A multi-scenario driven watershed flood dynamic simulation method and system

Abstract

The present application relates to the technical field of flood data processing, and particularly relates to a multi-scenario driven basin flood dynamic simulation method and system. The basin flood dynamic simulation method comprises the following steps: S1, fusing meteorological, geographical and hydrological data, constructing a surface grid based on the fused data, and dividing a basin catchment area by using the surface grid; S2, designing a rainstorm scenario set according to a historical rainfall sequence, climate characteristics, a return period and rainfall duration, constructing a response relationship of a runoff coefficient dynamically changing with rainfall time sequence, and constructing a rainstorm-runoff scenario by using the rainstorm scenario set and the response relationship; S3, constructing a flood simulation model and performing flood dynamic simulation on the rainstorm-runoff scenario. In the present application, a multi-dimensional rainstorm scenario set is built according to a historical rainfall sequence, climate characteristics, a return period and rainfall duration in a research area when designing a rainstorm-runoff scenario, which can cover different climate conditions and extreme weather events, and the adaptability to hydrological events under different climate change backgrounds is improved.

Description

Technical Field

[0001] This invention relates to the field of flood data processing, and in particular to a multi-scenario driven method and system for dynamic simulation of watershed floods. Background Technology

[0002] Floods are among the most destructive natural disasters globally, posing a serious threat to people's lives and property, and causing significant damage to social systems and infrastructure. Conducting dynamic simulations of river basin floods is of paramount strategic importance for improving flood control and disaster reduction capabilities and optimizing the design of water conservancy projects.

[0003] Watershed flood simulation is a key step in assessing flood disaster risk. From a research paradigm perspective, existing methods can be mainly categorized into two main types: data-driven and model-driven methods.

[0004] Data-driven approaches primarily rely on historical observation data, remote sensing imagery, and geographic information data, employing methods such as statistical analysis, spatial analysis, or machine learning to analyze flood evolution and assess disasters.

[0005] Historical disaster data analysis relies heavily on the accumulation of historical disaster data, which has problems such as delayed response. While remote sensing and GIS coupled analysis can intuitively reflect the spatial distribution characteristics of flood disasters, it lacks dynamic evolution simulation and is difficult to capture the flood propagation process. Although deep learning methods have improved the accuracy of flood prediction, they are highly dependent on the quality and quantity of data.

[0006] Model-driven methods simulate flood propagation processes based on hydrodynamics, fluid mechanics, or mathematical modeling techniques. Among these, hydrodynamic models can simulate flood formation and propagation, but they are complex to construct, require a large number of parameters, and have certain limitations in handling large-scale scenarios across multiple scenarios.

[0007] In recent years, flood simulation methods based on scenario construction theory have gradually attracted attention. However, the application of these methods at the watershed scale is still in its early stages, lacking the comprehensive ability to represent multiple rainfall scenarios, various surface conditions, and the dynamic evolution of floods. Although the above research has promoted the development of watershed hydrology, in the face of complex and ever-changing watershed environments, existing technologies still have shortcomings such as incomplete scenario simulation, lack of construction of different rainfall-runoff scenarios, and difficulty in adapting to hydrological events under different climate change backgrounds. Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of incomplete scenario simulation in the existing technology and to provide a multi-scenario driven method and system for dynamic simulation of watershed floods.

[0009] In a first aspect, the present invention provides a multi-scenario driven method for dynamic simulation of watershed floods, characterized by comprising the following steps:

[0010] S1. Meteorological, geographical, and hydrological data are fused and processed, and a surface grid is constructed based on the fused data. The surface grid is then used to delineate the catchment area of ​​the watershed.

[0011] S2. Designing a rainstorm-runoff scenario by combining the fused data and the watershed catchment area, including:

[0012] Design a set of rainstorm scenarios based on historical rainfall sequences, climate characteristics, recurrence periods, and rainfall durations;

[0013] Construct the response relationship between runoff coefficient and the dynamic change of rainfall time series;

[0014] A rainstorm-runoff scenario is constructed using the rainstorm scenario set and the response relationship;

[0015] S3. Construct a flood simulation model to perform dynamic flood simulation on the described rainstorm-runoff scenario.

[0016] According to a preferred embodiment, S1 includes: S11, data acquisition and fusion, including:

[0017] Obtain raw meteorological, geographical, and hydrological data within the study area; standardize the raw data to obtain a basic dataset;

[0018] By using spatial overlay analysis, the basic dataset is fused with digital elevation model data to obtain a raster dataset. Preferably, the standardization process includes at least: cropping, georegistration, resampling, abstraction, format conversion, and scene mapping.

[0019] According to a preferred embodiment, S1 further includes: S12, constructing a three-level refined grid using the raster dataset, including:

[0020] The continuous terrain surface is discretized into regular grid cells by rasterization to obtain the first-level macroscopic terrain grid.

[0021] An adaptive grid densification algorithm based on terrain features is used to densify the macroscopic terrain grid to obtain a second-level mesoscale feature grid.

[0022] A grid optimization algorithm based on hydrological processes is used to optimize the mesoscale feature grid by analyzing the confluence path and confluence time, thereby obtaining a third-level hydrological feature grid.

[0023] According to a preferred embodiment, S1 further includes: S13, dividing the watershed catchment area using the three-level refined grid, including:

[0024] Depression filling processing is performed on the DEM data;

[0025] The flow direction analysis of the three-level refined grid was performed using the D8 single-flow direction algorithm.

[0026] Calculate the cumulative flow for each grid cell;

[0027] Based on the flow direction analysis and the calculation results of the cumulative runoff, the catchment point is determined, the flow direction is traced in reverse, and all grid areas flowing toward the catchment point are identified to form a preliminary catchment surface. The topographic curvature and the runoff weight factor are introduced to optimize the boundary morphology of the catchment surface for the first time. The boundary smoothing algorithm and topological relationship correction are used to optimize the boundary morphology of the catchment surface for the second time.

[0028] According to a preferred embodiment, in S2, the step of designing a rainstorm scenario set based on historical rainfall sequences and climate characteristics includes:

[0029] Based on the climate conditions and meteorological data of the study area, rainfall parameters, rainfall variation parameters, rainfall duration correction parameters, and rainstorm attenuation index are obtained, and rainstorm intensity is designed.

[0030] The Chicago rainfall pattern method was used to assign rainfall intensity in the study area;

[0031] A set of rainstorm scenarios was designed based on the recurrence period and rainfall duration.

[0032] According to a preferred embodiment, in S2, constructing the response relationship of runoff coefficient with the dynamic change of rainfall time sequence includes:

[0033] Calculate the runoff curve over time based on soil conditions, soil saturation state, soil permeability, and rainfall duration;

[0034] Calculate the potential maximum retention based on the number of runoff curves;

[0035] Calculate runoff based on the potential maximum retention and rainfall;

[0036] The runoff coefficient is determined by the ratio of instantaneous runoff to current rainfall, and the response relationship of the runoff coefficient to the dynamic change of rainfall time sequence is constructed.

[0037] According to a preferred embodiment, the construction of the flood simulation model in S3 includes: constructing the flood simulation model based on cellular automata. The flood simulation model divides the study area into discrete square grid cells according to a preset resolution, constructing a computational space composed of several cells, and simulating the spatiotemporal process of flood flow in the basin within a discrete space and discrete time framework. The evolution of the cells is determined by the hydraulic state (velocity and water level changes) of the initial inflow boundary (such as the basin outlet section or distributed runoff generation unit); and the flood evolution of adjacent cells is calculated based on the height, roughness, water depth, unit width discharge in the X direction, and unit width discharge in the Y direction of the cell's location.

[0038] According to a preferred embodiment, the flood evolution is calculated using a discretized set of Saint-Venant equations. Non-physical factors during the evolution process are constrained and controlled, with rules including:

[0039] Water conservation: The sum of the remaining rainwater and the total water volume of the cells must not exceed the designed total intrusion water volume to ensure the total water balance in the simulation;

[0040] Traffic constraints: The increase in the bandwidth of a single cell cannot exceed the maximum bandwidth it can provide, in order to comply with physical traffic limits;

[0041] Cell state transition: If there is water around the cell and the number of neighboring cells with water is greater than 7, the cell state changes to water; if there is water around the cell and the number of neighboring cells with water is less than 2, the cell state changes to no water.

[0042] According to a preferred embodiment, the method further includes visualizing the dynamic simulation of the flood, including: obtaining the evolution results of the flood simulation model at each time step; and using the water depth information of the evolution results to construct a triangular network for flood modeling.

[0043] In a second aspect, the present invention also provides a multi-scenario driven watershed flood dynamic simulation system, which is capable of executing the multi-scenario driven watershed flood dynamic simulation method provided by the present invention.

[0044] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0045] This invention provides a multi-scenario driven method and system for dynamic simulation of watershed floods. It obtains a surface grid by fusing multi-source meteorological, geographical, and hydrological data, and delineates the watershed catchment area. Combining the fused data and the watershed catchment area, it designs storm-runoff scenarios, constructs a flood simulation model, and performs dynamic flood simulation on these scenarios. In designing the storm-runoff scenarios, this invention constructs a multi-dimensional storm scenario set based on the historical rainfall sequence and climate characteristics, return period, and rainfall duration of the study area. This set can cover different climatic conditions and extreme weather events, improving the adaptability of the watershed flood dynamic simulation method provided by this invention to hydrological events under different climate change backgrounds. Attached Figure Description

[0046] Figure 1 This is a schematic diagram of a multi-scenario driven dynamic simulation method for watershed floods according to a preferred embodiment of the present invention.

[0047] Figure 2 This is a schematic diagram of the D8 unidirectional flow algorithm.

[0048] Figure 3 This is a schematic diagram of the flow direction encoding for the D8 single-flow algorithm.

[0049] Figure 4 This is a schematic diagram illustrating the construction and simulation of a rainstorm-runoff scenario.

[0050] Figure 5 This is a schematic diagram of a cellular automata model.

[0051] Figure 6 This is a schematic diagram of a flood evolution method based on a cellular automata model.

[0052] Figure 7 A visual diagram of a flood. Detailed Implementation

[0053] The present invention will now be described in further detail with reference to specific embodiments. However, this should not be construed as limiting the scope of the present invention to the following embodiments; all technologies implemented based on the content of the present invention fall within the scope of the present invention.

[0054] Unless otherwise specified, the use of terms such as "upper," "lower," "left," "right," "center," "inner," and "outer" to indicate orientation or positional relationships in the description of specific embodiments of the present invention is based on the orientation or positional relationships shown in the accompanying drawings, or the orientation or positional relationship in which the product / equipment / device is typically placed during use. These terms are merely for the purpose of facilitating the description of the present invention or simplifying the description in specific embodiments, enabling those skilled in the art to quickly understand the solution, and do not indicate or imply that a particular device / component / element must have a specific orientation, or be constructed and operated in a specific positional relationship. Therefore, they should not be construed as limitations on the present invention.

[0055] Furthermore, the use of terms such as "horizontal," "vertical," "suspended," and "parallel" does not imply that the corresponding device / component / element must be absolutely horizontal, vertical, suspended, or parallel, but rather that it can be slightly tilted or have a deviation. For example, "horizontal" merely means that its direction is more horizontal relative to "vertical," not that the structure must be completely horizontal, but that it can be slightly tilted. Alternatively, it can be simplified to mean that the corresponding device / component / element, when set in a "horizontal," "vertical," "suspended," or "parallel" direction, can have an error / deviation of ±10% relative to the corresponding direction, more preferably within ±8%, more preferably within ±6%, more preferably within ±5%, and more preferably within ±4%. As long as the corresponding device / component / element is within the error / deviation range, it can still achieve its function in the present invention.

[0056] Furthermore, the use of terms such as "first," "second," and "third" in terminology is merely for distinguishing descriptions of identical or similar components and should not be interpreted as emphasizing or implying the relative importance of a particular component.

[0057] Furthermore, in the description of the embodiments of the present invention, "several", "more than", and "a number of" represent at least two. The number can be any number, such as 2, 3, 4, 5, 6, 7, 8, or 9, and can even exceed nine.

[0058] Furthermore, in the description of the technical solution of this invention, unless otherwise explicitly specified / limited / restricted, the terms "set up," "install," "connect," "link," "provided with," "laid out," and "arranged" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to common connection methods in the art, such as welding, riveting, bolting, and threaded connections. Such connections can be mechanical, electrical, or communication connections; they can be direct connections or indirect connections through an intermediate medium; and they can refer to the internal communication between two components.

[0059] Example 1

[0060] This embodiment provides a multi-scenario driven method for dynamic simulation of watershed floods. See also... Figure 1 The method for simulating watershed flood dynamics includes the following steps:

[0061] S1. Meteorological, geographical, and hydrological data are fused and processed, and a surface grid is constructed based on the fused data. The surface grid is then used to delineate the catchment area of ​​the watershed.

[0062] S2. Design stormwater-runoff scenarios by combining fused data and watershed catchment areas, including:

[0063] Design a set of rainstorm scenarios based on historical rainfall sequences, climate characteristics, recurrence periods, and rainfall durations;

[0064] Construct the response relationship between runoff coefficient and the dynamic change of rainfall time series;

[0065] Construct rainstorm-runoff scenarios using a set of rainstorm scenarios and response relationships;

[0066] S3. Construct a flood simulation model to simulate the dynamics of floods under the rainstorm-runoff scenario.

[0067] This embodiment provides a multi-scenario driven watershed flood dynamics simulation method. It obtains a surface grid by fusing multi-source meteorological, geographical, and hydrological data, and delineates the watershed catchment area. Combining the fused data and the watershed catchment area, it designs storm-runoff scenarios, constructs a flood simulation model, and performs dynamic flood simulation under these scenarios. In designing the storm-runoff scenarios, this embodiment constructs a multi-dimensional storm scenario set based on the historical rainfall sequence and climate characteristics, return period, and rainfall duration of the study area. This set can cover different climatic conditions and extreme weather events, improving the adaptability to hydrological events under different climate change backgrounds.

[0068] Example 2

[0069] This embodiment is a further improvement on embodiment 1, and the repeated content will not be described again.

[0070] Preferably, S1 includes:

[0071] S11. Data Acquisition and Fusion:

[0072] The study involves acquiring raw meteorological, geographical, and hydrological data within the study area; standardizing the raw data to obtain a basic dataset; and fusing the basic dataset with digital elevation model data through spatial overlay analysis to obtain a raster dataset. Preferably, the standardization process includes at least: cropping, georegistration, resampling, abstraction, format conversion, and scene mapping.

[0073] Preferably, this embodiment integrates remote sensing imagery, digital elevation model (DEM), and geographic information data to fuse meteorological, geographical, and hydrological data of the study area.

[0074] In watershed hydrological calculations, the quality and completeness of basic data directly affect the accuracy and reliability of catchment area delineation. This embodiment addresses the needs of watershed hydrological calculations by constructing a multi-level, multi-source data fusion preprocessing mechanism. Through systematic data cleaning, transformation, and alignment, it eliminates data heterogeneity and noise interference, laying the foundation for automatic catchment area delineation.

[0075] First, the data undergoes preprocessing. Based on the different data sources and characteristics, the raw data is standardized through processes such as cropping, georegistration, resampling, abstraction, format conversion, and scene mapping. Specifically, cropping extracts data from the target area based on the study region; georegistration ensures that various data types are aligned under a unified coordinate system, eliminating spatial deviations caused by different coordinate systems; the resampling process uses bilinear interpolation or nearest neighbor algorithms to unify spatial resolution for data of different resolutions; abstraction extracts features relevant to hydrological calculations and removes irrelevant information; format conversion converts various data types into a unified data format supported by the system; and scene mapping establishes a mapping relationship between data and calculation parameters according to application scenario requirements. After these processes, multi-source heterogeneous data is transformed into a basic dataset with a unified spatial benchmark, standardized format, and clearly defined features. Multi-source heterogeneous data includes: geographic information data, DEM, high-resolution remote sensing imagery, land use types, land vegetation data, climate and hydrological data, and geological information data.

[0076] Then, data fusion is performed. Through spatial overlay analysis, the data layers in the basic dataset are merged with the Digital Elevation Model (DEM) to form a unified raster dataset. Specifically, a weighted overlay method is used for fusion, with weights determined based on the accuracy and relevance of the data source. This process ensures the spatial consistency of the data and avoids deviations caused by coordinate system differences.

[0077] Step S1 effectively reduced the impact of noise in the basic data on the water catchment area delineation, laying the foundation for the construction of a refined surface grid.

[0078] S12. Construct a three-level fine-grained grid using a raster dataset:

[0079] The continuous terrain surface is discretized into regular grid cells by rasterization to obtain the first-level macroscopic terrain grid.

[0080] An adaptive grid densification algorithm based on terrain features is used to densify the macroscopic terrain grid and obtain a second-level mesoscale feature grid.

[0081] A grid optimization algorithm based on hydrological processes was adopted to optimize the mesoscale feature grid by analyzing the confluence path and confluence time, thereby obtaining the third-level hydrological feature grid.

[0082] Preferably, the first level of the three-level refined grid is a macro-topographic grid, based on digital elevation model (DEM) data, which discretizes the continuous topographic surface into regular grid cells through rasterization. Specifically, the study area is divided into square grids with a uniform spatial resolution, and the height value of each grid cell is calculated by interpolation from the DEM data. The macro-topographic grid mainly reflects the large-scale topographic relief characteristics and is the basic grid for hydrological simulation.

[0083] Preferably, the second level of the three-level refined grid is a mesoscale feature grid. Based on the macroscopic topographic grid, it integrates hydrological feature data such as soil type and vegetation cover to refine key areas. This embodiment adopts an "adaptive grid densification algorithm based on topographic features," which analyzes topographic curvature and runoff weights to densify the grid of key topographic features such as valleys and watersheds, thereby achieving accurate representation of important hydrological features.

[0084] For valley areas, the formula for determining the mesh refinement factor based on the valley depth and slope is as follows:

[0085]

[0086] in, The mesh density factor for the valley area. The depth of the valley, The average elevation of the region. The slope of the valley.

[0087] For watershed regions, the formula for determining the mesh refinement factor based on the radius of curvature of the watershed is as follows:

[0088]

[0089] in, The grid density factor for the watershed region. For the minimum radius of curvature, The mean radius of curvature of the region.

[0090] Preferably, the third level of the three-level refined grid is a hydrological characteristic grid. Based on the first two levels of grids and combined with the requirements of the hydrological model, a refined grid that meets the accuracy requirements for hydrological process simulation is constructed. For the hydrological characteristics of small watersheds, a grid optimization algorithm based on hydrological processes is adopted. By analyzing the confluence path and confluence time, the grid division is optimized to ensure that key hydrological processes are accurately simulated.

[0091] Through the aforementioned three-level grid system, this embodiment achieves a refined representation of the surface of a small watershed, ensuring both the overall efficiency of hydrological simulation and a refined description of key areas. This effectively balances the contradiction between computational accuracy and efficiency, providing a high-precision spatial foundation for flood simulation in small watersheds.

[0092] Surface fine-scale grids are the core foundation for small watershed flood simulation, and their accuracy directly affects the accuracy of hydrological calculations. This embodiment constructs a three-level fine-scale grid system suitable for the characteristics of small watersheds based on multi-source fusion data, realizing a full-scale fine-scale representation of the surface from macro-topography to micro-surface features.

[0093] S13. Delineate the catchment area of ​​the watershed using a three-level fine grid:

[0094] Depression filling processing is performed on the DEM data;

[0095] Flow direction analysis was performed on a three-level fine grid using the D8 unidirectional flow algorithm.

[0096] Calculate the cumulative flow for each grid cell;

[0097] Based on the flow direction analysis and the calculation results of the cumulative runoff, the catchment point is determined, the flow direction is traced in reverse, and all grid areas flowing toward the catchment point are identified to form a preliminary catchment surface. The topographic curvature and the runoff weight factor are introduced to optimize the boundary morphology of the catchment surface for the first time. The boundary smoothing algorithm and topological relationship correction are used to optimize the boundary morphology of the catchment surface for the second time.

[0098] Automatic extraction of catchment boundaries is a crucial step in small watershed hydrological calculations, and its accuracy directly impacts the calculated hydrological parameters. This invention proposes an automatic catchment boundary extraction method suitable for small watersheds, based on digital elevation models (DEM) and flow direction analysis techniques, combined with topographic features and hydrological principles. Preferably, the small watershed involved in this invention can be defined according to the water conservancy industry standard "Small Watershed Division and Coding Specification" (SL 653–2013): a small watershed refers to a catchment unit whose area generally does not exceed 50 km² enclosed by the waterline (which can be understood as a catchment area not exceeding 50 km²).

[0099] Flow direction analysis is fundamental to catchment boundary extraction. This embodiment uses the D8 single-flow direction algorithm for flow direction calculation; see [link to relevant documentation]. Figure 2 This algorithm compares the elevation difference between the central grid and its eight adjacent grids, which are located in the eight directions of east, south, west, north, southeast, southwest, northeast, and northwest of the central grid. It determines the direction of the water flow as the direction of the greatest elevation drop. (See also...) Figure 2 and Figure 3 The flow coding of the D8 algorithm is as follows: Figure 3 As shown, 2 to the power of n (n=0,1,2,...,7) represents 8 possible flow directions.

[0100] Preferably, before performing flow direction analysis, the DEM data needs to be filled to eliminate the influence of artificial depressions in the terrain on the flow path. After filling, the runoff accumulation for each grid cell is calculated, representing the number of upstream grid cells flowing towards that grid cell. The formula for calculating the runoff accumulation is as follows:

[0101]

[0102] in, The current flow accumulation of the raster. This is the cumulative flow of the upstream grid cells flowing from the current grid cells out of the eight adjacent grid cells.

[0103] Based on flow direction analysis and the calculation results of runoff accumulation, this embodiment proposes a catchment boundary optimization method based on topographic features. The algorithm first determines the catchment point according to the research requirements, and then determines all grid areas flowing towards the catchment point by reverse tracing of the flow direction, forming a preliminary catchment surface. Then, the topographic curvature and runoff weight factor are introduced to optimize the catchment boundary morphology for the first time.

[0104] The calculation formula for optimizing the boundary morphology of the catchment surface by introducing topographic curvature and runoff weighting factor is as follows:

[0105]

[0106] In the formula, For the optimized boundary position, This represents the original boundary location. and , where K is the terrain curvature factor and W is the confluence weighting factor.

[0107] Among them, the terrain curvature factor K reflects the local curvature of the terrain, and the confluence weight factor W reflects the influence of the terrain on the confluence path. The calculation formulas are as follows:

[0108]

[0109]

[0110] in, The radius of curvature of the terrain; For local slope, This represents the maximum slope in the area. The slope angle is denoted by .

[0111] The initially extracted catchment surface boundaries often exhibit jagged irregularities, requiring further refinement. This embodiment employs a boundary smoothing algorithm and topological correction to perform a second optimization of the catchment surface boundary morphology. Boundary smoothing utilizes a moving average method to smooth the boundary point coordinates; the calculation formula is as follows:

[0112]

[0113]

[0114] in, , These are the smoothed boundary point coordinates. , These are the original boundary point coordinates. To smooth the window radius.

[0115] Watershed hydrological calculations are a crucial foundation for flood analysis and engineering design. Their main components include design storms, design floods, and design flows. Among these, catchment area delineation is the most fundamental and critical step in design flow calculation. Traditional catchment area delineation relies heavily on visual estimation and manual division, failing to accurately reflect the impact of topography on runoff, resulting in low accuracy and inefficiency. Hydrological methods based on Digital Elevation Model (DEM) data can reflect the influence of topographic relief on runoff direction to some extent, but they are highly dependent on data quality and preprocessing accuracy, and are prone to errors under complex terrain or multi-source data conditions. Node-based geometric methods and the area-to-pipe-length ratio method rely too heavily on pipe network layout and often simplify runoff to a linear relationship, ignoring the influence of factors such as pipe network undulation on the runoff path, leading to calculation results that are often overestimated and fail to accurately reflect the natural surface runoff morphology. The above-mentioned methods for delineating catchment areas have shortcomings in terms of automation, accuracy consistency, and adaptability to complex scenarios. Therefore, this embodiment provides an automatic method for delineating catchment areas that can integrate multi-source geographic information and automatically identify micro-topographic features through S1, so as to achieve accurate extraction of watershed boundaries and automatic parameter mapping.

[0116] Preferably, the catchment area of ​​the watershed provides two types of key parameters to support the design of the rainstorm-runoff scenario: one is the geometric characteristic parameters of the catchment area (length of the main channel). Average slope Used to calculate convergence time This allows us to determine the intensity distribution of the rainstorms and the peak location of the Chicago rain pattern; secondly, the catchment area. As the spatial basis for calculating runoff volume, the runoff per unit area This is converted into the total actual runoff flowing into the river channel.

[0117] See Figure 4 Preferably, after combining fused data and watershed catchment area to design rainstorm-runoff scenarios, flood simulation models are used to perform flood dynamic simulation of the rainstorm-runoff scenarios.

[0118] Preferably, in S2, a set of rainstorm scenarios is designed based on historical rainfall sequences and climate characteristics, including:

[0119] Based on the climate conditions and meteorological data of the study area, rainfall parameters, rainfall variation parameters, rainfall duration correction parameters, and rainstorm attenuation index are obtained, and rainstorm intensity is designed.

[0120] Rainfall intensity was assigned using the Chicago rainfall pattern method combined with relevant regional parameters of the study area;

[0121] A set of rainstorm scenarios was designed based on the recurrence period, rainfall duration, and spatial distribution characteristics of rainstorms.

[0122] Preferably, in S2, constructing the response relationship of runoff coefficient with the dynamic change of rainfall time sequence includes:

[0123] Calculate the runoff curve over time based on soil conditions, soil saturation state, soil permeability, and rainfall duration;

[0124] Calculate the potential maximum retention capacity based on the number of runoff curves;

[0125] Calculate runoff based on potential maximum retention and rainfall;

[0126] The runoff coefficient is determined by the ratio of instantaneous runoff to current rainfall, and the response relationship of the runoff coefficient to the dynamic change of rainfall time sequence is constructed.

[0127] Preferably, when designing multiple rainstorm scenarios to generate a rainstorm scenario set, key factors such as regional climate characteristics, return period, and rainfall duration need to be considered. Based on regional meteorological data, a rainstorm intensity formula suitable for local characteristics should be designed:

[0128]

[0129] Among them, 1. E. , For parameters to be determined, specifically, 1 represents the rainfall intensity parameter, E represents the rainfall intensity variation parameter, b represents the rainfall duration correction parameter, and n represents the rainstorm attenuation index. These parameters are determined based on local climate conditions. To design the intensity of rainstorms , Design return period (years). The duration of rainfall (in minutes).

[0130] In terms of storm rainfall pattern design, the Chicago rainfall pattern method is adopted, which can better reflect the temporal distribution characteristics of rainfall. The Chicago rainfall pattern is based on the watershed surface extraction... and Calculate the convergence time The peak rainfall time (0.375τ) was determined, and the rainfall process was divided into an upward phase and a downward phase.

[0131] Convergence Time The calculation uses the following formula:

[0132]

[0133] in, The length of the main ditch (km). The average slope (‰) of the main ditch. , This is a regional empirical coefficient, the value of which is determined by looking up a table based on the geographical and climatic zoning of the specific study area. For specific values, please refer to the "Design Code for Highway Culverts".

[0134] The formula for rainfall intensity distribution is:

[0135]

[0136] in, Let be the rainfall intensity at time t, where t∈[0,T], which forms the basis for calculating runoff per unit time. , , For region-related parameters, specifically, denoted as , b as the rainfall peak intensity parameter, n as the rainfall duration correction parameter, and n as the rainfall attenuation index.

[0137] Taking into account the return period (10 years, 50 years, 100 years) and rainfall duration (12 hours, 24 hours), a multi-dimensional rainstorm scenario set is constructed, and typical rainstorm scenario examples are designed.

[0138] Table 1: Examples of typical rainstorm scenario designs

[0139]

[0140] The runoff coefficient is a key parameter connecting rainfall and runoff, influenced by various factors including antecedent soil moisture, rainfall intensity, and land cover type. Traditional methods typically use a fixed runoff coefficient, which fails to reflect the dynamic changes in runoff response during rainfall. This embodiment proposes a dynamic runoff coefficient construction method based on the SCS (Soil Conservation Service Curve Number) model (SCS-CN). The core of this method is the magnitude of the Curve Number (CN) value, a dimensionless parameter ranging from 0 to 100. A larger CN value indicates a greater probability of runoff generation. The basic formula of the SCS-CN method is:

[0141]

[0142] in, Runoff volume (mm) Rainfall (mm); The potential maximum retention (mm) is calculated using the following formula:

[0143]

[0144] In small watershed flood simulations, the CN value needs to consider the influence of rainfall time series. However, the CN value exhibits a non-linear variation with rainfall duration, therefore a time dynamic factor is introduced. Construct a formula for calculating the CN value as it changes over time:

[0145]

[0146] in, This is the initial CN value (reflecting previous soil conditions). The maximum CN value (reflecting soil saturation). This is the attenuation coefficient, which is related to soil permeability. These represent the various moments within the duration of the rainfall.

[0147] runoff coefficient It can then be determined by the ratio of instantaneous runoff to rainfall:

[0148]

[0149] in, Instantaneous runoff, This is the current rainfall. .

[0150] Based on instantaneous runoff coefficient With rainfall intensity The catchment area is obtained by automatically dividing the catchment surface. The design flow rate at the outlet section of the watershed can be calculated:

[0151]

[0152] This flow rate serves as the inflow boundary condition for dynamic flood simulation, driving the cellular automata model to perform flood evolution calculations.

[0153] By combining parameters such as land use type, soil hydrological grouping, and topographic slope, the response relationship of runoff coefficient with the dynamic change of rainfall duration was constructed. As shown in Table 2, there are significant differences in runoff coefficients for different land use types under different rainfall durations.

[0154] Table 2: Dynamic changes in runoff coefficients for different land use types

[0155]

[0156] Preferably, after obtaining the rainstorm scenario set and the response relationship of the runoff coefficient with the dynamic change of the rainfall time sequence, a rainstorm-runoff scenario can be constructed based on the rainstorm scenario set and the response relationship of the runoff coefficient with the dynamic change of the rainfall time sequence.

[0157] Preferably, the flood simulation model is constructed in S3, including: constructing a flood simulation model based on a cellular automata model. The cellular automata model is as follows: Figure 5 As shown, the flood simulation model divides the study area into discrete square grid cells with a preset resolution, constructing a computational space composed of several cells. Within the framework of discrete space and discrete time, it simulates the spatiotemporal process of flood flow in the basin. The evolution of the cells is determined by the hydraulic state (velocity and water level changes) of the initial inflow boundary (such as the basin outlet section or distributed runoff generation unit); and the flood evolution of adjacent cells is calculated based on the height, roughness, water depth, unit width discharge in the X direction, and unit width discharge in the Y direction of the cell's location.

[0158] Preferably, based on the refined elevation grid and the results of the storm-runoff multi-scenario design, this embodiment constructs a flood simulation model based on cellular automata. The flood simulation model defines a gridded surface of the study area, constructing a computational space composed of numerous cells. Within a discrete space and discrete time framework, it simulates the spatiotemporal process of flood flow in the watershed, thereby achieving accurate simulation of the watershed flood evolution process.

[0159] A cellular automaton consists of a quadruple of a cell, its neighborhood, its state, and its transition rule.

[0160]

[0161] Where C represents a complete cellular automaton system; G represents the cell space; D represents the dimension of the cell space, which is a positive integer; Q represents a finite, discrete set of states of the cell; N represents the cell neighborhood; and F represents the transition rule. The state set Q can be specifically represented as:

[0162]

[0163] v, d, r, z These are hydrological state variables, representing flow velocity, water depth, roughness, elevation, and time, respectively.

[0164] Preferably, the cellular space G corresponds to the gridded surface of the study area; the spatial dimension D is 2 (two-dimensional plane); the state set Q is specified as state variables such as flow velocity, water depth, roughness, and height; the neighborhood N adopts the Mohr neighborhood structure; and the transformation rule F is specified as the Saint-Venant equations.

[0165] Flood simulation models, by defining cell states and transition rules, can effectively simulate the temporal evolution of complex systems. In watershed flood simulation, the study area is divided into discrete square grid cells with a certain resolution, and a fixed time step is used. Simulate the evolutionary process.

[0166] According to the calculation method for flood evolution, the evolution of a cell is determined by the hydraulic state (velocity and water level changes) of the initial inflow boundary (such as the watershed outlet section or distributed runoff generation unit). When calculating the flood evolution of adjacent cells, the following five variable attributes need to be considered comprehensively: the height (DSM) of the cell location, roughness coefficient, water depth, unit width discharge in the X direction, and unit width discharge in the Y direction. Based on this, five state attributes are set for each cell space to describe its dynamic evolution characteristics.

[0167] The flood evolution calculation uses a discretized version of the Saint-Venant equations, and the calculation process is as follows:

[0168] 1. The water surface height and flow rate of the inflow boundary cell in the initial state are derived from the watershed hydrological scenario data and the rainfall data calculated in S2. Except for the inflow boundary region, all other cells are set to a waterless state, and the water depth and directional unit width flow rate at their locations are initialized to 0.

[0169] 2. Calculate the unit width discharge at time t+1 from the unit water depth at time t. The calculation method is as follows:

[0170]

[0171] 3. Then, calculate the cell depth at time t+1 from the cell unit width flow rate at time t. The calculation method is as follows:

[0172]

[0173] The meanings of the parameters in the formula are as follows:

[0174] , , , , , These represent the unit width flux of cell (i, j) in the X and Y directions at time t, respectively. , water depth Water surface elevation Flow velocity in the X and Y directions ; The roughness of cell (i, j) ; The unit of time represents the iteration; g is the acceleration due to gravity. , Cellular size in the X and Y directions .

[0175] Meanwhile, to ensure that the overall effect of the flood evolution simulation conforms to physical reality, non-physical situations in the evolution process are constrained and controlled. The cell state transition rules are as follows: if there is water around the cell and the number of neighboring cells with water is greater than 7, the cell state changes to water; if there is water around the cell and the number of neighboring cells with water is less than 2, the cell state changes to no water.

[0176] Finally, the flood simulation model outputs disaster information for the study area at each time step, that is, disaster information for the time period from time T=T0 to time T=Tn. The disaster information includes: inundated area, water depth, flow velocity, and time.

[0177] See Figure 6 When constructing a flood simulation model, a cellular automata model can be built based on the rainstorm-runoff scenario, the geographical information of the study area, and the high-precision surface grid of the study area to generate the flood simulation model. The flood simulation model simulates the disaster situation from time T=T0 to time T=Tn.

[0178] The storm-runoff scenario obtained through S2 includes: design flow rate, catchment area, inflow boundary, runoff coefficient, etc.

[0179] Geographic information includes surface roughness, grid resolution, etc.

[0180] Preferably, this embodiment also visualizes the flood dynamic simulation, including: obtaining the evolution results of the flood simulation model at each time step; and using the water depth information of the evolution results to construct a triangular network for flood modeling.

[0181] To intuitively represent the dynamic process of floods, this invention proposes a method for visualizing and simulating the spatiotemporal evolution process. See also... Figure 7 Specifically, a triangular mesh is constructed frame by frame based on the index of the simulation results to create a dynamic evolution effect. Since the flood simulation model is a two-dimensional cellular automaton flood model, the obtained simulation results actually represent water depth elements based on the inundation area represented by a two-dimensional array of cellular units. It is necessary to elevate the two-dimensional model to form a three-dimensional model and present the effect of dynamic undulation of the water surface frame by frame to make the visualization effect more realistic.

[0182] The flood simulation model calculates a set of evolution results at each time step. The evolution results record water depth information, flow velocity information, etc. at that time point. The water depth information in the evolution results is used to construct a triangular network for flood modeling.

[0183] The output of the flood simulation model provides detailed and reliable data support for three-dimensional dynamic visualization, thereby more comprehensively showcasing the evolution characteristics of flood processes in small watersheds.

[0184] Preferably, after constructing the basic geographic scene of the study area, grid cells with water depth values ​​greater than 0 are identified as modeling objects by reading the water depth file output from the flood simulation model; a linked list is constructed to store relevant vertex attribute information, including parameters such as the center point coordinates of the water-bearing grid, water depth, flow velocity, and elevation; simultaneously, a triangle index table is generated, recording the three vertex numbers of each water surface triangle in clockwise order. During the modeling process, the vertex indices in the triangle vertex index table are extracted sequentially according to the numbers of the water surface triangles, and then the spatial position of each vertex and its corresponding water depth attribute are combined to complete the reconstruction of the three-dimensional flood model; finally, the three-dimensional flood model is rendered frame by frame, and the corresponding color bands are matched according to the depth information of the triangles to present a visual effect of undulating gradients.

[0185] This embodiment provides a multi-scenario driven dynamic simulation method for watershed floods:

[0186] First, the automatic delineation of watershed catchment areas through multi-source data fusion is achieved by integrating data such as DEM, high-resolution remote sensing images and land use types, performing standardized preprocessing and multi-dimensional information fusion, constructing attribute-enhanced map grids, and introducing an adaptive grid densification algorithm for key hydrological feature areas to establish a dynamic resistance coefficient model. Through iterative optimization, accurate boundary extraction driven by physical mechanisms is achieved.

[0187] Secondly, multi-scenario driven watershed flood dynamics simulation designs multiple intensities of rainstorm scenarios based on historical rainfall sequences and climate characteristics, and constructs dynamic response relationships of runoff coefficients.

[0188] Then, a cellular automata simulation model is established using a refined grid to simulate the flood flow process within a discrete spatiotemporal framework, and the spatiotemporal evolution is visualized through dynamic rendering technology.

[0189] The target of this embodiment of watershed flood dynamic simulation is limited to "small watersheds" to highlight the innovative advantages of this invention in refined, small-scale simulation; however, this invention is not absolutely limited to small watersheds, but rather proposes a technical solution that integrates multi-source high-resolution data, refined grid construction and cellular automata dynamic simulation, taking into account the characteristics of small watershed flood processes such as fast response and strong suddenness.

[0190] Example 3

[0191] This embodiment provides a multi-scenario driven watershed flood dynamic simulation system, which can execute the multi-scenario driven watershed flood dynamic simulation method involved in Embodiment 1, Embodiment 2, or Embodiment 3.

[0192] Preferably, the multi-scenario driven watershed flood dynamic simulation system includes an input unit, an output unit, and a processing unit.

[0193] The input unit is used for front-end user interaction. Users can input basic data and related parameters such as meteorological, geographical, and hydrological data through the input unit.

[0194] The processing unit is used to process the data input by the user through the input unit.

[0195] The processing unit includes: a flood simulation module, a storage module, and a visualization processing module.

[0196] The flood simulation module can design rainstorm-runoff scenarios, set flood simulation model parameters, and perform flood evolution calculations based on basic data and related parameters such as meteorological, geographical, and hydrological data input by the input unit, thereby realizing dynamic flood simulation.

[0197] The storage module stores the simulation results of the flood simulation module and generates a triangle index table by combining it with the terrain file of the study area.

[0198] The visualization module reads simulation results from the storage module to display disaster information and constructs a basic geographic scene of the study area based on the terrain file. The visualization module can also read the water depth file output by the flood simulation model and, in conjunction with the basic geographic scene of the study area, reconstruct a 3D flood model, rendering a 3D flood model.

[0199] The output unit is used to display processing results / view input information, intermediate parameters, etc. Preferably, the output unit can display disaster information and a 3D flood model.

[0200] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-scenario driven method for dynamic simulation of watershed floods, characterized in that, Includes the following steps: S1. Meteorological, geographical, and hydrological data are fused and processed, and a surface grid is constructed based on the fused data. The surface grid is then used to delineate the catchment area of ​​the watershed. S2. Designing a rainstorm-runoff scenario by combining the fused data and the watershed catchment area, including: A set of rainstorm scenarios was designed based on the historical rainfall sequence and climate characteristics, recurrence period, and rainfall duration of the study area. Constructing the response relationship of runoff coefficient to the dynamic change of rainfall time series, including: Calculate the runoff curve over time based on soil conditions, soil saturation state, soil permeability, and rainfall duration; in, The initial CN value, The maximum CN value is the numerical value of the runoff curve. This is the attenuation coefficient, which is related to soil permeability. These represent various moments within the duration of the rainfall. Calculate the potential maximum retention based on the number of runoff curves; Calculate runoff based on the potential maximum retention and rainfall; The runoff coefficient is determined based on the ratio of instantaneous runoff volume to current rainfall, and the response relationship of the runoff coefficient to the dynamic change of rainfall time sequence is constructed. in, This is the runoff coefficient. Instantaneous runoff, This is the current rainfall. ; Based on instantaneous runoff coefficient With rainfall intensity Combined with the water catchment area obtained from the division Calculate the design flow rate at the outlet section of the watershed: The flow As an inflow boundary condition for flood dynamics simulation; A rainstorm-runoff scenario is constructed using the rainstorm scenario set and the response relationship; S3. Construct a flood simulation model based on cellular automata, perform flood dynamic simulation on the rainstorm-runoff scenario, and calculate flood evolution using the discretized Saint-Venant equations. The inflow boundary conditions are used to drive the cellular automata model for flood evolution calculation. The dynamic evolution characteristics of the cells are described by the height (DSM), roughness, water depth, unit width flow rate in the X direction, and unit width flow rate in the Y direction at the location of the cells.

2. The multi-scenario driven watershed flood dynamic simulation method according to claim 1, characterized in that, S1 includes: S11, data acquisition and fusion, including: The study involves acquiring raw meteorological, geographical, and hydrological data within the research area; standardizing the raw data to obtain a basic dataset; and fusing the basic dataset with digital elevation model data through spatial overlay analysis to obtain a raster dataset. The standardization process includes at least: cropping, georegistration, resampling, abstraction, format conversion, and scene mapping.

3. The multi-scenario driven watershed flood dynamic simulation method according to claim 2, characterized in that, S1 further includes: S12, constructing a three-level refined grid using the raster dataset, including: The continuous terrain surface is discretized into regular grid cells by rasterization to obtain the first-level macroscopic terrain grid. An adaptive grid densification algorithm based on terrain features is used to densify the macroscopic terrain grid to obtain a second-level mesoscale feature grid. A grid optimization algorithm based on hydrological processes is used to optimize the mesoscale feature grid by analyzing the confluence path and confluence time, thereby obtaining a third-level hydrological feature grid.

4. The multi-scenario driven watershed flood dynamic simulation method according to claim 3, characterized in that, S1 further includes: S13, using the three-level refined grid to divide the watershed catchment area, including: Depression filling processing is performed on the DEM data; The flow direction analysis of the three-level refined grid was performed using the D8 single-flow algorithm. Calculate the cumulative flow for each grid cell; Based on the flow direction analysis and the calculation results of the cumulative runoff, the catchment point is determined, the flow direction is traced in reverse, and all grid areas flowing toward the catchment point are identified to form a preliminary catchment surface. The topographic curvature and the runoff weight factor are introduced to optimize the boundary morphology of the catchment surface for the first time. The boundary smoothing algorithm and topological relationship correction are used to optimize the boundary morphology of the catchment surface for the second time.

5. The multi-scenario driven watershed flood dynamic simulation method according to claim 4, characterized in that, In S2, the design of a rainstorm scenario set based on historical rainfall sequences and climate characteristics includes: Based on the climate conditions and meteorological data of the study area, rainfall parameters, rainfall variation parameters, rainfall duration correction parameters, and rainstorm attenuation index are obtained, and rainstorm intensity is designed. The Chicago rainfall pattern method was used to assign rainfall intensity in the study area; A set of rainstorm scenarios was designed based on the recurrence period and rainfall duration.

6. The multi-scenario driven watershed flood dynamic simulation method according to claim 5, characterized in that, S3 The flood simulation model divides the study area into discrete square grid cells according to a preset resolution, constructs a computational space composed of several cells, and simulates the spatiotemporal process of flood flow in the basin within the framework of discrete space and discrete time. The evolution of a cell is determined by the hydraulic state of the initial inflow boundary; and the flood evolution of adjacent cells is calculated based on the height, roughness, water depth, unit width discharge in the X direction, and unit width discharge in the Y direction of the cell location.

7. The multi-scenario driven watershed flood dynamic simulation method according to claim 6, characterized in that, The rules constrain and control non-physical aspects of the evolution process, including: Water conservation: The sum of the remaining rainwater and the total water volume of the cells must not exceed the designed total intrusion water volume to ensure the total water balance in the simulation; Traffic constraints: The increase in the bandwidth of a single cell cannot exceed the maximum bandwidth it can provide, in order to comply with physical traffic limits; Cell state transition: If there is water around the cell and the number of neighboring cells with water is greater than 7, the cell state changes to water; if there is water around the cell and the number of neighboring cells with water is less than 2, the cell state changes to no water.

8. The multi-scenario driven watershed flood dynamic simulation method according to claim 7, characterized in that, It also includes visualization of flood dynamics simulations, including: Obtain the evolution results of the flood simulation model at each time step; The water depth information from the evolution results is used to construct a triangular network for flood modeling.

9. A multi-scenario driven watershed flood dynamic simulation system, characterized in that, Perform the multi-scenario driven watershed flood dynamic simulation method as described in any one of claims 1 to 8.

Images