

[0012] Step S6: For the classified and partitioned areas, set the flash flood warning thresholds differently and construct corresponding warning models. First, formulate thresholds that match the inundation range and basin lag time indicators for different risk levels, and achieve real-time linkage through meteorological forecasts and real-time monitoring data of rainfall and water levels. Combine with the GIS platform and the emergency management system to achieve linkage.

[0013] Further, the step S1 includes the following sub-steps:

[0014] Step S1-1: Summarize and organize the historical or potential flash flood disaster points in the basin in combination with historical documents, and initially delimit the scope of the risk area.

[0015] Step S1-2: Use high-resolution DEM or satellite images to extract the key underlying surface characteristic parameters of the basin: confluence path, slope, roughness, land use, and river network density, and perform format conversion and error correction on the data.

[0016] Step S1-3: Collect and organize the historical rainfall observation records, rainfall station distributions, and measured runoff data of hydrological stations in the basin; if there are radar precipitation products or meteorological forecast data, include them as well to provide multi-source support for subsequent flood simulations.

[0017] Step S1-4: Based on the obtained meteorological and hydrological data, use the regional storm frequency analysis method to derive the design storm intensity or rainfall process of the basin; combine with runoff generation and confluence calculations or verified empirical formulas to obtain the peak flow or design flood hydrograph at different design frequencies for subsequent model calculations and risk assessments.

[0018] Further, in step S2, different formulas are used to calculate the basin lag time according to the basin area. Specifically, the basin lag time is the time interval between the centroid of the flow process at the outlet section of the rainfall basin and the centroid of the corresponding net rainfall process, which is equivalent to the average basin confluence time during the basin confluence process. Determine the corresponding basin lag time for each risk area according to the confluence time; when the basin area is less than or equal to 50 km 2 , use the flood routing formula for iterative calculation; when the basin area is greater than 50 km 2 , use the NRCS equation for solution. Since the terrain in small and medium-sized basins varies greatly, use distributed calculation based on digital elevation.

[0019] Further, the step S3 includes the following sub-steps:

[0020] Step S3-1: Initialize the model and set boundary conditions, that is, initialize the hydrodynamic model according to the actual topographic features and data conditions of small and medium-sized river basins. Specifically, extract the longitudinal and cross-sectional information of the river channel in the basin through high-resolution DEM data, adjust the section spacing to adapt to the bending degree of the river channel and complex water flow; after extracting the underlying surface characteristics such as slope and river network density, specify the river width, river depth and cross-sectional shape for each section, and calculate the roughness coefficient value according to the Manning formula; the setting of boundary conditions is to use the design flood peak flow rate or flow process curve as the upstream boundary condition; the design flood peak flow rate is calculated by combining the regional rainfall intensity with the rainfall area and runoff coefficient; the downstream boundary condition is to use a constant water level or free outflow condition; when there are hydraulic structures, the downstream water level needs to be corrected through hydraulic calculations; the rainfall input part is mapped to the model grid through the rainfall intensity-duration curve; the river channel and the land area are coupled. The river channel area is divided into one-dimensional units with a grid size within 1 / 5 of the river width; the land area uses two-dimensional unstructured triangular grids to ensure the dynamic consistency of flood overflow and river channel backflow.

[0021] Step S3-2: Discretize the shallow water equations using the finite volume method to capture the non-linear characteristics of flood propagation; when performing multi-scenario simulations under different rainfall scenarios and flood frequencies, perform minimum time control through a parallel architecture.

[0022] Step S3-3: Uncertainty analysis. On the input side, perform Monte Carlo sampling on the design parameters: rainfall amount, rainfall duration, and boundary flow rate. Assume that the rainfall intensity follows the Chicago rainfall pattern distribution or normal distribution. By generating multiple sets of random samples, quantify the uncertainty of the water level simulation results; on the parameter side, for the key parameters of roughness coefficient, bottom bed morphology, and weir dams, use Bayesian statistics or sensitivity analysis tools for quantification, adjust the parameter perturbation range one by one, and evaluate its influence range on the simulation results. Through this analysis, the confidence interval of the water level simulation can be obtained.

[0023] Step S3-4: After the simulation is completed, generate the water level hydrographs of key sections under different return periods; superimpose the water depth calculation results on the DEM data, and output the flood inundation range and flow velocity distribution map, presented in vector format or raster format, providing input for subsequent grid division, risk assessment, and classification and zoning.

[0024] Furthermore, the step S4 includes the following sub-steps:

[0025] Step S4-1: Determine the watershed boundary. Specifically, based on the spatial distribution of flash flood warning targets, select the downstream position of the watershed to which the flash flood warning targets belong as the outlet section, and control the watershed area. Then, based on the DEM data of the study area, the spatial location data of flash flood warning targets, and the river system vector data, use ArcGIS hydro tools to carry out terrain depression filling, flow direction analysis, catchment capacity analysis, and digital river system generation, and delimit the watershed boundary according to the watershed outlet section.

[0026] Use the hydraulic engineering layout plans of river channel water boundaries, bridge piers, and water-blocking weirs and dams to cut the watershed boundary and clarify the land and water areas.

[0027] Step S4-2: Discretize the watershed space and divide the unstructured grid. Conduct watershed characteristic analysis, classify the land, water, and river channel grids, divide the unstructured grid, and check and adjust the grid.

[0028] Step S4-3: Identify and embed flash flood risk sources. Specifically, overlay the investigated potential hazard sources in the grid database; and give special marks to the grids where key bridges, traffic arteries, and other infrastructure are located. The potential hazard sources include but are not limited to debris flow gullies, steep cliffs, and weirs and dams.

[0029] After completing the division of the unstructured grid, import the DEM data, basic hydrological data, and rainfall data into the grid, and assign corresponding physical properties according to the characteristics of different regions. The characteristics of different regions include but are not limited to soil type, vegetation cover, slope, roughness, and underlying surface.

[0030] Furthermore, step S4-2 further includes the following sub-steps:

[0031] Step S4-2a: Watershed characteristic analysis. Specifically, pay special attention to the flash floods in small and medium-sized watersheds vulnerable to local heavy precipitation, including precipitation distribution, slope changes in the watershed terrain, and characteristics of the river channels in the watershed.

[0032] Step S4-2b: Classify the land, water, and river channel grids. Specifically, for the upstream areas and large-slope areas in the mountainous regions, reduce the grid size to finely simulate the changes in water flow and the process of precipitation infiltration; but for low-slope areas and catchment areas, increase the grid size; for the river channel parts, river channel sharp bends, and confluence areas of small flash flood watersheds, high-resolution grids are required for simulation.

[0033] Step S4-2c: Unstructured grid division. Specifically, starting from the watershed boundary, select grid nodes and conduct triangular meshing based on the Delaunay criterion, and ensure that the interior angles of the triangles are maximized.

[0034] Step S4-2d: After unstructured grid division, grid inspection and dynamic adjustment are carried out by manual identification to ensure that the shape of each grid cell is reasonably set.

[0035] Further, the said step S5 includes the following sub-steps:

[0036] Step S5-1: Establishment of multi-dimensional data feature vectors and Gaussian clustering GMM of grids, including constructing multi-dimensional feature vectors for each grid cell, constructing risk factors and mountain flood risk quantification indicators for each grid, defining the number of clusters, calculating the belonging probability, and fitting the data;

[0037] Step S5-2: Hierarchical optimization and GIS display, that is, after Gaussian clustering is completed, the classification results of the dangerous areas are optimized and displayed; First, the clustering results are manually checked, and combined with historical disaster records, the classification results that deviate from the actual situation are adjusted; Subsequently, the clustering results are overlaid with the administrative region boundaries, and the partitions are refined according to the township or village boundaries to make the partition results meet the actual management requirements.

[0038] Further, the said step S5-1 also includes the following sub-steps:

[0039] Step S5-1a: Construct multi-dimensional feature vectors for each grid cell, that is, the features of each grid include multiple dimensions of risk factors related to mountain flood disasters, including hydraulic characteristic parameters, flood forecast period, and risk source information parameters;

[0040] Step S5-1b: Construct the non-linear relationship between risk factors and mountain flood risk quantification indicators for each grid; The feature vector of each grid is represented as a multi-dimensional database containing risk factors and mountain flood risk determination index parameters;

[0041] Step S5-1c: Define the number of clusters, that is, by defining the risk level, and then define the number of clusters, the number of clusters is defined according to the basin area and the degree of differentiation; Then initialize the parameters of each cluster, including: mean, covariance matrix, and weight;

[0042] Step S5-1d: Based on the feature data of each grid: mean, covariance matrix, and weight, calculate the posterior probability of each grid cell belonging to each Gaussian distribution, so as to calculate the belonging probability of each grid cell in different clusters;

[0043] Step S5-1e: Optimize the partition results by maximizing the probability that the grid points belong to multiple Gaussian distributions, so that each cluster can better fit the data, and this process is achieved by maximizing the likelihood function, and its expression is: ; where, N represents the total number of samples in the dataset, that is, the number of data points; is the set of all parameters, , K is the total number of clusters (groups); Represents the i-th sample data point in the data set; represents the weight of each Gaussian distribution, For the The probability density function of a Gaussian distribution is and are the mean vector and covariance matrix respectively;

[0044] Step S5-1f, alternately perform steps S5-1d and S5-1e for multiple iterations until the mean of the model is , covariance matrix And weight It becomes stable or reaches the maximum number of iterations. Each iteration updates the grid's attribution probability based on the current model and gradually optimizes the cluster division.

[0045] In step S5-1g, the GMM model outputs the flash flood risk classification result of each grid, that is, the cluster to which each grid belongs.

[0046] Furthermore, the step S6 also includes the following sub-steps:

[0047] Step S6-1, differentiated setting of warning thresholds, i.e., based on the hierarchical zoning results of flash flood hazard areas, combined with the flood inundation range, basin lag time, key section water level process and time required for transfer in each area, determine the zoning warning thresholds, and establish differentiated warning trigger conditions for risk areas of different levels in the zone by combining historical data and field survey results;

[0048] Step S6-2, design and implementation of differentiated early warning modes, that is, after completing the setting of early warning thresholds, differentiated early warning response modes are designed in combination with the risk characteristics of different regions. By introducing weather forecasts, rainfall monitoring, water level monitoring and other multi-source real-time data, a hierarchical early warning mechanism is established, and the early warning level is dynamically adjusted;

[0049] Step S6-3: Release and linkage management of warning results, that is, the warning information generated by the differentiated warning mode needs to be released within the region and linked with the emergency management system; the warning information is released through multiple channels, including text messages, broadcasts, warning terminal display screens and social media platforms, to ensure that different groups can obtain warning signals in a timely manner; and combined with the linkage function of the GIS platform and the warning system, the dynamically updated warning information is shared in real time with management departments at all levels to dispatch and coordinate resources.

[0050] Furthermore, the step S6-1 further includes the following sub-steps:

[0051] Step S6-1a: Define corresponding warning indicators and threshold ranges for different risk-level regions;

[0052] Step S6-1b: Use the weight analysis method to quantitatively allocate the influence degrees of different indicators to calibrate the warning thresholds; the different indicators include water depth, flow velocity, rainfall, and basin lag time;

[0053] Step S6-1c: Verify the setting of the regional warning threshold with the measured data, and calibrate the warning conditions in real time through retrospective analysis of historical flash flood events.

[0054] The beneficial effects of the present invention are as follows:

[0055] Through technical means such as basin lag time calculation, simplified hydrodynamic model, and grid-based risk assessment, combined with the Gaussian clustering algorithm, the present invention realizes the hierarchical and zonal classification and differential risk warning of flash flood disasters, significantly improving the accuracy and scientificity of risk assessment. By setting differential warning thresholds and dynamic adjustment mechanisms, it is possible to optimize the warning response according to different risk levels and regional characteristics, ensure the priority prevention and control of high-risk areas, and improve the accuracy and timeliness of warnings. At the same time, by linking meteorological forecasts, real-time monitoring data with the model, and dynamically updating the warning level, the intelligent level and emergency response ability of warnings are significantly enhanced. The present invention is applicable to the prevention and control of flash flood disasters in small and medium-sized basins under various terrain conditions, has wide applicability and popularization value, provides scientific support for improving the level of disaster prevention and mitigation and optimizing resource allocation, and has important social and economic benefits. Description of the Drawings

[0056] Figure 1 is the flowchart of the method for hierarchical and zonal classification and differential risk warning of flash flood disasters in small and medium-sized basins of the present invention. Detailed Embodiments

[0057] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application.

[0058] The present invention uses the method of hierarchical and zonal classification and differential risk warning to improve the effect of flash flood disaster prevention. By comprehensively considering the topographic features, flood propagation process, and risk distribution law within the basin, and setting differential warning modes and response mechanisms for different regions, it is possible to more scientifically match the warning time, transfer time, and prevention measures, thereby significantly improving the accuracy and timeliness of warnings. At the same time, by clarifying the danger levels of different regions through hierarchical and zonal classification, it is possible to effectively optimize resource allocation and ensure that the prevention needs of key regions are met. The specific technical solutions are implemented as follows:

[0059] As Figure 1 shown, the present invention provides a method for differentiated risk early warning of mountain flood disasters at different levels and in different regions in small and medium-sized river basins, including the following steps:

[0060] Step S1: Taking small and medium-sized river basins as units, collect topographic data, satellite images and meteorological and hydrological data, and analyze underlying surface characteristics such as confluence paths, basin slopes and roughness coefficients. Combining the existing basic investigation results of mountain flood disasters, clarify the preliminary distribution of risk areas within the basin. And through data analysis and derivation of design rainstorms and design floods, lay a foundation for subsequent model calculations.

[0061] Step S2: After clarifying the basin area and underlying surface characteristics, calculate the basin lag time for the determined risk areas. Approximate the time interval between the centroid of the rainfall-runoff process at the basin outlet section and the centroid of the corresponding net rainfall process as the average basin confluence time, and calculate the basin lag time using different formulas according to the basin area.

[0062] Step S3: Calculate the water levels at different grades and the inundation ranges under the scenarios of mountain flood disasters at dangerous sections of key villages for different flood frequencies by constructing a two-dimensional or three-dimensional hydrodynamic model.

[0063] Step S4: Divide the mountain flood risk areas and identify risk sources based on unstructured grids. Divide the land area and water area on the basis of unstructured grids, clarify the grid scale, conduct flow direction analysis and terrain filling depression treatment on the study area based on DEM (Digital Elevation Model data) and river system vectors, determine the basin boundary, and superimpose water conservancy facilities such as bridges and weirs and potential hazard sources on the grid database, so as to form a grid unit system and a risk source information matrix to support risk analysis.

[0064] Step S5: Based on the Gaussian clustering method (Gaussian Mixture Model, GMM), combine the flood simulation results, basin lag time and regional risk source analysis to dynamically classify and partition the mountain flood hazard areas. The partition results are displayed on the GIS platform after manual verification and administrative boundary optimization, and are linked with real-time water regime and rainfall data to achieve dynamic update.

[0065] Step S6: For the classified and partitioned regions, differentially set mountain flood warning thresholds and construct corresponding warning modes. First, formulate thresholds matching indicators such as inundation range and basin lag time for different risk levels, and realize the linkage through the real-time linkage of meteorological forecasts and rainfall and water level monitoring data, and combine the GIS platform and the emergency management system.

[0066] Furthermore, the specific content of step S1 is as follows:

[0067] Step S1-1: Combine historical documents, existing investigation reports, and on-site surveys to summarize and organize the existing or potential flash flood disaster points in the basin, and preliminarily delimit the scope of the risk area.

[0068] Step S1-2: Use high-resolution DEM or satellite images to extract key underlying surface characteristic parameters such as the confluence path, slope, roughness coefficient, land use, and river network density of the basin, and perform necessary format conversion and error correction on the data.

[0069] Step S1-3: Collect and organize the historical rainfall observation records, rainfall station distributions, and measured runoff data of hydrological stations in the basin; if there are radar precipitation products or meteorological forecast data, include them as well to provide multi-source support for subsequent flood simulations.

[0070] Step S1-4: Based on the obtained meteorological and hydrological data, use the regional storm frequency analysis method (such as P-III distribution, Lognormal, etc.) to derive the design storm intensity or rainfall process of the basin; combine appropriate runoff yield and concentration calculations or verified empirical formulas to obtain the peak flood flow or design flood hydrograph at different design frequencies, laying a quantitative foundation for model calculations and risk assessments.

[0071] Further, the specific steps of step S2 are as follows:

[0072] The basin lag time is the time interval between the centroid of the flow process at the outlet section of the rainfall basin and the centroid of the corresponding net rainfall process, which is equivalent to the average basin confluence time during the basin confluence process. Determine the corresponding basin lag time for each risk area according to the confluence time.

[0073] Optionally, when the basin area is less than or equal to 50 km 2 the iterative calculation is performed using the Zhejiang design flood inference formula.

[0074] Optionally, when the basin area is greater than 50 km 2 the NRCS equation is used for solution, considering the characteristics of large terrain variations in small and medium-sized basins, and distributed calculations are performed based on digital elevation.

[0075] Further, the specific steps of step S3 are as follows:

[0076] Step S3-1, initialize the model and set boundary conditions: Initialize the hydrodynamic model according to the actual topographic features and data conditions of small and medium-sized river basins. First, extract the longitudinal and cross-sectional information of the river channels in the basin through high-resolution DEM data, and adjust the cross-section spacing to 50 to 200 meters to adapt to the curvature of the river channels and the complexity of the water flow. After extracting the underlying surface features such as slope and river network density, specify the river width, river depth, and cross-sectional shape (such as rectangular, trapezoidal, or compound cross-section) for each cross-section, and calculate the roughness coefficient value according to the Manning formula. The typical roughness coefficient value range is 0.020 to 0.025 for sandy riverbeds and 0.030 to 0.035 for rocky riverbeds, and the specific values can be adjusted according to on-site investigations.

[0077] In the setting of boundary conditions, the design flood peak flow rate or the flow hydrograph is adopted as the upstream boundary condition. The design flood peak flow rate can be calculated by combining the regional rainfall intensity with the rainfall area and runoff coefficient. The downstream boundary condition usually adopts a constant water level or a free outflow condition. However, if there are hydraulic structures such as weirs and dams, the downstream water level needs to be corrected through hydraulic calculations. The rainfall input part is mapped to the model grid through the rainfall intensity-duration curve. The river channel and the land area are treated in a coupled manner. The river channel area is divided into one-dimensional units, and the grid size is within 1 / 5 of the river width; the land area adopts two-dimensional unstructured triangular grids, and the grid size is 10 to 30 meters to ensure the dynamic consistency of flood overflow and river channel backflow.

[0078] Step S3-2, discretize the shallow water equations using the finite volume method to capture the nonlinear characteristics of flood propagation. When performing multi-scenario simulations under different rainfall scenarios and flood frequencies, the computational time can be significantly reduced through a parallel architecture. For example, when the basin area is less than 50 square kilometers, the time to complete one simulation is controlled within 1 hour.

[0079] Step S3-3, uncertainty analysis. Specifically, on the input side, perform Monte Carlo sampling on parameters such as the design rainfall amount, rainfall duration, and boundary flow rate. Assume that the rainfall intensity follows the Chicago rainfall pattern distribution or the normal distribution. By generating multiple sets of random samples, quantify the uncertainty of the water level simulation results;

[0080] On the parameter side, for key parameters such as roughness coefficient, bottom bed morphology, and weirs and dams, use Bayesian statistics or sensitivity analysis tools for quantification. Adjust the parameter perturbation range one by one, and evaluate its influence range on the simulation results. Through this analysis, the confidence interval of the water level simulation can be obtained.

[0081] After the simulation is completed in Step S3-4, generate the water level hydrographs of key cross-sections under different recurrence periods (such as 5 years, 20 years, 50 years). Superimpose the water depth calculation results on the DEM data, and output the flood inundation range and velocity distribution maps, presented in vector format or raster format. These results provide accurate inputs for subsequent grid division, risk assessment, and classification and zoning.

[0082] Furthermore, step S4 is specifically as follows:

[0083] Step S4-1, determination of the watershed boundary. According to the spatial distribution of flash flood warning targets (Si), select the downstream position of the watershed to which the flash flood warning target belongs as the outlet section, and try to ensure that the controlled watershed area is less than 200 km². Based on the DEM data of the study area, the spatial position data of flash flood warning targets, and the river system vector data, use ArcGIS hydro tools to carry out terrain depression filling, flow direction analysis, catchment capacity analysis, digital river system generation, etc., and delimit the watershed boundary based on the watershed outlet section.

[0084] Use the plane layout diagrams of water conservancy projects such as river channel water boundaries, bridge piers, and water-blocking weirs to cut the watershed boundary and clarify the land and water areas.

[0085] Step S4-2, discrete disposal of the watershed space and unstructured grid division, specifically including the following sub-steps:

[0086] Step S4-2a, watershed characteristic analysis. Flash floods in small and medium-sized watersheds are usually affected by local heavy precipitation, and the terrain is complex with large slopes. Therefore, special attention should be paid to the precipitation distribution, slope changes in the watershed terrain, and the characteristics of the river channels within the watershed.

[0087] Step S4-2b, grid grading for land, water, and river channels. For the upper reaches of mountainous areas and areas with large slopes, it is recommended to use smaller grid sizes (such as about 10 meters) to finely simulate the changes in water flow and the precipitation infiltration process. For low-slope areas and catchment areas, the grids can be slightly larger (not more than 30 meters) to improve the calculation efficiency. For the river channel part of small flash flood watersheds, the grids should be more dense, usually not more than 5-10 meters. Especially in areas with large water flow changes such as river channel sharp bends and tributary confluences, higher-resolution grids need to be used for simulation.

[0088] Step S4-2c, unstructured grid division. Starting from the watershed boundary, select grid nodes and perform triangular meshing based on the Delaunay criterion (ensuring the maximization of the interior angles of triangles). Preferably, in key areas such as flood propagation paths and runoff convergence areas, use smaller grids to refine the simulation accuracy and ensure accurate capture of the hydrological processes of the watershed.

[0089] Step S4-2d, after unstructured grid division, conduct grid inspection and dynamic adjustment through manual identification to ensure that the shape of each grid cell is reasonable and there are no overly long and narrow or extreme shapes.

[0090] Step S4-3, Identification and embedding of mountain flood risk sources. Specifically, overlay the potential hazard sources (such as debris flow outlets, steep cliffs, weirs and dams) obtained from the investigation on the grid database; further, give special marks to the grids where infrastructure such as key bridges and traffic arteries are located.

[0091] After completing the unstructured grid division in Step S4-4, import DEM, basic hydrological data, rainfall data, etc. into the grid, and assign corresponding physical properties (such as soil type, vegetation cover, slope, roughness, underlying surface, etc.) according to the characteristics of different regions.

[0092] Further, Step S5 is specifically as follows:

[0093] Step S5-1, Establishment of multi-dimensional data feature vectors and Gaussian mixture model (GMM) clustering of grids, which specifically includes the following sub-steps:

[0094] Step S5-1a, Construct multi-dimensional feature vectors for each grid cell. The features of each grid include multiple dimensions of risk factors related to mountain flood disasters, including hydraulic characteristic parameters, flood forecast period, risk source information parameters, etc.

[0095] Specifically, the calculation results of the hydrodynamic model in Step S3 can provide grid-level hydraulic characteristic parameters, such as dynamic hydrological data such as maximum water depth, maximum flow velocity and flood arrival time; the basin lag time in Step S2 can calculate the flood forecast period of each grid area; the regional risk source information in Step S4 (such as population density, distribution of important facilities and historical disaster points, etc.) can further supplement the spatial and attribute integrity of the data. These features are used to create a multi-dimensional data vector of risk factors for each grid cell.

[0096] Step S5-1b, Construct the non-linear relationship between risk factors (hydraulic characteristic parameters, risk source information) and the mountain flood risk quantification indicators of each grid (different levels of water levels, flood forecast period, inundation range under the scenario of mountain flood disasters are determined at dangerous sections of key villages for different flood frequencies); finally, the feature vector of each grid can be represented as a multi-dimensional database containing parameters such as risk factors and mountain flood risk judgment indicators.

[0097] Step S5-1c, Define the number of clusters. Suppose we have 3 risk levels (high risk, medium risk, low risk), so the number of clusters is K = 3 (the number of clusters is defined according to the basin area and the degree of differentiation, and can also be 4, 5, 6, etc.). Then, initialize the parameters of each cluster, including: mean, covariance matrix, weight.

[0098] Step S5-1d: Based on the feature data (mean, covariance matrix, weight) of each grid, calculate the posterior probability of each grid cell belonging to each Gaussian distribution, and thus calculate the membership probability of each grid cell in different clusters (risk levels).

[0099] Step S5-1e: Optimize the partitioning result by maximizing the probability that grid points belong to multiple Gaussian distributions, so that each cluster can better fit the data. This is achieved by maximizing the likelihood function, and the expression is: ; where N represents the total number of samples in the dataset, that is, the number of data points; is the set of all parameters, , K is the total number of clusters (groupings); represents the i-th sample data point in the dataset; represents the weight of each Gaussian distribution, is the -th probability density function of the Gaussian distribution, and are the mean vector and covariance matrix respectively;

[0100] Step S5-1f: Alternately perform multiple iterations of Step S5-1d and Step S5-1e until the mean of the model, covariance matrix and weight tend to be stable, or reach the set maximum number of iterations; where in each iteration, the membership probability of the grid is updated according to the current model, and the partitioning of the clusters is gradually optimized;

[0101] Step S5-1g: The GMM model will output the mountain flood risk classification result of each grid, that is, which cluster (high risk, medium risk or low risk) each grid belongs to. For example, the grid is divided into three categories: high risk, medium risk and low risk. Among them, the high-risk area is characterized by high water depth (such as ≥1.5m), fast flow velocity (such as ≥2.5m / s), short arrival time (such as ≤30 minutes), and high population density and facility exposure, while the low-risk area is characterized by shallow water depth, slow flow velocity and long arrival time.

[0102] Step S5-2: Hierarchical optimization and GIS display: After completing Gaussian clustering, optimize and display the classification results of the dangerous areas. First, manually check the clustering results, and combine expert experience or historical disaster records to adjust the classification results that deviate from the actual situation. For example, when the risk level of some grids does not match their historical disaster records, they can be reclassified according to the actual situation. Subsequently, the clustering results are overlaid with the administrative region boundaries, and the partitions are refined according to the township or village boundaries, so that the partition results meet the actual management requirements.

[0103] The zoning results are visually presented in different colors on the GIS platform to distinguish high, medium, and low-risk areas. Through the linkage with real-time rainfall monitoring and water level monitoring data, the risk zone classification can be dynamically updated. For example, when an extreme rainfall event occurs, the real-time data triggers the model to recalculate or call the pre-simulated results, quickly adjusting the zoning level to ensure the timeliness of early warning response. The zoning results include vector format hazard zone layer files and dynamic update reports, which are used to support flood control command and emergency decision-making.

[0104] Through the above process, the zoning method for mountain flood hazard zone classification based on Gaussian clustering realizes scientific zoning and dynamic adjustment, provides technical support for the precise prevention and control of hazard zones, and enhances applicability and response efficiency in actual management.

[0105] Furthermore, step S6 is specifically as follows:

[0106] Step S6-1, differential setting of early warning thresholds: Based on the zoning results of mountain flood hazard zones, comprehensively considering the flood inundation range, basin lag time, water level process at key sections, and time required for evacuation in each region, determine the zoning early warning thresholds. By combining historical data and on-site investigation results, establish differential early warning trigger conditions for different risk level regions within the zone. Specifically, it includes the following sub-steps:

[0107] Step S6-1a, define corresponding early warning indicators and threshold ranges for different risk level regions (Level I, Level II, Level III). For example, for Level I high-risk regions, prioritize the characteristic of short flood arrival time, and set strict rainfall and water level thresholds. For example, when the rainfall intensity exceeds the critical value (such as 50 mm / h) or the cross-section water level exceeds 1.5 meters, immediately initiate the early warning; for Level II and Level III regions, relatively loose trigger conditions can be set, with emphasis on the flood overflow range and the time required for personnel evacuation.

[0108] Step S6-1b, adopt a weight analysis method (such as the analytic hierarchy process or entropy weight method) to quantitatively allocate the influence degree of different indicators (such as water depth, flow velocity, rainfall, basin lag time) to calibrate the early warning thresholds.

[0109] Step S6-1c, verify the setting of regional early warning thresholds with measured data, and calibrate the scientificity and applicability of early warning conditions through retrospective analysis of historical mountain flood events.

[0110] Step S6-2, design and implementation of differential early warning modes: After completing the setting of early warning thresholds, design differential early warning response modes in combination with the risk characteristics of different regions. By introducing multi-source real-time data such as meteorological forecasts, rainfall monitoring, and water level monitoring, establish a hierarchical early warning mechanism and dynamically adjust the early warning level.

[0111] For high-risk areas of level I, focus on monitoring the impact of short-term heavy rainfall. Combining radar precipitation prediction and the measured data of rain gauges, set up a rapid response mechanism. When the cumulative rainfall or the monitored water level approaches the warning threshold, trigger early warnings and simultaneously transmit information downstream. During this process, high-frequency data acquisition devices and communication networks can be utilized to ensure the timely transmission and execution of warning signals.

[0112] For areas of level II, adopt a combined trigger mode. Determine whether to initiate a warning through comprehensive analysis of rainfall and water levels. When the water level at the monitoring station continuously exceeds the threshold and the cumulative rainfall exceeds the warning line, enter the warning state.

[0113] For areas of level III, the cumulative rainfall is the main trigger condition. Combine the evaluation of the inundation range to determine the activation time and provide a sufficient response window.

[0114] To enhance the flexibility of warning responses, a dynamic adjustment mechanism can be introduced into the warning mode. When extreme rainfall events or abnormal hydrological changes occur, combine real-time monitoring data with model simulation results to quickly adjust the warning level and trigger conditions. For example, when the monitoring data exceeds the preset threshold range, the warning system automatically upgrades the warning level for the corresponding area and prompts the upgrade of defense measures.

[0115] Step S6-3, Release and linkage management of warning results: The warning information generated through the differentiated warning mode needs to be quickly released within the regional scope and carry out linkage work in combination with the emergency management system. The release of warning information can adopt multiple channels, including text messages, broadcasts, warning terminal displays, social media platforms, etc., to ensure that different groups can obtain warning signals in a timely manner.

[0116] In high-risk areas, the warning information should include the specific arrival time of floods, the expected inundation range, suggestions for evacuation routes, etc., to guide the personnel and facilities in relevant areas to enter the emergency state in advance;

[0117] In medium- and low-risk areas, the information released can focus more on preventive measures and suggestions for dynamic attention;

[0118] Furthermore, combined with the linkage function of the GIS platform and the warning system, share the dynamically updated warning information with all levels of management departments in real time, so that the flood control command center can quickly dispatch resources and coordinate all parties.

[0119] Through the above steps, the setting of differentiated warning thresholds, mode design, and release management achieve precise prevention and control of different regions and different risk levels, providing a reliable guarantee for the effective defense of mountain flood disasters.

[0120] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all fall within the protection scope of the present invention.

Claims

1. A differentiated risk warning method for flash flood disasters in small and medium-sized river basins based on classification and zoning, characterized in that: The method comprises the following steps: Step S1, collect the underlying surface characteristics with small and medium-sized watersheds as units; clarify the preliminary distribution of risk areas in the watershed, and analyze and infer the design rainstorm and flood information through data analysis; Step S2, calculate the basin delay time for the determined risk area, and use different formulas to calculate the basin delay time according to the basin area; the basin delay time is calculated according to different formulas according to the basin area, specifically: the basin delay time is the time distance between the centroid of the flow process of the outlet section of the rainfall basin and the centroid of the corresponding net rain process, which can be equivalent to the average confluence time of the basin in the basin confluence process, and the basin delay time corresponding to each risk area is determined according to the confluence time; when the basin area is less than or equal to 50km 2 When the basin area is greater than 50km 2 The NRCS equation is used to solve the problem. Since the terrain of small and medium-sized watersheds varies greatly, distribution calculation based on digital elevation is used. Step S3, by constructing a two-dimensional or three-dimensional hydrodynamic model to calculate different flood frequencies at dangerous sections of key villages to determine the inundation range under different levels of water levels and flash flood disaster scenarios; Step S4, dividing the flash flood risk area and identifying the risk source based on the unstructured grid, dividing the land and water areas on the basis of the unstructured grid, clarifying the grid scale and determining the basin boundary, and then obtaining the grid unit system and risk source information matrix; Step S5, combining flood simulation results, basin lag time and regional risk source analysis, dynamically classify and partition the flash flood danger areas, and the partition results are displayed on the GIS platform and dynamically updated in conjunction with real-time water and rainfall data; Step S6, for the classified and zoned areas, differentiated flash flood warning thresholds are set and corresponding warning modes are constructed, and the release and linkage management of warning results are carried out; specifically, the following sub-steps are included: Step S6-1, differentiated setting of warning thresholds, i.e., based on the hierarchical zoning results of flash flood hazard areas, combined with the flood inundation range, basin lag time, key section water level process and time required for transfer in each area, determine the zoning warning thresholds, and establish differentiated warning trigger conditions for risk areas of different levels in the zone by combining historical data and field survey results; Step S6-2, design and implementation of differentiated early warning modes, that is, after completing the setting of early warning thresholds, differentiated early warning response modes are designed in combination with the risk characteristics of different regions. By introducing weather forecasts, rainfall monitoring, water level monitoring and other multi-source real-time data, a hierarchical early warning mechanism is established, and the early warning level is dynamically adjusted; Step S6-3 is the release and linkage management of warning results, that is, the warning information generated by the differentiated warning mode needs to be released within the region and linked with the emergency management system; the warning information is released through multiple channels, including text messages, broadcasts, warning terminal display screens and social media platforms; and combined with the linkage function of the GIS platform and the warning system, the dynamically updated warning information is shared in real time with management departments at all levels.

2. According to claim 1, a differentiated risk warning method for mountain torrent disasters in small and medium-sized river basins is characterized by: The step S1 comprises the following sub-steps: Step S1-1, summarize and sort out the flash flood disaster points that have occurred or are potential in the basin in combination with historical documents, and preliminarily define the scope of the risk area; Step S1-2, using high-resolution DEM or satellite images, extracting key underlying surface characteristic parameters of the watershed: confluence path, slope, roughness, land use and river network density, and performing data format conversion and error correction; Step S1-3, collect and organize the rainfall observation records, rainfall station distribution, and measured runoff data of the watershed over the years; if there are radar precipitation products or weather forecast data, they are also included to provide multi-source support for subsequent flood simulation; Step S1-4, based on the acquired meteorological and hydrological data, a regional rainstorm frequency analysis method is used to deduce the design rainstorm intensity or rainfall process of the basin; Combined with runoff calculations or verified empirical formulas, peak flows or design flood process lines under different design frequencies are obtained for subsequent model calculations and risk assessments.

3. The method for differentiated risk warning of flash flood disasters in small and medium-sized river basins according to claim 1 is characterized in that: The step S3 comprises the following sub-steps: Step S3-1, initialize the model and set boundary conditions, that is, initialize the hydrodynamic model according to the actual terrain characteristics and data conditions of the small and medium-sized watersheds; specifically, extract the longitudinal and cross-sectional information of the river channel of the watershed through high-resolution DEM data, and adjust the section spacing to adapt to the curvature of the river channel and complex water flow; after extracting the underlying surface characteristics of the slope and river network density, specify the river width, river depth and cross-sectional shape for each section, and calculate the roughness value according to the Manning formula; The boundary conditions are set by using the designed flood peak flow or flow process curve as the upstream boundary conditions; the designed flood peak flow is calculated by combining the regional rainstorm intensity with the rainfall area and the runoff coefficient; the downstream boundary condition is to use a constant water level or a free outflow condition; when there are hydraulic structures, the downstream water level needs to be corrected by hydraulic calculation; the rainfall input part is mapped to the model grid through the rainfall intensity-duration curve; the river channel and the land area are coupled, the river channel area is divided into one-dimensional units, and the grid size is within 1 / 5 of the river width; the land area uses a two-dimensional unstructured triangular grid to ensure the dynamic consistency of flood overflow and river channel backflow; Step S3-2, using the finite volume method to discretize the shallow water equations to capture the nonlinear characteristics of flood propagation; when performing multi-scenario simulations under different rainfall scenarios and flood frequencies, the parallel architecture is used for minimum time control; Step S3-3, uncertainty analysis, that is, on the input side, Monte Carlo sampling is performed on the design parameters: rainstorm amount, rainfall duration and boundary flow, assuming that the rainfall intensity obeys the Chicago rain pattern distribution or normal distribution, and multiple groups of random samples are generated to quantify the uncertainty of the water level simulation results; on the parameter side, Bayesian statistics or sensitivity analysis tools are used to quantify the key parameters of roughness, bed morphology and weirs, and the parameter disturbance range is adjusted one by one to evaluate its impact on the simulation results. Through this analysis, the confidence interval of the water level simulation can be obtained; Step S3-4, after the simulation is completed, the water level process lines of key sections under different recurrence periods are generated; the water depth calculation results are superimposed on the DEM data, and the flood inundation range and flow velocity distribution map are output and presented in vector format or raster format to provide input for subsequent grid division, risk assessment and hierarchical zoning.

4. The method for differentiated early warning of flash flood disasters in small and medium-sized river basins according to claim 1 is characterized in that: The step S4 comprises the following sub-steps: Step S4-1, basin boundary determination, that is, according to the spatial distribution of flash flood warning objects, the downstream position of the basin to which the flash flood warning object belongs is selected as the outlet section, and the basin area is controlled; then, based on the DEM data of the study area, the spatial location data of the flash flood warning object, and the vector data of the river system, ArcGIS hydro tools are used to carry out terrain filling, flow direction analysis, water collection capacity analysis, and digital water system generation, and the basin boundary is delineated according to the basin outlet section; the basin boundary is cut using the plan layout of the water conservancy project of the river water boundary, bridge piers, and water blocking weirs, and the land and water areas are detailed; Step S4-2: Discretization of watershed space and unstructured grid division, watershed characteristic analysis, land, water and river grid classification, unstructured grid division, and grid inspection and adjustment; Step S4-3, flash flood risk source identification and embedding, specifically, the potential hazard sources obtained from the investigation are superimposed on the grid database; and special marks are given to the grids where key bridges, traffic arteries and other infrastructure are located; the potential hazard sources include but are not limited to debris flow gullies, steep cliffs and weirs; Step S4-4, after completing the unstructured grid division, the DEM data, basic hydrological data and rainfall data are imported into the grid, and corresponding physical properties are assigned according to the characteristics of different areas. The characteristics of different areas include but are not limited to soil type, vegetation cover, slope, roughness and underlying surface.

5. The method for differentiated early warning of flash flood disasters in small and medium-sized river basins according to claim 4 is characterized in that: The step S4-2 also includes the following sub-steps: Step S4-2a, watershed characteristics analysis, i.e., paying special attention to mountain torrents in small and medium-sized watersheds that are susceptible to local heavy precipitation, including precipitation distribution, slope changes in watershed terrain, and characteristics of river channels within the watershed; Step S4-2b, land, water, and river grid classification, that is, the grid size is reduced for the upstream of the mountainous area and the area with a large slope; but for the low-slope area and the catchment area, the grid size is enlarged; for the river part of the mountain torrent small watershed, the river bend, and the confluence of tributaries, high-resolution grids are required for simulation; Step S4-2c, unstructured meshing, i.e. starting from the boundary of the watershed, selecting mesh nodes, performing triangulation based on the Delaunay criterion, and ensuring that the inner angle of the triangle is maximized; Step S4-2d, after the unstructured grid is divided, the grid is checked and dynamically adjusted by manual identification to ensure that the shape of each grid unit is reasonably set.

6. The method for early warning of differentiated risk of flash flood disasters in small and medium-sized river basins according to claim 1 is characterized in that: The step S5 comprises the following sub-steps: Step S5-1, establishment of multidimensional data feature vector and Gaussian clustering GMM of grids, including construction of multidimensional feature vector for each grid unit, construction of risk factors and quantitative indicators of flash flood risk for each grid, definition of the number of clusters, calculation of attribution probability and fitting of data; Step S5-2, classification optimization and GIS display, that is, after completing Gaussian clustering, the classification results of the danger zone are optimized and displayed; first, the clustering results are manually checked, and the classification results that deviate from the actual situation are adjusted in combination with historical disaster records; Subsequently, the clustering results are superimposed on the administrative district boundaries, and the divisions are refined according to township or village boundaries to make the zoning results meet actual management needs.

7. The method for differentiated early warning of flash flood disasters in small and medium-sized river basins according to claim 6 is characterized in that: The step S5-1 further comprises the following sub-steps: Step S5-1a, constructing a multi-dimensional feature vector for each grid unit, that is, the characteristics of each grid contain multiple dimensions of risk factors related to flash flood disasters, including hydraulic characteristic parameters, flood forecast period, and risk source information parameters; Step S5-1b, constructing a nonlinear relationship between risk factors and flash flood risk quantitative indicators of each grid; the characteristic vector of each grid is represented as a multidimensional database containing risk factors and flash flood risk determination indicator parameters; Step S5-1c, defining the number of clusters, that is, defining the risk level and then defining the number of clusters, wherein the number of clusters is defined according to the watershed area and the degree of differentiation; Then initialize the parameters of each cluster, including: mean, covariance matrix and weight; Step S5-1d, based on the feature data of each grid: mean, covariance matrix and weight, by calculating the posterior probability that each grid unit belongs to each Gaussian distribution, thereby calculating the probability of each grid unit belonging to different clusters; Step S5-1e, optimize the partitioning result by maximizing the probability that the grid points belong to multiple Gaussian distributions so that each cluster can better fit the data. This process is achieved by maximizing the likelihood function, which is expressed as: Where N represents the total number of samples in the data set, that is, the number of data points; π k represents the weight of each Gaussian distribution, X i |μ k ,∑ k is the probability density function of the kth Gaussian distribution, μ k and∑ k are the mean vector and covariance matrix respectively, X i Represents the i-th sample data point in the data set; Step S5-1f, alternately perform steps S5-1d and S5-1e for multiple iterations until the mean μ of the model is k , covariance matrix∑ k and weight π k It becomes stable or reaches the maximum number of iterations. Each iteration updates the grid's attribution probability based on the current model and gradually optimizes the cluster division. In step S5-1g, the GMM model outputs the flash flood risk classification result of each grid, that is, the cluster to which each grid belongs.

8. The method for differentiated early warning of flash flood disasters in small and medium-sized river basins according to claim 1 is characterized in that: The step S6-1 further comprises the following sub-steps: Step S6-1a, defining corresponding warning indicators and threshold ranges for areas of different risk levels; Step S6-1b, using a weight analysis method to quantify the impact of different indicators to calibrate the warning threshold; the different indicators include water depth, flow velocity, rainfall, and basin hysteresis; Step S6-1c, verifying the setting of the regional warning threshold with the measured data, and calibrating the warning conditions in real time through retrospective analysis of historical flash flood events.

Images