Urban flood simulation and safety assessment method

Abstract

The invention discloses an urban flood simulation and safety evaluation method, and belongs to the technical field of hydrological data processing and flood rehearsal. The method comprises the following steps: firstly, collecting DEM data of a to-be-processed range, performing projection depression filling preprocessing through ArcGIS software, calculating a flow direction and a flow rate, extracting a river network, and constructing a catchment area and a rainfall flood model; carrying out rainfall simulation and one-dimensional hydrodynamic force calculation by using SWMM software to obtain a river channel and node time sequence flow; then, the flow data are input into HEC-RAS software, two-dimensional hydrodynamic simulation is completed in combination with the terrain, and the flood inundation range and depth are obtained; finally, repeatedly simulating rainfall in different recurrence periods, and dividing flood risk areas according to indexes such as a water depth threshold value and a flooded area. According to the method, full-process automatic processing from rainfall input to flood routing is realized, the timeliness and precision of flood rehearsal are improved, and support is provided for disaster prevention and reduction decisions.

Description

Technical Field

[0001] This invention relates to the field of hydrological data processing and flood prediction technology, specifically to a method for urban flood simulation and safety assessment. Background Technology

[0002] Floods are natural disasters caused by sudden increases in water volume or rapid rises in water levels due to natural factors such as heavy rainfall, snowmelt, and storm surges. They are among the major natural disasters threatening human life and property. In recent years, influenced by global climate change and accelerated urbanization, extreme rainfall events have become more frequent, significantly increasing the frequency and destructive power of floods and severely impacting urban safety, infrastructure, and socio-economic activities. Therefore, how to conduct rapid and accurate flood forecasting and risk assessment has become a key technical issue in disaster prevention, mitigation, and watershed management.

[0003] Traditional methods for calculating small watershed floods primarily rely on topographic maps, hydrological handbooks, and empirical formulas, involving the manual extraction of parameters such as watershed area, river length, gradient, and average rainfall for step-by-step calculations. These methods are cumbersome and inefficient, making large-scale batch processing difficult and prone to significant human error. They fail to meet the real-time and accuracy requirements of the "four early warnings" (forecasting, early warning, rehearsal, and contingency planning) for flash flood disaster prevention. Especially in areas with complex terrain and scarce data, traditional methods are insufficient in depicting the formation and evolution of floods and fail to reflect flood dynamics.

[0004] Currently, flood simulation research mainly employs two technical approaches: one is based on empirical hydrological models, which are computationally simple but have limited accuracy; the other is based on physical mechanisms and hydrodynamic models, which can simulate the entire process of rainfall-runoff-confluence-flood evolution and output key elements such as flow velocity, water depth, and inundation extent. However, existing models such as SWAT, HEC, and MIKEFLOOD still have shortcomings in terms of data interfaces, coupling efficiency, and spatial scale adaptability, especially in the multi-dimensional, rapid-response flood simulation scenarios where computational efficiency is low.

[0005] With the development of Geographic Information System (GIS) technology, spatial data analysis and visualization capabilities have provided new support for flood simulation. GIS can integrate topographic, hydrological, land use, meteorological, and real-time monitoring data, and through spatial overlay analysis and dynamic visualization, it can achieve a detailed representation of flood propagation paths and inundation ranges. However, traditional GIS-based water balance analysis remains at a static spatial level and cannot dynamically reflect the spatiotemporal changes during flood processes.

[0006] Therefore, there is an urgent need for a comprehensive method that combines GIS spatial analysis capabilities with rainfall-runoff physical simulation to achieve coupled calculation of the entire process from rainfall input to flood evolution. This method should have high spatial accuracy and meet the needs of rapid simulation and risk assessment, thus providing scientific support for flood prevention and disaster reduction. Summary of the Invention

[0007] To address the aforementioned issues, this invention proposes a method for urban flood simulation and safety assessment. It constructs a comprehensive flood prediction process that integrates GIS spatial analysis, hydrological simulation, and hydrodynamic calculation, achieving fully automated processing from rainfall input to flood evolution. This provides reliable technical support for regional flood risk assessment and disaster prevention and mitigation decision-making.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: a method for urban flood simulation and safety assessment, comprising the following steps:

[0009] S1: Collect DEM data of the area to be processed, and place the DEM data in ArcGIS software for projection and depression filling preprocessing;

[0010] S2: Based on the DEM tif file, quickly calculate the flow direction and flow rate in ArcGIS software, extract the river network and construct the catchment area, and realize the construction of the rainfall and flood model;

[0011] S3: Based on SWMM software, rainfall simulation is performed on the rainfall and flood model, and one-dimensional hydrodynamic calculations are carried out to obtain the time-series flow of each river channel and node;

[0012] S4: Using the time-series flow data calculated by SWMM software as input, perform two-dimensional hydrodynamic simulation on the corresponding terrain based on HEC-RAS software to obtain the final inundation range and depth of the flood caused by extreme rainfall;

[0013] S5: Process the flooded areas under different return periods of rainfall in step S4, overlay the output images, determine whether the area is prone to flooding, and then delineate flood risk zones.

[0014] Furthermore, in step S1, after the DEM data is projected using ArcGIS software, the method also includes filling depressions in the terrain data to eliminate false depressions and ensure the continuity of surface water flow paths, thereby improving the accuracy of subsequent flow direction and confluence calculations.

[0015] Furthermore, in step S2, based on the extracted river network layer, the river is classified by analyzing the river length, catchment area and topological relationship to distinguish the main river channel from the tributary system; and the DEM file after depression filling is converted into a vector format shapefile and spatially connected with the river network nodes to generate a basic model file for rainfall-runoff simulation.

[0016] Furthermore, in step S3, the rainfall-runoff simulation uses SWMM software for one-dimensional hydrodynamic calculations. The model sets the drainage outlet, rainfall sequence, simulation duration, and calculation step size. The model is automatically run and batch results are extracted through the interface between Python software and SWMM software to obtain the flow time series data of the river channel and nodes.

[0017] Furthermore, in step S4, the two-dimensional hydrodynamic simulation uses HEC-RAS software, with DEM as the calculation basis, the simulation area is delineated according to the river network distribution, and a flow input port is set at the boundary node. The flow time series data obtained by SWMM software simulation is used as the input boundary condition to perform two-dimensional flood evolution simulation, and the results of water depth and inundation range at different time steps are output.

[0018] Furthermore, in step S5, steps S3-S4 are repeated for rainfall conditions with different return periods to obtain the flood inundation range and water depth distribution under each scenario. Based on indicators such as water depth threshold, inundated area and duration, flood risk levels are classified to achieve regional flood risk zoning assessment and level determination.

[0019] Advantages and beneficial effects of this invention: Based on DEM (Digital Elevation Model) topographic data, this invention utilizes ArcGIS software to automatically extract watershed spatial features, combines SWMM software for rainfall-runoff simulation, and further employs HEC-RAS software for two-dimensional hydrodynamic calculations. This enables accurate acquisition of flood inundation range and water depth distribution under different return period rainfall conditions. This method achieves fully automated processing from rainfall input to flood evolution, significantly improving the timeliness and accuracy of flood forecasting, and providing reliable technical support for regional flood risk assessment and disaster prevention and mitigation decision-making. Attached Figure Description

[0020] Figure 1 This is a flowchart of the steps of the urban flood simulation and safety assessment method provided in the embodiments of the present invention;

[0021] Figure 2 This is a schematic diagram of the DEM model of Shapingba District, the demonstration area provided in this embodiment of the invention;

[0022] Figure 3 This is a schematic diagram of the flow direction for DEM calculation in Shapingba District provided in an embodiment of the present invention;

[0023] Figure 4 This is a schematic diagram of the river network in Shapingba District provided in an embodiment of the present invention;

[0024] Figure 5 This invention provides a schematic diagram of a generalized river network in Shapingba District.

[0025] Figure 6 This is a schematic diagram of a rainfall-runoff model for Shapingba District provided in an embodiment of the present invention;

[0026] Figure 7 This is a schematic diagram of the rainfall-runoff simulation results provided in an embodiment of the present invention;

[0027] Figure 8 This is a schematic diagram of the HEC-RAS model provided in an embodiment of the present invention;

[0028] Figure 9 This is a schematic diagram of the running results of the HEC-RAS model provided in the embodiment of the present invention;

[0029] Figure 10 This is a schematic diagram of the flood risk zone division in Shapingba District provided in an embodiment of the present invention; Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] This invention discloses a method for flood prediction and flood risk assessment based on GIS and rainfall-runoff simulation. The specific process is as follows: Figure 1 As shown, it includes the following steps:

[0032] Consult local regulations to obtain the rainfall data for the corresponding historical return period, and calculate the corresponding rainfall intensity curve for subsequent simulations;

[0033] Before conducting flood simulations, typical rainfall data under different return periods are collected based on meteorological, hydrological, and planning data for the target area, such as 1-year, 5-year, 10-year, 50-year, and 100-year return periods. Through statistical analysis of these historical rainfall data, the relationship between rainfall duration and rainfall intensity is fitted using rainfall intensity formulas or empirical models to generate a rainfall intensity curve. This curve reflects the pattern of regional rainfall changes over time and is an important input parameter for subsequent calculations of rainfall runoff and flood evolution.

[0034] Obtain the 30m resolution digital elevation model (DEM) of the corresponding region, such as... Figure 2 As shown, the tif file is imported into ArcGIS for projection processing because most original DEM coordinates are in latitude and longitude format, which is not suitable for subsequent flood simulation. It needs to be converted into a projected coordinate system.

[0035] To ensure the accuracy of topographic features and the spatial precision of the simulation results, DEM data with a resolution of 30m was selected as the base topographic file. After importing the DEM data into ArcGIS software, a suitable projection coordinate system (such as WGS_1984_UTM_Zone_XXN) was selected based on the geographical location of the study area, and the TIFF format data was transformed by projection. This step can eliminate the inaccuracies in distance and area calculations under latitude and longitude coordinates, making it meet the spatial calculation requirements of subsequent flow direction analysis and hydrodynamic simulation.

[0036] To improve the spatial accuracy of subsequent hydrodynamic simulations, this invention explicitly defines the projected coordinate system and DEM preprocessing parameters. The study area is located at 106°15′–106°30′ E, corresponding to the 48N zone of the UTM projection. Therefore, WGS_1984_UTM_Zone_48N is selected as the unified projected coordinate system to ensure the accuracy of distance, slope, and area calculations. During the DEM depression filling operation, to avoid over-smoothing of real depressions, this invention limits the depression filling depth threshold to no more than 0.5m. When the depression depth exceeds this threshold, only local smoothing is performed without complete filling. By setting this threshold, erroneous alterations to the terrain structure can be avoided, while ensuring the spatial continuity of the confluence path.

[0037] The reprojected tif file will then undergo a fill operation to fill any gaps that may exist in the tif file;

[0038] Topographic preprocessing, including depression filling, is performed on the projected DEM to eliminate spurious depressions caused by sampling errors or interpolation algorithms. Depression filling ensures the continuity of surface water flow, avoids interruptions in flow paths or incorrect convergence in runoff calculations, and provides a smooth topographic surface for flow direction and flow rate calculations.

[0039] Perform flow direction calculation on the tif file using the D8 algorithm, and save it as a new layer as follows: Figure 3 As shown;

[0040] Based on the preprocessed DEM, the flow direction of each raster cell was calculated using ArcGIS's "Flow Direction" tool, and the D8 algorithm was used to determine the maximum slope direction of each cell. The calculation results were stored in the form of direction codes to generate a flow direction layer, providing topographic constraints for subsequent runoff calculations and river network extraction. Runoff accumulation was calculated on the new layer to obtain a runoff accumulation map, and a flow threshold was defined.

[0041] The "Flow Accumulation" tool is used to accumulate the upstream confluence units of each pixel to obtain a cumulative surface runoff distribution map. Based on the watershed size and topographic features, topological analysis is performed on the extracted results to calculate the break ratio and false stream ratio. When the threshold is below 1000, overly dense river channels and false confluence paths are likely to occur; when the threshold is above 5000, tributary loss and river network fragmentation may result. Therefore, 1000–5000 is determined as the optimal calculation range, and a suitable confluence threshold is set to distinguish between major confluence paths and general surface runoff areas, thus providing a basis for river channel identification and extraction.

[0042] The flow map is filtered based on flow thresholds to obtain a preliminary river network generated according to the terrain, such as... Figure 4 As shown. Using the aforementioned flow threshold, the confluence accumulation map is filtered to extract the main confluence channels controlled by the terrain, generating a preliminary river network layer. This river network reflects the potential water flow paths under natural terrain conditions, providing a foundation for the construction of the river network in subsequent models;

[0043] The river network layer is hierarchically processed to classify rivers into different levels, forming the final generalized river network, as shown below. Figure 5 As shown;

[0044] Based on river length, drainage area, and confluence relationships, the extracted river network is classified and graded to distinguish between the main channel and tributary systems. This invention uses the Strahler classification method as the standard classification method for river network grading. Its calculation rules are as follows: the source segment (without upstream tributaries) is defined as first-order; when two segments of the same order converge, the downstream segment's grade increases by one level; when segments of different orders converge, the downstream segment's grade is the higher grade among the upstream segments. For the case of multiple tributary confluences, if the highest upstream grade r_max occurs at least twice, the downstream grade is r_max+1; otherwise, the downstream grade is r_max. In implementation, this invention utilizes the ArcGIS Hydrology toolbox for classification: After preprocessing the DEM using Fill, FlowDirection, and Flow Accumulation to generate a confluence accumulation map, the river network is extracted using empirical or calibration thresholds (generating a binary river network raster). The river network is then divided into independent links using Stream Link, and the Strahler grade for each segment is calculated using Stream Order (selecting the Strahler method). Finally, the grades are written into a vector attribute table for subsequent parameter assignment (roughness, gradient, width, etc.). To enhance the reliability of the classification, this invention can use Shreve values ​​as an auxiliary discrimination indicator in complex or urbanized areas, and performs length and connectivity filtering on excessively short or suspected pseudo-river segments before classification and parameter assignment. The classification rules described in this section are consistent with the topology calculation logic of ArcGIS tools, ensuring the repeatability and traceability of the river network classification results. It also provides clear and operable classification standards for subsequent hydrodynamic parameter zoning and model boundary condition setting. Furthermore, the ArcGIS river network classification tool can automatically generate river grade attribute fields, facilitating subsequent hydraulic feature analysis and simulation structure definition.

[0045] The processed DEM raster file is converted to vector shapefile (shp) format for spatial overlay and neighborhood analysis with the river network node layer. Spatial join operations are used to determine watershed boundaries, river network nodes, and catchment area relationships, generating a basic input file that can be imported into the SWMM model, such as... Figure 6 As shown, topographic and structural inputs are provided for rainfall-runoff simulation;

[0046] The generated .inp file was imported into SWMM. Drainage outlets and rainfall sequences were designed, with the outlets located in four low-lying areas at the edge of the DEM region. Rainfall time series with different return periods and a resolution of 1 minute, totaling 2 hours of rainfall duration, were used. Then, the simulation duration and simulation step size were defined. The duration was set to 3 hours to cover the rainfall duration and the 1-hour one-dimensional water flow duration. The simulation step size was designed to be 1 minute, which enabled relatively fine output of flow simulation results. Subsequently, a one-dimensional hydrodynamic simulation was performed in SWMM.

[0047] Load the model file generated in the previous step into the SWMM software, arrange the drainage outlets according to the actual drainage system of the area, input the rainfall intensity curve and duration data, and set the simulation time and time step. Run the SWMM model to obtain the runoff generation and confluence process under the influence of rainfall, and generate the flow time series output for each node and river segment, as shown below. Figure 7 As shown, the Python interface with SWMM is used to batch acquire flow time-series data from SWMM simulation results, which will then be used as data input for subsequent two-dimensional hydrodynamic simulations.

[0048] By connecting to the SWMM API interface via Python scripts, the model can be run automatically and results extracted. The output nodal flow time series are exported in batches and formatted, and used as boundary condition inputs for the two-dimensional hydrodynamic model (HEC-RAS), thus achieving data coupling between the one-dimensional and two-dimensional models.

[0049] In ArcGIS, river networks are segmented and combined, dividing them into main streams and tributaries. The gradient of the main stream and tributaries is calculated using ArcGIS's built-in Python implementation. The specific calculation formula is as follows: Where S is the river gradient (dimensionless, usually expressed as m / m), ΔH is the elevation difference (m) from the beginning to the end of the river segment, and L is the actual length of the river channel (m), used for subsequent two-dimensional hydrodynamic simulation;

[0050] The spatial analysis capabilities of ArcGIS were used to combine and segment the river network, clarifying the topological relationships between the main stream and tributaries. Gradient parameters for each river segment were calculated using ArcPy scripts, generating a geometric attribute table of the river network, which provides a foundation for setting channel roughness and velocity distribution in two-dimensional hydrodynamic calculations.

[0051] The DEM was imported into HEC-RAS as the base file for the two-dimensional hydrodynamic simulation. The simulation area was divided according to the river network coverage. At the outer edge of the area, the flow inlets were marked according to the node locations of the river network. Figure 8 In subsequent processes, the flow time series results obtained from the one-dimensional hydrodynamic simulation based on SWMM in the previous step will be input into HEC-RAS;

[0052] Import the DEM into HEC-RAS software and generate a terrain mesh. Combine this with the river network layer to define the calculation area and divide the mesh into 30m*30m grids, maintaining the same resolution as the DEM. Set flow input ports at the model boundaries, corresponding to the outflow nodes of the SWMM model. Import the flow time-series data output from the SWMM into the model as time-varying boundary conditions to achieve dynamic coupling between rainfall generation and flood evolution.

[0053] By adjusting the computation time and step size of HEC-RAS, a two-dimensional hydrodynamic simulation was performed to obtain time-series results of the flood inundation area and depth caused by rainfall, as shown below. Figure 9 As shown;

[0054] Based on the area of ​​the study area and the accuracy of the computational grid, the time step and computation time of HEC-RAS are adjusted, the two-dimensional hydrodynamic solution is started, and the time series data such as water depth, flow velocity and inundation range at different time steps are output to generate dynamic results of the flood evolution process.

[0055] Repeat the above steps to simulate rainfall with different return periods for the region, obtain the corresponding inundation area, and define the flood risk level as follows: Figure 10 As shown, the classification of flood risk levels is based on marking areas with simulated rainfall events that occur once every 20 years as high-risk areas, once every 100 years as medium-risk areas, and once every 300 years as high-risk areas.

[0056] The above process is repeated using rainfall intensity curves with different return periods to obtain the maximum inundation area and water depth distribution under each scenario. Based on the water depth threshold and inundation area percentage from the simulation results, the regional flood risk level is determined. This method is applicable to simulating flooding caused by rainfall over large areas. It can directly perform one-dimensional and two-dimensional hydrodynamic simulations based on DEM and corresponding rainfall time-series data even in the absence of urban pipe networks, realizing the entire process calculation from rainfall input to flood risk assessment, and providing quantitative basis for flood control planning and emergency decision-making.

Claims

1. A method for simulating and assessing urban flooding safety, characterized in that, Includes the following steps: S1: Collect DEM data of the area to be processed, and place the DEM data in ArcGIS software for projection and depression filling preprocessing; S2: Based on the DEM tif file, quickly calculate the flow direction and flow rate in ArcGIS software, extract the river network and construct the catchment area, and realize the construction of the rainfall and flood model; S3: Based on SWMM software, rainfall simulation is performed on the rainfall and flood model, and one-dimensional hydrodynamic calculations are carried out to obtain the time-series flow of each river channel and node; S4: Using the time-series flow data calculated by SWMM software as input, perform two-dimensional hydrodynamic simulation on the corresponding terrain based on HEC-RAS software to obtain the final inundation range and depth of the flood caused by extreme rainfall; S5: Process the flooded areas under different return periods of rainfall in step S4, overlay the output images, determine whether the area is prone to flooding, and then delineate flood risk zones.

2. The method for urban flood simulation and safety assessment according to claim 1, characterized in that, In step S1, after the DEM data is projected using ArcGIS software, the method also includes filling depressions in the terrain data to eliminate false depressions and ensure the continuity of surface water flow paths, thereby improving the accuracy of subsequent flow direction and confluence calculations.

3. The urban flood simulation and safety assessment method according to claim 1, characterized in that, In step S2, based on the extracted river network layer, the river is classified by analyzing the river length, catchment area and topological relationship to distinguish the main river channel from the tributary system; and the DEM file after depression filling is converted into a vector format shapefile and spatially connected with the river network nodes to generate a basic model file for rainfall-runoff simulation.

4. The urban flood simulation and safety assessment method according to claim 1, characterized in that, In step S3, the rainfall-runoff simulation uses SWMM software for one-dimensional hydrodynamic calculations. The model sets the drainage outlet, rainfall sequence, simulation duration, and calculation step size. The model is automatically run and batch results are extracted through the interface between Python software and SWMM software to obtain the flow time series data of the river channel and nodes.

5. The urban flood simulation and safety assessment method according to claim 1, characterized in that, In step S4, the two-dimensional hydrodynamic simulation uses HEC-RAS software, with DEM as the calculation basis. The simulation area is delineated according to the river network distribution, and a flow input port is set at the boundary node. The flow time series data obtained by SWMM software simulation is used as the input boundary condition to perform two-dimensional flood evolution simulation and output the water depth and inundation range results at different time steps.

6. The urban flood simulation and safety assessment method according to claim 1, characterized in that, In step S5, steps S3-S4 are repeated for rainfall conditions with different return periods to obtain the flood inundation range and water depth distribution under each scenario. Based on indicators such as water depth threshold, inundated area and duration, flood risk levels are classified to achieve regional flood risk zoning assessment and level determination.

Images