Flood scenario construction method and system based on multi-model coupling

Abstract

The present application relates to the technical field of flood simulation risk assessment, and provides a flood scenario construction method and system based on multi-model coupling, which aims at the problems of traditional model scale mismatch, process fragmentation, low resolution, efficiency and accuracy contradiction and the like, and constructs a full-space coupling simulation architecture of underground, river channel and ground. A three-dimensional geological structure model is first established to provide a lower boundary for hydrological simulation; then a land surface hydrological model is coupled with meteorological and geological parameters to generate surface runoff; then a river channel confluence model is constructed to output the water level of the section; and finally, the HAND topographic index is combined to realize the rapid conversion of the river water level to the two-dimensional flooded range. The present application penetrates the whole physical process, takes into account the simulation efficiency and the hundred-meter-level resolution, can accurately identify the flooding risk of key facilities, and is suitable for urban storm flood simulation and resilient city planning.

Description

Technical Field

[0001] This invention relates to the field of flood simulation risk assessment technology, and more specifically, to a method and system for constructing flood scenarios based on multi-model coupling. Background Technology

[0002] Traditional urban flood simulation primarily relies on hydrological, hydrodynamic, and one-dimensional or two-dimensional coupled models. Hydrological models generalize based on sub-catchments, calculate runoff using empirical formulas (such as SCS), and employ the Muskingen method for river confluence without two-dimensional inundation. They can only output cross-sectional flow and water level, with a resolution at the kilometer level. One-dimensional pipe network models simulate underground pipe flow, calculating nodal overflows and focusing only on pipe hydraulics, without directly calculating surface inundation. Two-dimensional hydrodynamic models simulate surface runoff based on shallow water equations, offering high accuracy but extremely slow computation (requiring several days for a single scenario), making them unsuitable for multi-scenario processing. The semi-distributed two-dimensional surface formed by the coupling of one- and two-dimensional models only triggers surface inundation during pipe network overflows, lacking sufficient physical mechanisms for direct surface runoff during heavy rainfall.

[0003] The aforementioned traditional urban flood simulation schemes focus primarily on surface and pipeline processes, without considering underground three-dimensional geological structures, and have the following main shortcomings: (1) Model scale mismatch: Geological models, hydrological models and topographic analysis usually operate independently, lack effective coupling, and are difficult to reflect the complete physical process of "rainfall-infiltration-runoff-confluence-inundation".

[0004] (2) Process fragmentation: Most simulations only focus on the two-dimensional evolution of surface floods, ignoring the boundary control effect of underground geological structures on infiltration and groundwater transport, resulting in distorted runoff calculations.

[0005] (3) Insufficient resolution: The spatial resolution of traditional flood risk maps is mostly at the kilometer level (1-10 km), which cannot support the refined risk assessment and resilience planning and design at the block and community scale, especially the ability to identify the inundation of key infrastructure (roads, hospitals, schools, etc.).

[0006] (4) Efficiency versus accuracy: High-precision two-dimensional hydrodynamic models have high computational costs and are difficult to meet the needs of multi-scenario and rapid evaluation.

[0007] Therefore, there is an urgent need for a refined flood scenario construction method that can connect the entire space of "underground-river channel-ground" and take into account both physical mechanisms and computational efficiency. Summary of the Invention

[0008] In view of the problems existing in traditional urban flood simulation schemes, such as lack of underground three-dimensional hydrogeological constraints, model scale fragmentation, and insufficient coupling, the purpose of this invention is to provide a flood scenario construction method and system based on multi-model coupling.

[0009] On the one hand, the present invention provides a flood scenario construction method based on multi-model coupling, comprising the following steps: S1: Establish a three-dimensional geological structure model of the underground area and perform geological volume element processing to obtain the hydrogeological parameters of each volume element, and then rasterize the volume elements into a grid that matches the pre-constructed land surface hydrological model; wherein, the hydrogeological parameters include at least lithology, soil layer thickness, saturated permeability coefficient and porosity; S2: Based on the land surface hydrological model, coupled with the pre-acquired meteorological data and the hydrogeological parameters, simulate the rainfall-infiltration-runoff process and output hourly rasterized surface runoff; S3: The surface runoff is allocated to the pre-divided river segments, and the water level of each river section is obtained by calculating the evolution of each river segment through a preset river confluence model; wherein, the river segments are divided based on the vector river network extracted from the DEM of the study area; S4: The water level of the river section is interpolated into a grid water level field. The grid water level is compared with the HAND topographic index grid by grid to obtain the two-dimensional flood inundation range and water depth distribution. The HAND topographic index is calculated based on the DEM of the study area. When the HAND topographic index is less than or equal to the grid water level, it is determined to be flooded. The flood depth is equal to the grid water level minus the HAND topographic index. S5: Repeat steps S2 to S4 for different return periods or rainfall scenarios to obtain multi-scenario refined flood inundation maps.

[0010] Meanwhile, this invention also provides a flood scenario construction system based on multi-model coupling, applied to electronic devices, including: The underground three-dimensional geological structure modeling unit is used to establish an underground three-dimensional geological structure model of the study area and perform geological volume element processing to obtain the hydrogeological parameters of each volume element, and to rasterize the volume elements into a grid that matches the pre-constructed land surface hydrological model; wherein, the hydrogeological parameters include at least lithology, soil layer thickness, saturated permeability coefficient and porosity; The land surface hydrological model coupling unit is used to couple pre-acquired meteorological data and hydrogeological parameters based on the land surface hydrological model to simulate the rainfall-infiltration-runoff process and output hourly rasterized surface runoff. The river confluence model evolution unit is used to allocate the surface runoff to pre-divided river segments. It calculates the evolution of each river segment through a preset river confluence model to obtain the water level of each river section. The river segments are divided based on the vector river network extracted from the DEM of the study area. The interpolation comparison unit is used to interpolate the water level of the river section into a grid water level field, and compare the grid water level with the HAND topographic index grid by grid to obtain the two-dimensional flood inundation range and water depth distribution; wherein, the HAND topographic index is calculated based on the DEM of the study area, and when the HAND topographic index ≤ the grid water level, it is determined to be flooded, and the flood water depth = grid water level - HAND topographic index; The return period cycle unit is used to repeat steps S2 to S4 for different return periods or rainfall scenarios to obtain a multi-scenario refined flood inundation map.

[0011] The flood scenario construction method and system based on multi-model coupling according to the present invention uses a three-dimensional underground geological structure model as the physical basis for hydrological process simulation. It embeds core hydrogeological parameters such as saturated permeability coefficient and porosity corresponding to soil layers, rock masses, and aquifers into the land surface hydrological model. This effectively solves the industry pain points of traditional flood simulation models, which neglect underground seepage boundary constraints and overgeneralize geological parameters, leading to distorted runoff calculation results and poor simulation fit. The present invention relies on high-precision DEM data to calculate the HAND topographic index and constructs a rapid conversion algorithm for river cross-section water level to surface inundation range and inundation depth. This replaces the computationally cumbersome and computationally expensive traditional two-dimensional hydrodynamic model. With hourly computational efficiency and hundreds-meter spatial resolution, it can accurately identify the flood risk of key infrastructure such as streets, roads, medical facilities, and educational institutions, perfectly achieving a two-way balance between flood simulation accuracy and computational efficiency. Meanwhile, this invention adopts a pluggable model architecture of VIC / RAPID / HAND, which can adapt to application scenarios of different research areas and simulation scales, and has strong compatibility and scalability. It can support the rapid simulation of multiple flood scenarios such as current rainstorms, dynamic rainfall, and flood control and regulation, and can output refined and quantifiable flood risk results, providing efficient and accurate technical tools to support resilient urban planning, regional flood control and disaster reduction, and emergency decision-making and deployment.

[0012] To achieve the foregoing and related objectives, one or more aspects of the present invention include specific features of the present invention, which will be described in detail below. The following description and accompanying drawings illustrate certain exemplary aspects of the invention. However, these aspects indicate only a few of the various ways in which the principles of the invention can be used. Furthermore, the invention is intended to encompass all such aspects and their equivalents. Attached Figure Description

[0013] Other objects and results of the invention will become more apparent and readily understood with reference to the following description taken in conjunction with the accompanying drawings. In the drawings: Figure 1 This is a flowchart illustrating the flood scenario construction method based on multi-model coupling according to an embodiment of the present invention. Figure 2 This is a schematic diagram illustrating the technical chain and data flow of the flood scenario construction method based on multi-model coupling according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the framework structure of a flood scenario construction system based on multi-model coupling according to an embodiment of the present invention.

[0014] In all the accompanying drawings, the same reference numerals indicate similar or corresponding features or functions. Detailed Implementation

[0015] In the following description, numerous specific details are set forth for illustrative purposes and to provide a thorough understanding of one or more embodiments. However, it will be apparent that these embodiments may also be implemented without these specific details. In other instances, well-known structures and devices are shown in block diagram form for ease of description of one or more embodiments.

[0016] To better illustrate the technical solution of the present invention, some of the technical terms involved in the present invention will be briefly explained below.

[0017] (1) DEM: Digital Elevation Model is a digital simulation of the ground topography using limited topographic elevation data. It records the surface elevation information of the study area in the form of a regular grid matrix and is the basic topographic data for hydrological simulation analysis such as water system extraction, topographic analysis, and HAND topographic index calculation.

[0018] (2) HAND Topographic Index: Normalized Elevation Topographic Index is a topographic feature parameter calculated based on DEM data. It represents the vertical elevation difference between each grid in the study area and the nearest river grid. It can effectively eliminate the influence of the overall topographic slope and accurately reflect the local water catchment and water accumulation potential. It is a core topographic parameter for quickly determining the flood inundation range and calculating the inundation depth.

[0019] (3) Kriging interpolation is a high-precision spatial interpolation algorithm based on spatial statistics. By analyzing the spatial correlation of sample data, it makes the optimal unbiased estimation of the attribute values ​​of unknown spatial locations. This invention is used for spatial interpolation of underground strata and hydrogeological parameters to realize the refined and continuous assignment of three-dimensional geological body parameters.

[0020] (4) DSI algorithm: Discrete smooth interpolation algorithm is a high-precision interpolation algorithm suitable for three-dimensional geological modeling. It can fit the undulation characteristics of the stratum interface, smoothly fit complex geological surfaces, accurately depict the stratum layering and spatial distribution of aquifers, and adapt to the construction of complex underground three-dimensional geological structures and the assignment of volume element attributes.

[0021] (5) VIC model: Variable infiltration capacity model is a distributed land surface hydrological model that can simulate regional rainfall infiltration, soil water storage, surface runoff and other hydrological processes in a refined manner based on meteorological data and underlying hydrogeological parameters. It is suitable for multi-scale regional surface runoff fine simulation.

[0022] (6) Noah-MP model: Noah multi-parameter land surface model is an optimized and upgraded land surface hydrological simulation model. It has a complete soil water, land-air exchange and hydrological response simulation mechanism, which can accurately depict the rainfall generation and infiltration process under different lithology and soil layer conditions, and is suitable for hydrological simulation and deduction of complex underlying surface areas.

[0023] (7) RAPID model: A large-scale rapid river confluence model, which is an efficient river confluence calculation model for large-scale watersheds. Based on the topological relationship of river segments, it can complete the flood evolution calculation of multiple river segments in parallel, and quickly output the water level and flow of each river section. It is suitable for batch and multi-scenario flood confluence simulation.

[0024] (8) Musking Matrix Evolution Algorithm: Based on the traditional Musking Matrix confluence principle, the matrix evolution algorithm is optimized. By constructing the confluence matrix of the river segment, it solves the flood evolution process of the entire river segment in a unified manner. It supports parallel computing and can greatly improve the efficiency of river confluence calculation and the overall simulation accuracy.

[0025] (9) D8 algorithm, a classic raster flow direction calculation algorithm, defines the flow direction of a single raster as the steepest descending direction among the eight surrounding neighboring rasteres. It is the basic algorithm for DEM hydrological analysis and is used to calculate raster flow direction and cumulative runoff, providing core support for raster river network extraction and HAND topographic index calculation.

[0026] To address the problems of existing urban flood simulation schemes, such as the lack of underground three-dimensional hydrogeological constraints, fragmented model scale, and insufficient coupling, this invention proposes a flood scenario construction method and system based on multi-model coupling. By simulating the entire space from "underground-river channel-ground surface", the underground geological structure model is used as the physical basis for the hydrological process. In the later stage, the HAND topographic index is used to achieve efficient conversion of river water level to surface inundation, forming a closed-loop coupling technology system.

[0027] In general, the process of connecting "underground-river channel-surface" in this invention mainly includes the following two parts: First, the underground physical foundation (underground space) is constructed, mainly comprising two parts: geological volumetric modeling and underground space coupling. Geological volumetric modeling involves assigning different physical property parameters to each stratum (artificial fill, soft soil, silty clay, bedrock, etc.) using volumetric modeling technology, such as permeability coefficient k, void ratio e, and shear strength parameter c. The hydrogeological parameters involved in this invention include at least lithology, soil layer thickness, saturated permeability coefficient, and porosity. Underground space coupling involves embedding engineering models such as underground pipe networks and subways into the volumetric model to establish spatial topological relationships and identify seepage risk points.

[0028] Next, the physical transfer of hydrological processes from the river channel to the land surface is investigated, mainly including two parts: HAND topographic index calculation and water level-inundation conversion. In the HAND topographic index calculation, based on 30-meter or higher resolution topographic data, the relative elevation of each grid within the watershed relative to the nearest drainage point is calculated, generating a HAND index layer. In the water level-inundation conversion, the river channel water level rise value is directly compared with the HAND layer using spatial algebraic operations to quickly determine the surface inundation boundary.

[0029] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0030] Figure 1 and Figure 2 The flowchart and complete technical chain of the flood scenario construction method based on multi-model coupling according to embodiments of the present invention are shown respectively.

[0031] like Figure 1 and Figure 2 As shown in the figure, the flood scenario construction method based on multi-model coupling provided in this embodiment mainly includes the following steps: S1: Establish a three-dimensional geological structure model of the underground area and perform geological volumetric processing to obtain the hydrogeological parameters of each volumetric element, and then rasterize the volumetric elements into a grid that matches the pre-constructed land surface hydrological model.

[0032] This step is mainly used to establish a three-dimensional underground geological structure model of the study area, and to obtain the spatial distribution and hydrogeological parameters of the soil layer, rock layer and aquifer, which serve as the lower boundary conditions for hydrological simulation.

[0033] The process of establishing a three-dimensional underground geological structure model of the study area mainly includes the following three stages: (1) Multi-source data integration and preprocessing Borehole data, geological maps, and hydrogeological test data for the study area were collected. Spatial distribution models of soil layers, weathered layers, bedrock layers, and aquifers were constructed using 3D geological modeling software (such as GMS and Petrel). Parameters such as lithology, thickness, saturated permeability coefficient, and porosity of each grid cell were output and rasterized into a format matching the hydrological model.

[0034] The three-dimensional geological structure model of the present invention may include artificial fill, soft soil, silty clay, bedrock and aquifer. Kriging interpolation or DSI algorithm is used to assign volume element attributes, and pre-constructed underground pipe network and subway engineering model are embedded to establish topological association.

[0035] Specifically, as an example, 1:5000 high-precision topographic data, borehole columnar sections, geological maps, and real-time groundwater level monitoring data can be collected within the demonstration area. The core of integrating and preprocessing multi-source data lies in the unification of spatial, temporal, and semantic dimensions. First, all data are aligned to the same spatial benchmark (such as CGCS2000) through coordinate transformation, and the borehole elevation is corrected using a 1:5000 DEM as the geometric top cover. Second, the boreholes and geological maps are digitized and structured, extracting stratigraphic positions and attributes, and using spatial interpolation algorithms (such as Kriging or DSI) to transform discrete point and line data into three-dimensional voxel models. Finally, the real-time water level data is denoised and standardized, and mapped to the voxel mesh according to timestamps. Through semantic alignment and unified lithological coding, a seamless digital physical foundation is ultimately formed, integrating surface topography, subsurface structure, and dynamic flow field.

[0036] (2) Geometric framework construction

[0037] Using geological modeling software, surface interpolation is performed based on borehole stratigraphic information to construct contact surfaces between strata (such as bedrock surface undulations and bottom boundaries of loose deposits). Specifically, as an example, the process of 3D geological modeling (GOCAD, Petrel, etc.) consists of three steps: First, data extraction is performed, extracting the X, Y, and Z coordinate points of the top and bottom plates of each lithological layer from a structured borehole database; then, surface interpolation is performed, using the discrete smooth interpolation radial basis function (RBF) algorithm to fit a continuous stratigraphic trend surface between discrete borehole points; finally, topological trimming and constraints are performed, using a high-precision DEM as the top boundary, and Boolean operations are performed according to the stratigraphic depositional sequence (old strata being covered or cut by new strata) to handle pinch-outs, unconformities, and fault cutting relationships between strata, thereby generating a geometric framework that can reflect bedrock undulations or loose layer boundaries.

[0038] (3) Geological body elementalization

[0039] The geological volume transformation described in this invention refers to cutting three-dimensional space into regular or irregular voxel grids, so that the geological model is transformed from a simple "empty shell" surface into a solid body with internal properties.

[0040] Specifically, as an example, the geological body elementalization process includes: The underground three-dimensional space of the study area is divided into a volumetric mesh; Map the various stratigraphic interfaces of the study area to the individual voxels of the voxel grid; The hydrogeological parameters are assigned to each volume element according to the lithological type of the study area.

[0041] Solidification in three-dimensional space is typically achieved using spatial partitioning techniques. The partitioning process begins with Boolean operations, dividing the space into different geological units based on the previously constructed strata contact surfaces (curved surfaces). Then, each unit is filled with a mesh, commonly using regular voxel partitioning (such as uniform-sized Voxel pixel arrays) or irregular voxel partitioning (such as tetrahedral or prismatic meshes adapted to complex strata boundaries). The partitioning criterion is the consistency of spatial topology and properties, ensuring that each voxel contains only the same lithology or physical parameter, and that the top surface of the mesh strictly conforms to the high-precision DEM, thus transforming the geometry of the geological structure into computational units capable of bearing physical properties such as mechanical properties and permeability coefficients.

[0042] This invention transforms a three-dimensional geological structure model into a three-dimensional geological volumetric model, which is a digital representation of the underground geological structure model. Without a structural model, the volumetric model lacks shape constraints; without a volumetric model, the structural model is merely a few thin films, unable to support physical property calculations. This integration of above-ground and underground elements allows flood simulation to go beyond the surface, further considering the influence of groundwater depth and soil structure on seepage.

[0043] (4) Attribute assignment and spatial coupling

[0044] The physical parameters obtained from hydrogeological tests are mapped to each volume element mesh, and the above-ground and underground penetration function is enabled to embed the underground space engineering (subway, pipeline network) model into it and establish topological association.

[0045] (5) Rasterization output: Export the processed three-dimensional voxel model as a standard raster format to ensure that its resolution matches the surface hydrological model, thereby supporting dynamic calculations from rainfall infiltration to geological response.

[0046] The generation of the 3D voxel model is the final integration and physical instantiation of all previous processing steps. Its logical progression is as follows: First, the high-precision DEM from the multi-source integration stage is used as the top boundary to ensure accurate surface morphology. Second, using the interlayer contact surfaces generated by interpolation in the geometric framework stage as spatial constraints, the continuous geological space is divided into subdivided grid units through voxel cutting. Subsequently, preprocessed borehole attributes and real-time water level data are mapped to the corresponding grids, completing the transformation from a geometric shell to a physical entity. The final rasterization output is a dimensionality reduction mapping process that perfectly aligns the 3D voxels with the grid of the surface hydrological model on the horizontal projection, allowing every millimeter of rainfall infiltration on the surface to directly enter the corresponding voxel unit for geological response calculation in the vertical dimension, thus achieving numerical coupling between the above-ground and underground.

[0047] The three-dimensional voxel model of this invention adopts a hierarchical architecture of "bottom-level data - physical voxels - business coupling" to ensure that it can serve as the physical basis for hydrological processes. The following shows an example of the data correspondence between the base layer of the multi-source heterogeneous data layer, the three-dimensional geological model layer, and the visualization support layer according to an embodiment of this invention:

[0048] After establishing a three-dimensional geological structure model of the underground area and performing geological volume element processing, the land surface hydrological simulation stage was entered.

[0049] S2: Based on the land surface hydrological model, coupled with the pre-acquired meteorological data and the hydrogeological parameters, simulate the rainfall-infiltration-runoff process and output hourly rasterized surface runoff.

[0050] In one specific embodiment of this invention, the land surface hydrological model employs either the VIC or Noah-MP model, with a calculation time step of 1 hour. Specifically, as an example, the VIC model is used, inputting historical or designed heavy rainfall meteorological data for the study area (time step 1 hour), and coupling the geological parameters (such as permeability coefficient) output in step 1 as soil stratification parameters. The model simulates evapotranspiration, surface runoff, baseflow, etc., for each grid, outputting hourly rasterized surface runoff.

[0051] More specifically, the VIC model simulation process can include four stages: data preprocessing, parameter coupling, model solving, and result output.

[0052] In the data preprocessing stage, the first step is to prepare the gridded base: 1) The study area was divided into regular computational grids based on the high-precision DEM (30-100m); 2) Define each grid as an independent computational unit containing multiple soil layers (usually 3 layers).

[0053] Then, meteorological forcing data is input: 1) Input hourly data such as rainfall, temperature, wind speed, and humidity; 2) The land surface hydrological model spatially allocates rainfall according to a preset time step (e.g., 1 hour) and initiates water balance calculation.

[0054] In the parameter coupling stage, geological parameters are coupled. Specifically, the saturated permeability coefficient (k), porosity, and soil thickness obtained in step S1 are mapped to the soil parameter file of the VIC model. This coupling ensures that infiltration calculations are no longer empirical values, but rather physical responses based on measured geological body properties. In this invention, soil thickness refers to the thickness of the loose deposit layer or artificial fill layer in the corresponding geological model, which is an important factor affecting soil water storage capacity and slope instability criteria.

[0055] During the model solution phase, a moisture balance solution is performed: 1) The model calculates evapotranspiration, infiltration, surface runoff, and baseflow in parallel on each grid. 2) Rainfall first fills the pores in the soil surface, and the portion exceeding the infiltration capacity forms surface runoff.

[0056] In the results output stage, the runoff results are output in a rasterized form: the hourly rasterized surface runoff is output as input data for subsequent river confluence simulation.

[0057] S3: Distribute the surface runoff to the pre-divided river segments, and calculate the water level of each river segment by segment through a preset river confluence model.

[0058] The river sections in this invention are divided based on the vector river network extracted from the DEM of the study area. Specifically, as an example, the vector river network of the study area is extracted based on a 30m resolution DEM to establish the river channel topology. The surface runoff output in step S2 is used as the lateral inflow and input into the RAPID model to calculate the river flood evolution, obtaining the flow process and water level of each river section.

[0059] This step is used to simulate river confluence. A specific simulation process example is as follows: S31: Vector River Network Topology Construction: Utilize a 30m high-precision DEM to extract flow direction and cumulative flow, generate a vectorized river network, and establish a topology connection table including river segment ID and upstream / downstream logical relationships, as shown in the example below.

[0060]

[0061] S32: Runoff Spatial Weight Mapping: The rasterized surface runoff output in step S2 is used as lateral inflow. Through spatial overlay analysis, the runoff of each grid is weighted and accumulated to the corresponding vector river segment.

[0062] Spatial overlay refers to overlaying the generated vector river network sub-basin layer with the runoff raster layer. If a grid belongs entirely to the catchment area of ​​a certain river segment, its weight is 1.0; if a grid crosses a boundary, it is assigned according to its area proportion.

[0063] Specifically, as an example, suppose at a certain moment, there are four 30m grid cells within the sub-basin Sub_A, and the calculated surface runoff depth is as follows:

[0064] S33: RAPID Matrix Evolution Calculation: The Muskingum matrix evolution algorithm based on matrix form is used to solve the flow of each river segment in parallel under the whole basin topology.

[0065] The following is an example of the core matrix evolution formula: At time step Δ t Within the entire basin, the flow evolution of all river sections can be represented by the following matrix equation: Q out ( t +Δ t )=A·Q out ( t )+B·Q in ( t +Δ t )+C·Q in ( t ) Among them, Q out Let be the vector of outlet discharges for all river segments in the entire basin that needs to be solved. t Q represents the current computation time (or the current starting point in the time domain). out ( t +Δ t Let Q be the vector of outlet discharges for all river segments in the entire basin at the next time step. out ( t Q represents the current moment's outflow vector for all river segments in the entire basin (based on known historical data). in ( t +Δ t Q represents the inflow vector of all river segments in the entire basin at the next moment (known future inflows). in ( t Let A, B, and C be the inflow vectors of all river segments in the entire basin at the current moment (known historical inflows), and let A, B, and C be the Muskingan parameters (including propagation time and flow weights) and time step Δ for each river segment in the entire basin.t The calculated coefficient matrix.

[0066] To achieve full basin coupling, the RAPID model integrates upstream inflow relationships into a matrix, ultimately resulting in the following simplified form:

[0067] M1·Q out ( t +Δ t )=M2·Q out ( t )+V in

[0068] Where M1 is the implicit operator matrix of the system at the next time step, and M2 is the explicit operator matrix at the current time step. Both M1 and M2 are constructed by Muskingan parameters, time step size, and river network topology. V in The integrated inflow vector includes the inter-regional inflow and upstream inflow contribution for each river segment.

[0069] S34: Water Level-Flow Process Output: Real-time output of hourly flow rate Q at each river cross-section, and conversion of the flow rate Q into dynamic water level using the Manning formula or water level-flow curve, taking into account the river cross-sectional geometry (such as trapezoidal or rectangular cross-section). H This process supports subsequent flooding analysis.

[0070] Taking one hour during a rainfall event as an example, the conversion process is as follows: Input: The RAPID model outputs the flow rate Q = 150 m³ / h of this river segment at 14:00. 3 / s; Match: Call the geometric parameters of the section (e.g., bottom width 20m, roughness 0.03).

[0071] Calculation: If the water level H =2.0m, Q is calculated calc =142m 3 / s (too small); If the water level H =2.1m, Q is calculated calc =155m 3 / s (too large); Output: The dynamic water level at 14:00 is determined through high-frequency iteration. H =2.06m.

[0072] S4: Interpolate the water level of the river section into a grid water level field, compare the grid water level with the HAND topographic index grid by grid to obtain the two-dimensional flood inundation range and water depth distribution.

[0073] The HAND topographic index is calculated based on the DEM of the study area. When the HAND topographic index is less than or equal to the grid water level, it is determined to be flooded. The flood depth is calculated as grid water level - HAND topographic index.

[0074] In calculating the HAND topographic index based on the DEM of the study area, the DEM is filled with depressions, the river network is extracted, the elevation difference between each raster and the nearest river channel raster is calculated, and a HAND raster layer is generated. The specific steps include the following: S41: The DEM of the study area is subjected to depression filling treatment to eliminate local depressions and obtain a depression-free DEM; S42: Based on the depression-free DEM, the flow direction of each grid is calculated using the D8 algorithm; S43: Calculate the cumulative flow of each grid cell based on the flow direction, and extract the grid river network according to the cumulative flow and the preset cumulative flow threshold. S44: Based on the grid river network, determine the nearest river grid corresponding to each grid in the study area, and calculate the elevation difference between each grid and the corresponding nearest river grid as the HAND topographic index.

[0075] In one specific embodiment of the present invention, the method for calculating the HAND terrain index includes: Step 1: DEM Preprocessing. To eliminate isolated depressions in the DEM caused by data errors or micro-topography and to ensure the continuity of water flow on the terrain, the original 30m / 90m DEM was globally filled with depressions using the algorithm of Wang & Liu et al.

[0076] Step 2: Calculate the flow direction. The D8 algorithm is used to determine the direction with the steepest slope among eight adjacent pixels. For example, for a pixel... i Its direction to adjacent pixels j The slope is:

[0077] in, Z For elevation, Z i Represents a pixel i elevation, Z j Represents a pixel j elevation, d i,j Represents a pixel i and pixels j The distance between the center points, S i,j Represents a pixel j Relative to a pixel i The slope. The flow direction is in the direction of the maximum slope value S.

[0078] Step 3: Extract the river network. The total number of pixels A contributed by the upstream of each grid is calculated as the cumulative flow. Then, a threshold is set. When A is greater than the threshold, the grid is defined as a river grid.

[0079] Step 4: Determine the elevation of the nearest drainage point. Using the flow direction map, trace each non-channel grid downstream until you reach the first channel grid in that flow path, and then record the elevation value of that channel grid.

[0080] Step 5: Generate the HAND grid.

[0081]

[0082] in, Z ( x , y ) indicates the current grid ( x , y The original elevation value of ) Z stream ( x , y The value represents the elevation of the river channel grid at the end of the flow path corresponding to this grid.

[0083] Then, a typical flood scenario (such as a 100-year design rainstorm) is selected, and the river level at the corresponding moment in step S3 is extracted. For each grid cell in the study area, the river level is compared with the HAND value of that grid cell: if the river level > the HAND value, then the grid cell is submerged, and the submerged depth = river level - HAND value; otherwise, it is not submerged. Finally, a two-dimensional submerged range and depth distribution map (i.e., a flood submerged map) is generated, and the spatial resolution of the generated flood submerged map is consistent with the resolution of the DEM of the study area (e.g., 30 m).

[0084] Specifically, as an example, the submersion conversion process is as follows: 1. Scene and water level extraction (input preparation): Extract the river cross-section water level H at a specific time t (such as the flood peak) or a specific return period (such as a 100-year return period) from the RAPID model in step S3. Here, the water level is the relative incremental height relative to the bottom of the riverbed, rather than the absolute elevation.

[0085] 2. Spatial scale matching: The river water level data is spatially interpolated from discrete "cross-section points" along the river network direction to generate a continuous raster layer of river water level with the same resolution as the HAND raster (30m).

[0086] 3. Grid-by-grid logical discrimination (core calculation): Traverse each grid cell (x, y) within the study area and read the HAND(x, y) value of that grid cell. If the river level at that point can "overflow" its relative height difference, it is considered flooded.

[0087] 4. Water depth and range generation: The flooding depth is calculated based on the discrimination results, and the unflooded areas are set to invalid values ​​or transparent; finally, a two-dimensional raster layer (TIFF format) consistent with the original DEM spatial reference is generated.

[0088] S5: Repeat steps S2 to S4 for different return periods or rainfall scenarios to obtain multi-scenario refined flood inundation maps.

[0089] This step is used for scenario construction and visualization. Steps S2-S4 are repeated for different return periods (e.g., 20-year, 50-year, 100-year return periods) or different rainfall durations to generate multi-scenario inundation maps. The results are imported into a 3D visualization platform (e.g., Cesium, Unity), and combined with vector data of buildings, roads, etc., to achieve a 3D dynamic display and risk analysis of the inundation process.

[0090] The core of scenario building lies in transforming discrete rainfall data into a continuous logic of risk evolution. As an example, the scenario building process is as follows: 1. Set trigger conditions: Based on long-term historical meteorological data, set different levels of rainfall drivers (e.g., 24-hour rainfall ≥ 200mm corresponds to a 50-year return period scenario). 2. Parameter cyclic correction: Repeatedly execute VIC (land surface), RAPID (river channel), and HAND (inundation) calculations; Specifically, for a 20-year return period, the focus is on water accumulation in normal flood control areas and low-lying road sections; for a 100-year return period, the focus is on the risk of floods exceeding the standard for dikes and the overflow risk of important infrastructure (such as pumping stations and subway entrances).

[0091] 3. Risk correlation calculation: The generated raster flood map is overlaid with the vector base (buildings, population distribution, road network) to calculate quantitative indicators such as the number of threatened people and the estimated damage to infrastructure.

[0092] The above describes a flood scenario construction method based on multi-model coupling according to an embodiment of the present invention. This method can be implemented in software, hardware, or a combination of both. Through this multi-model coupling-based flood scenario construction method, fully coupled simulation of the underground-channel-surface area can be achieved, ensuring the accuracy of hydrophysical processes and rapidly completing two-dimensional inundation conversion through the HAND topographic index, greatly improving the efficiency and resolution of refined flood scenario construction.

[0093] Corresponding to the above-mentioned flood scenario construction method based on multi-model coupling, the present invention also provides a flood scenario construction system based on multi-model coupling.

[0094] Figure 3 The framework structure of a flood scenario construction system 300 based on multi-model coupling according to an embodiment of the present invention is shown. Figure 3 As shown, the flood scenario construction system based on multi-model coupling provided by this invention generates response information for public communication events using the flood scenario construction method based on multi-model coupling described above. According to the functions implemented, the flood scenario construction system 300 based on multi-model coupling mainly includes an underground three-dimensional geological structure modeling unit 310, a land surface hydrological model coupling unit 320, a river confluence model evolution unit 330, an interpolation comparison unit 340, and a return period cycle unit 350.

[0095] The unit of this invention can also be called a module, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and are stored in the memory of the electronic device.

[0096] Furthermore, in one possible implementation of the present invention, the functions of each module / unit are as follows: The underground three-dimensional geological structure modeling unit 310 is used to establish an underground three-dimensional geological structure model of the study area and perform geological volume element processing to obtain the hydrogeological parameters of each volume element, and to rasterize the volume elements into a grid that matches the pre-constructed land surface hydrological model; wherein, the hydrogeological parameters include at least lithology, soil layer thickness, saturated permeability coefficient and porosity. The land surface hydrological model coupling unit 320 is used to couple the pre-acquired meteorological data and the hydrogeological parameters based on the land surface hydrological model to simulate the rainfall-infiltration-runoff process and output hourly rasterized surface runoff. The river confluence model evolution unit 330 is used to allocate the surface runoff to pre-divided river segments and calculate the water level of each river section by evolving segment by segment through a preset river confluence model; wherein, the river segments are divided based on the vector river network extracted from the DEM of the study area. Interpolation comparison unit 340 is used to interpolate the water level of the river section into a grid water level field, and compare the grid water level with the HAND topographic index grid by grid to obtain the two-dimensional flood inundation range and water depth distribution; wherein, the HAND topographic index is calculated based on the DEM of the study area, and when the HAND topographic index ≤ the grid water level, it is determined to be flooded, and the flood water depth = grid water level - HAND topographic index; The return period cycle unit 350 is used to repeat steps S2 to S4 for different return periods or rainfall scenarios to obtain a multi-scenario refined flood inundation map.

[0097] The flood scenario construction system based on multi-model coupling described above is an implementation method corresponding to the flood scenario construction method based on multi-model coupling described above. Its specific implementation scheme can be referred to the specific embodiments of the flood scenario construction method based on multi-model coupling described above, and will not be described in detail here.

[0098] This invention provides a flood scenario construction method and system based on multi-model coupling. For the first time, it uses an underground geological structure model as the physical basis for hydrological processes and achieves efficient conversion of river water levels to surface inundation through the HAND topographic index, forming a closed-loop coupling technology system. It utilizes a high-precision DEM and physical mechanism model to improve the spatial resolution of flood inundation range to the hundred-meter level, clearly identifying the flood risk of roads and critical facilities, and supporting refined design of health and safety units. The introduction of the HAND topographic index as a rapid topographic analysis tool effectively avoids the high computational cost of traditional two-dimensional hydrodynamic models, shortening the inundation scenario construction time from several days to several hours, while ensuring the physical rigor of the hydrological process. It can also flexibly integrate different hydrological models and confluence models, supporting rapid evaluation of various rainstorm scenarios and planning schemes, providing a quantitative tool for rainstorm disaster prevention and mitigation and resilient spatial planning in megacities.

[0099] In the several embodiments provided by this invention, it should be understood that the disclosed devices, systems, and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0100] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0101] Furthermore, the functional modules in the various embodiments of this invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in a combination of hardware and software functional modules. Moreover, the word "comprising" obviously does not exclude other units or steps, and the singular does not exclude the plural. The multiple units or systems described in the flood scenario construction system based on multi-model coupling can also be implemented by a single unit or system through software or hardware.

[0102] The flood scenario construction method and system based on multi-model coupling according to the present invention have been described above by way of example with reference to the accompanying drawings. However, those skilled in the art should understand that various modifications can be made to the flood scenario construction method and system based on multi-model coupling proposed in the present invention without departing from the scope of the present invention.

Claims

1. A flood scenario construction method based on multi-model coupling, characterized in that, Includes the following steps: S1: Establish a three-dimensional geological structure model of the underground area and perform geological volume element processing to obtain the hydrogeological parameters of each volume element, and then rasterize the volume elements into a grid that matches the pre-constructed land surface hydrological model; wherein, the hydrogeological parameters include at least lithology, soil layer thickness, saturated permeability coefficient and porosity; S2: Based on the land surface hydrological model, coupled with the pre-acquired meteorological data and the hydrogeological parameters, simulate the rainfall-infiltration-runoff process and output hourly rasterized surface runoff; S3: The surface runoff is allocated to the pre-divided river segments, and the water level of each river section is obtained by calculating the evolution of each river segment through a preset river confluence model; wherein, the river segments are divided based on the vector river network extracted from the DEM of the study area; S4: The water level of the river section is interpolated into a grid water level field. The grid water level is compared with the HAND topographic index grid by grid to obtain the two-dimensional flood inundation range and water depth distribution. The HAND topographic index is calculated based on the DEM of the study area. When the HAND topographic index is less than or equal to the grid water level, it is determined to be flooded. The flood depth is equal to the grid water level minus the HAND topographic index. S5: Repeat steps S2 to S4 for different return periods or rainfall scenarios to obtain multi-scenario refined flood inundation maps.

2. The flood scenario construction method based on multi-model coupling as described in claim 1, characterized in that, The three-dimensional geological structure model includes artificial fill, soft soil, silty clay, bedrock and aquifer. Kriging interpolation or DSI algorithm is used to assign volume element attributes, and pre-constructed underground pipe network and subway engineering model are embedded to establish topological associations.

3. The flood scenario construction method based on multi-model coupling as described in claim 2, characterized in that, The geological body elementalization process includes: The underground three-dimensional space of the study area is divided into a volumetric mesh; Map each stratigraphic interface of the study area to each volume element of the volume element mesh; The hydrogeological parameters are assigned to each volume element according to the lithological type of the study area.

4. The flood scenario construction method based on multi-model coupling as described in any one of claims 1 to 3, characterized in that, The land surface hydrological model adopts either the VIC or Noah-MP model, and the calculation time step of the land surface hydrological model is 1 hour.

5. The flood scenario construction method based on multi-model coupling as described in claim 4, characterized in that, The meteorological data includes hourly rainfall data, temperature data, wind speed data, and humidity data.

6. The flood scenario construction method based on multi-model coupling as described in claim 5, characterized in that, The river confluence model adopts the RAPID model, which calculates the flow of each river segment in parallel based on the Musking matrix evolution algorithm, and converts the water level of each river segment into the cross-sectional geometric relationship of each river segment.

7. The flood scenario construction method based on multi-model coupling as described in claim 6, characterized in that, The method for calculating the HAND topographic index based on the DEM of the study area includes: The DEM of the study area was subjected to depression filling treatment to eliminate local depressions and obtain a depression-free DEM; Based on the depression-free DEM, the flow direction of each grid cell is calculated using the D8 algorithm; The cumulative flow of each grid is calculated based on the flow direction, and the grid river network is extracted according to the cumulative flow and the preset cumulative flow threshold. Based on the grid river network, the nearest river grid corresponding to each grid in the study area is determined, and the elevation difference between each grid and the corresponding nearest river grid is calculated as the HAND topographic index.

8. The flood scenario construction method based on multi-model coupling as described in claim 7, characterized in that, The resolution of the DEM in the study area is 30–100 m, and the spatial resolution of the generated flood inundation map is consistent with that of the DEM in the study area.

9. The flood scenario construction method based on multi-model coupling as described in claim 1, characterized in that, The different return periods include design rainstorm scenarios with a return period of 20 years, 50 years, and 100 years.

10. A flood scenario construction system based on multi-model coupling, applied to electronic devices, characterized in that, include: The underground three-dimensional geological structure modeling unit is used to establish an underground three-dimensional geological structure model of the study area and perform geological volume element processing to obtain the hydrogeological parameters of each volume element, and to rasterize the volume elements into a grid that matches the pre-constructed land surface hydrological model; wherein, the hydrogeological parameters include at least lithology, soil layer thickness, saturated permeability coefficient and porosity; The land surface hydrological model coupling unit is used to couple pre-acquired meteorological data and hydrogeological parameters based on the land surface hydrological model to simulate the rainfall-infiltration-runoff process and output hourly rasterized surface runoff. The river confluence model evolution unit is used to allocate the surface runoff to pre-divided river segments. It calculates the evolution of each river segment through a preset river confluence model to obtain the water level of each river section. The river segments are divided based on the vector river network extracted from the DEM of the study area. The interpolation comparison unit is used to interpolate the water level of the river section into a grid water level field, and compare the grid water level with the HAND topographic index grid by grid to obtain the two-dimensional flood inundation range and water depth distribution; wherein, the HAND topographic index is calculated based on the DEM of the study area, and when the HAND topographic index ≤ the grid water level, it is determined to be flooded, and the flood water depth = grid water level - HAND topographic index; The return period cycle unit is used to repeat steps S2 to S4 for different return periods or rainfall scenarios to obtain a multi-scenario refined flood inundation map.

Images