A hydraulic consistency surface data construction method for urban rainstorm waterlogging simulation

Abstract

This invention discloses a method for constructing hydraulically consistent surface data for urban stormwater flooding simulation. Its key feature is that it uses a high-precision satellite stereo mapping (DSM) of the city as a foundation. Through DSM preprocessing, topographic filtering, and continuous hydraulic datum reconstruction, combined with road vectors, water bodies, watercourses, building outlines, and land cover data, the original visible surface is hydraulically consistent, constructing hydraulically effective surface elevations and distributed surface parameters suitable for two-dimensional stormwater simulation. Compared with existing technologies, this invention can still recover major surface runoff channels and retain major barrier structures even in the absence of fine-grained pipe networks, forming a surface parameter raster suitable for two-dimensional urban flooding simulation. Furthermore, it can recover major urban surface runoff channels and retain major barrier structures such as buildings even when fine-grained pipe network data is missing or unshareable, providing replicable, scalable, and physically interpretable foundational data for two-dimensional urban flooding simulation.

Description

Technical Field

[0001] This invention relates to the field of remote sensing terrain modeling and geographic information processing technology, specifically to a method for constructing hydraulically consistent surface data for urban stormwater flooding simulation based on satellite stereo mapping (DSM), road, water body, building, and land cover data. Background Technology

[0002] Urban flooding is jointly controlled by rainfall intensity, topographic slope, surface cover, drainage channels, and urban micro-topography. In densely built-up areas, shallow surface runoff is influenced not only by macro-topographic gradients but also by meter-level structures such as roads, rivers, bridges, elevated roads, buildings, local low-lying areas, and green spaces. These structures determine whether rainwater is transported, blocked, or temporarily stored; therefore, urban flooding models need to be able to represent hydraulically effective surface morphology.

[0003] Satellite stereo mapping can generate meter-level DSMs over a large area, but the original DSM represents the visible urban surface, not the hydraulically effective shape required for two-dimensional surface runoff calculations. Artifacts such as water reflection, weak texture, stereo matching errors, vehicles, tree canopies, bridges, overpasses, and building edges can all create abnormal elevations or depressions, resulting in false water blockages, false depressions, broken drainage paths, and distorted water accumulation areas. Furthermore, many cities suffer from problems such as the inability to share detailed drainage network data, frequent updates, incomplete spatial coverage, or fragmented management departments, directly hindering the widespread adoption of conventional network-coupled urban flooding models.

[0004] In summary, existing terrain editing methods rely on manual canal construction, road smoothing, or localized obstacle repair, resulting in high subjectivity and low standardization, making them unsuitable for replicable urban-scale modeling. Therefore, it is necessary to propose a method for constructing hydraulically consistent surface data based on high-resolution DSM and commonly used urban geographic data. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by providing a method for constructing hydraulically consistent surface data for urban stormwater flooding simulation. Based on a high-precision satellite stereo mapping urban datum (DSM), and combined with road vectors, water surface, river system lines, building outlines, and land cover data, the original visible surface undergoes hydraulic consistency processing to construct hydraulically effective surface elevations and distributed surface parameters suitable for two-dimensional stormwater simulation. This method achieves automated or semi-automated construction of basic data for two-dimensional urban flooding simulation through satellite stereo DSM preprocessing, continuous hydraulic datum reconstruction, statistical replacement of road and river segment elevations, priority consistency processing of confluence water bodies, constraint filling, building barrier restoration, and distributed surface parameter construction. Even in the absence of a fine-grained pipe network, this invention can still restore major surface runoff channels, retain major barrier structures, and form a surface parameter raster suitable for two-dimensional urban flooding simulation. Furthermore, even with missing or unshareable fine-grained pipe network data, it can restore major urban surface runoff channels and retain major barrier structures such as buildings, providing replicable, scalable, and physically interpretable basic data for two-dimensional urban flooding simulation.

[0006] The specific technical solution to achieve the purpose of this invention is: a method for constructing hydraulically consistent surface data for urban stormwater flooding simulation, characterized by the following steps:

[0007] Step 1: Acquire satellite stereo imagery, road vectors, water surface, water system lines, building outlines, building heights, and land cover data, and unify them to the same plane coordinate system, vertical reference, and raster resolution.

[0008] Step 2: Generate the original DSM from satellite stereo imagery, and perform quality control on the original DSM to remove strip artifacts, isolated elevation spikes, invalid values, and local holes. Specifically, this includes: identifying and removing isolated elevation spikes based on local window statistics, smoothing and weakening strip noise, identifying invalid pixels based on effective value masks, and repairing local holes using neighborhood constraint interpolation, morphological closing operations, and multi-scale smoothing methods, thereby obtaining the preprocessed DSM.

[0009] Step 3: Perform terrain filtering on the DSM to separate ground and non-ground components, and reconnect fragmented ground pixels according to local void density, void size and spatial configuration to form a continuous hydraulic base.

[0010] Step 4: Rasterize the road vectors, water surfaces, water system lines, and building outlines into a DSM mesh to form road masks, water masks, and building masks.

[0011] Step 5: Divide long roads and waterways into local processing sections based on intersections, confluence points, geometric abrupt change points, or preset lengths.

[0012] Step 6: Extract effective DSM pixels from the road mask within each road segment, and use the low envelope elevation within the segment to represent the hydraulically effective road corridor.

[0013] Step 7: Extract effective DSM pixels from the water mask within each river segment, and use the median elevation within the segment to represent the effective hydraulic elevation of the water body.

[0014] Step 8: At the point where the road mask and the water mask overlap, restore water connectivity by prioritizing the water elevation over the road elevation.

[0015] Step 9: Constrain and fill the remaining non-physical depressions, and restore the main urban barrier structures using building outlines and heights.

[0016] Step 10: Construct a distributed surface parameter raster based on land cover, road, building and water body categories, and output a hydraulically consistent DSM and model parameter raster.

[0017] The satellite stereo imagery consists of GF-7 forward-looking and backward-looking images and their RPC data. The forward-looking imagery has a spatial resolution of 0.6 m, and the backward-looking imagery has a spatial resolution of 0.8 m. A 1 m resolution DSM is generated through dense stereo matching and forward intersection based on the RPC model.

[0018] The plane coordinate system and vertical reference adopt the mapping reference of the target city or target area.

[0019] The water mask consists of an OSM water surface and an OSM water system line, wherein the water system line generates a buffer zone based on the water level or available channel width attribute, and merges with the water surface to form a continuous water mask.

[0020] The low envelope elevation within the segment is calculated using the following formula:

[0021] z road_i =min{z(x)|x∈M road_i, And z(x) is the effective DSM elevation}.

[0022] Among them, M road_i Let z(x) be the road mask for the i-th road segment; z(x) is the DSM elevation of raster pixel x.

[0023] The median elevation within the segment is calculated using the following formula:

[0024] z river_j =median{z(x)|x∈M river_j , and z(x) is the effective DSM elevation}.

[0025] Among them, M river_j It serves as a water cover for the j-th river segment.

[0026] The water body elevation adopts the rule of covering road elevation for areas where roads intersect with rivers, roads cross rivers, and bridges and viaducts cover rivers, so that the river corridor in the two-dimensional model is not misinterpreted as a complete water-blocking structure by the visible bridge surface or viaduct surface.

[0027] The constrained depression filling method is a priority flood depression filling or equivalent connectivity correction method that limits the maximum depression filling depth. It is used to remove artificial non-physical depressions while retaining real local depression storage space that may have hydrological significance.

[0028] The parameters of the distributed surface include: Manning roughness coefficient, impermeability, maximum deficit, initial deficit, and constant loss rate. These parameters are assigned by land cover category, and the original built-up area is further reclassified into building category and road category through building outline and road vector.

[0029] The parameters of the distributed surface adopt the Deficiency and Constant infiltration scheme, in which the impermeability of building categories is 100%, and the maximum deficit, initial deficit, and constant loss rate are 0; the impermeability, maximum deficit, initial deficit, and constant loss rate of roads and other categories are determined according to the road paving materials, road grades, green belt configuration, permeable paving ratio, drainage facility density, and historical water accumulation calibration results of the target city, so as to adapt to the differences in the surface infiltration capacity of roads in different cities.

[0030] The hydraulically consistent DSM and model parameter grid are used as inputs to two-dimensional stormwater models, two-dimensional shallow water equation models, or Rain-on-Grid models for urban stormwater flooding simulation, historical event review, risk assessment, or subsequent deep learning flood prediction model training.

[0031] Compared with the prior art, the present invention has the following beneficial technical effects and significant technical progress:

[0032] 1) Without explicitly relying on detailed drainage network data, high-precision DSM and available urban geographic data are used to reconstruct the main surface runoff paths, reducing the data dependence of urban flooding modeling.

[0033] 2) By adopting median, low envelope, and water priority rules for water bodies, roads, and intersections respectively, non-hydraulic effective elevations caused by water surface reflection, vehicles, tree canopies, bridges, overpasses, and stereo matching noise can be effectively reduced.

[0034] 3) By restoring the barriers of buildings and filling depressions with constraints, the surface connectivity is improved, while the real urban barrier structure and local water storage space are preserved, thereby enhancing the physical rationality of the two-dimensional urban flooding simulation.

[0035] 4) By constructing a distributed parametric grid through land cover, road and building reclassification, it can simultaneously express the differences in roughness, impermeability and infiltration loss of different urban surfaces such as roads, buildings, water bodies and green spaces.

[0036] 5) This method has strong standardization and reproducibility, and can be used for historical rainstorm flooding review, urban flooding risk assessment, scenario simulation, and supervised sample generation for deep learning flood prediction models. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the overall process of the present invention;

[0038] Figure 2 Flowchart for DSM pretreatment and continuous hydraulic datum reconstruction;

[0039] Figure 3 This is a schematic diagram of DSM pretreatment and continuous hydraulic datum reconstruction.

[0040] Figure 4 A schematic diagram illustrating the construction of a water body mask and the replacement of the median elevation of the river channel;

[0041] Figure 5 Flowchart for road segmentation and low-envelope elevation extraction;

[0042] Figure 6 This is a schematic diagram of road segmentation and low-envelope elevation extraction;

[0043] Figure 7 A schematic diagram of priority treatment for road-water body overlapping areas and the framework of urban hydraulic connectivity;

[0044] Figure 8 A schematic diagram for constructing distributed surface parameters. Detailed Implementation

[0045] This invention includes the following steps:

[0046] (1) Acquisition and unification of basic data

[0047] 1-1: Acquire GF-7 forward and rearward stereo images, RPC data, road vectors, water surfaces, water system lines, building outlines, building heights, and land cover data. The spatial resolution of the GF-7 forward image is 0.6 m, and the spatial resolution of the rearward image is 0.8 m. After dense stereo matching and forward intersection with RPC, a 1 m DSM is generated.

[0048] 1-2: Unify the basic data collected above to the local coordinate system and elevation datum of the corresponding city.

[0049] (2) DSM pretreatment and hydraulic foundation reconstruction

[0050] 2-1: Perform strip artifact removal, isolated elevation spike removal, invalid value removal, and local hole removal on the original DSM;

[0051] 2-2: Terrain filtering is used to separate ground and non-ground components;

[0052] 2-3: To address the problem of fragmented ground pixels in densely built-up areas, fragmented ground pixels are reconnected based on local void density, void size, and spatial structure, and small discontinuities are filled to obtain a continuous hydraulic base.

[0053] (3) Construction of water and road cover

[0054] 3-1: Project and rasterize the OSM water surface, OSM water system line, OSM road vector and building outline into a 1m DSM grid;

[0055] 3-2: The water system line constructs a buffer zone based on the river channel grade or width attribute, and merges it with the water surface to form a water body mask;

[0056] 3-3: Road vectors construct road masks based on road network topology.

[0057] (4) Sections of roads and waterways

[0058] Long roads and rivers are divided into local processing segments at road intersections, river confluence points, major geometric abrupt changes, and excessively long continuous elements. By segmenting these segments, the application of a single elevation statistical value to roads or rivers with strong internal heterogeneity is avoided, thereby suppressing artifacts while preserving elevation changes along the route.

[0059] (5) Low envelope elevation replacement of road sections

[0060] For each road segment, the effective DSM pixels in its road mask are extracted, and the lowest effective representative elevation within the segment is calculated. This is used as the hydraulically effective surface elevation of the road segment. This processing is used to reduce artifacts caused by vehicles, tree canopies, bridge decks, overpasses, building edges, and artificial obstacles caused by the misalignment of the road vector with the visible road surface plane, thus restoring the road's function as a shallow surface runoff corridor.

[0061] (6) Replacement of median elevation of river section

[0062] For each river segment, the effective DSM pixels in its water mask are extracted, the median elevation within the segment is calculated, and the value is assigned to the corresponding water pixel. The statistics of the median elevation are used to suppress bidirectional outliers caused by water surface reflection, weak texture, local occlusion, and stereo matching failure, so that the river, ditch, and water corridor remain continuous in the two-dimensional model.

[0063] (7) Consistency treatment of road-water confluence area

[0064] In areas where road and water body masks overlap, a water body elevation replacement rule is adopted that takes precedence over road elevation, so that the river elevation can cover the road elevation. This rule is used to prevent bridges, viaducts, or road surfaces from being misidentified as complete water barriers by the two-dimensional model, thereby restoring the connectivity of water bodies under roads or bridges.

[0065] (8) Restoration of depressions and building barriers

[0066] 8-1: After road and river elevation processing, prioritize flood filling or equivalent connectivity correction with limited maximum filling depth is applied to the remaining artificial non-physical depressions to improve water flow connectivity without deleting real depression storage space.

[0067] 8-2: Rasterize the building outline to a 1 m grid and use the building height to restore the blocking effect of the main building.

[0068] (9) Distributed surface parameter construction

[0069] 9-1: Based on Dynamic World land cover data, and combined with road vectors and building outlines, classify the road and building categories of the original built-up areas;

[0070] 9-2: Assign parameters such as Manning roughness coefficient, impermeability, maximum deficit, initial deficit, and constant loss rate according to land cover category to form the parameter grid required for two-dimensional model calculation.

[0071] (10) Output of basic model data

[0072] Output hydraulically consistent DSM, building barrier grid, water and road constraint grid, and distributed surface parameter grid for input to a 2D Rain-on-Grid urban flooding model, enabling automated or semi-automated construction of basic data for 2D urban flooding simulation.

[0073] The embodiments of the present invention are described in detail below. This embodiment uses 1 m DSM and OSM road and water system data, building outlines and heights, and Dynamic World land cover data generated from GF-7 stereo imagery as a basis to construct hydraulically consistent surface data for a two-dimensional urban flooding model. Unless otherwise specified, the software, hardware, and geographic information processing steps in the embodiments are conventional methods in the art. This embodiment is only used to explain the present invention and should not be construed as limiting the scope of protection of the present invention.

[0074] Example 1

[0075] The implementation data and parameters are detailed in Table 1 below:

[0076] Table 1. Main Input Data (Information not provided or not required will be processed by default)

[0077]

[0078] Among them, OSM refers to OpenStreetMap, an open, crowdsourced geographic information data source that provides vector geographic features such as roads, waterways, buildings, and points of interest. CMAB refers to the China Multi-Attribute Building Dataset, a multi-attribute building data product for urban buildings in China, providing building outlines, heights, and related building attribute information. Dynamic World refers to a 10-meter resolution global land use / land cover data product generated based on Sentinel-2 imagery, providing land cover categories such as water bodies, trees, grasslands, farmland, built-up areas, and bare land, along with their probability information. GF-7 refers to the Gaofen-7 satellite, an optical stereo mapping satellite in China's high-resolution Earth observation system, possessing forward and backward stereo imaging capabilities, which can be used to acquire high-resolution stereo imagery and generate digital surface models.

[0079] The land cover category parameters are detailed in Table 2 below:

[0080] Table 2 Land Cover Category Parameters

[0081]

[0082] When the land cover category parameters in Table 2 above are applied in other regions, the parameter values ​​can be adjusted based on measured data or local experience without changing the procedure of this method.

[0083] See Figure 1 A method for constructing hydraulically consistent surface data for urban stormwater flooding simulation is described, with the following specific steps:

[0084] Step 1: DSM generation and coordinate unification

[0085] 1-1: Acquire forward and rearward stereo images and RPC data from the GF-7, perform dense stereo matching, and use the RPC sensor model for forward intersection.

[0086] 1-2: Rasterize the output point cloud into a 1 m DSM. Unify the DSM, road, water body, building, and land cover data to the same spatial reference and elevation datum; DSM: Digital Surface Model, used to represent the elevation information of the land surface and its overlying features, including the top surface elevation of buildings, vegetation, road facilities, and other surface structures.

[0087] Step 2: DSM Preprocessing and Base Surface Reconstruction

[0088] See Figure 2 DSM pretreatment and continuous hydraulic datum reconstruction specifically include:

[0089] 2-1: Perform local elevation filtering on the original DSM to remove strip artifacts and isolated spikes;

[0090] 2-2: Use SAGA or equivalent terrain filtering methods to separate ground and non-ground components; SAGA: System for Automated Geoscientific Analyses, is an open-source geographic information processing and geoscientific analysis software platform that provides functions such as terrain analysis, raster processing, spatial interpolation, geomorphic parameter calculation and terrain filtering;

[0091] 2-3: Reconnect fragmented ground pixels in densely built-up areas and fill small holes to form a continuous hydraulic base.

[0092] See Figure 3 The local spikes, non-surface uplifts, and fragmented ground in the original DSM are weakened or reconnected into a continuous hydraulic datum to express the hydraulically effective shape required for two-dimensional surface runoff calculation.

[0093] Step 3: Generation of Road, Water, and Building Masks

[0094] 3-1: Rasterize road vectors, water surfaces, water system lines, and building outlines to a 1 m DSM mesh;

[0095] 3-2: Construct a buffer zone for the water system line and merge it with the water surface to obtain a continuous water mask;

[0096] 3-3: Generate a road mask based on the road vector;

[0097] 3-4: Generate a building blocking mask based on the building outline.

[0098] Step 4: Segmenting roads and waterways

[0099] See Figure 4 A water body mask is constructed based on the water surface, water system line and river buffer zone, and the river is divided into local segments according to the river segment to provide processing units for subsequent median elevation replacement within the segment.

[0100] See Figure 5 A road mask is constructed based on road vectors, and the road is segmented locally according to road intersections, geometric changes and excessively long continuous road segments, providing processing units for the subsequent low envelope elevation extraction of road segments.

[0101] Step 5: Water body elevation correction

[0102] See Figure 4For each river or water segment, effective DSM pixels are extracted, the median elevation within the segment is calculated and assigned to the water cell of that segment, suppressing abnormal elevations caused by water surface reflection, weak texture, local occlusion and stereo matching failure, so that the river, ditch and water corridor remain continuous in the two-dimensional model.

[0103] Step 6: Road elevation correction

[0104] See Figure 5 The effective road elevation is extracted for each road segment, and the processed road elevation is assigned to the corresponding road cell to restore the role of the road as a shallow surface runoff corridor.

[0105] See Figure 6 By using low envelope elevation or minimum representative elevation within the segment, the local uplift caused by vehicles, tree canopies, overpasses and bridge decks is weakened, thus avoiding the formation of false water obstruction in the two-dimensional model by the aforementioned non-hydraulic effective elevations.

[0106] Step 7: Consistency processing of intersection regions and construction of connectivity skeleton

[0107] See Figure 7 In areas where road and water body masks overlap, a water body elevation replacement rule is adopted that takes precedence over road elevation, so that the elevation of the river or ditch covers the elevation of the road, viaduct or bridge deck, thereby weakening the non-hydraulic effective obstacles caused by bridge deck obstruction and road crossing of the river; at the same time, water body channels, road channels and building barriers are integrated to construct the urban hydraulic connectivity framework.

[0108] Step 8: Restoring the confined depressions and the building barrier

[0109] Connectivity corrections with maximum filling depth limited are applied to residual non-physical depressions. The main urban barrier structures are restored using building outlines and heights to prevent excessive smoothing from weakening the obstruction effect of buildings on surface runoff.

[0110] Step 9: Distributed parameter construction

[0111] See Figure 8 The land cover / land use data is spatially aligned, resampled, and reclassified with road vectors, building outlines, and water surface / water system data. Based on the reclassification results, a parameter mapping relationship is established, and each grid is assigned distributed surface parameters such as Manning roughness coefficient, impermeability, maximum deficit, initial deficit, and constant infiltration rate to form the parameter grid required for the calculation of the two-dimensional stormwater model.

[0112] Step 10: Result Output and Application

[0113] Output hydraulically consistent DSM, urban hydraulic connectivity framework, building barrier grid, and distributed surface parameter grid, which can be used by the 2D Rain-on-Grid urban flooding model for historical rainstorm simulation, scenario analysis, or deep learning proxy model sample generation.

[0114] The above embodiments are merely illustrative of the present invention and are not intended to limit the scope of the patent. All equivalent substitutions, modifications, or variations made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for constructing hydraulically consistent surface data for urban stormwater flooding simulation, characterized in that, The method includes the following steps: Step 1: Acquire satellite stereo imagery, road vectors, water surface, water system lines, building outlines, building heights, and land cover data, and unify them to the same plane coordinate system, vertical reference, and raster resolution; Step 2: Generate the original DSM from the satellite stereo imagery, and process the original DSM for strip artifacts, isolated elevation spikes, invalid values, and local holes; Step 3: Perform terrain filtering on the DSM to separate ground and non-ground components, and reconnect fragmented ground pixels according to local void density, void size and spatial configuration to form a continuous hydraulic datum. Step 4: Rasterize the road vectors, water surfaces, water system lines, and building outlines into a DSM mesh to form road masks, water masks, and building masks; Step 5: Divide long roads and waterways into local processing sections based on intersections, confluence points, geometric abrupt change points, or preset lengths; Step 6: Extract effective DSM pixels from the road mask within each road segment, and use the low envelope elevation within the segment to represent the hydraulically effective road corridor; Step 7: Extract effective DSM pixels from the water mask within each river segment, and use the median elevation within the segment to represent the effective hydraulic elevation of the water body; Step 8: At the point where the road mask and the water mask overlap, restore water connectivity by prioritizing the water elevation over the road elevation; Step 9: Fill the remaining non-physical depressions with constraints, and restore the urban barrier structure using the building outlines and heights; Step 10: Construct a distributed surface parameter raster based on land cover, road, building and water body categories, and output a hydraulically consistent DSM and model parameter raster.

2. The method for constructing hydraulically consistent surface data for urban stormwater flooding simulation according to claim 1, characterized in that, The satellite stereo imagery consists of forward-looking and backward-looking images and their RPC data, and a DSM is generated through dense stereo matching and forward intersection based on the RPC model.

3. The method for constructing hydraulically consistent surface data for urban stormwater flooding simulation according to claim 1, characterized in that, The plane coordinate system and vertical reference adopt the mapping reference of the target city or target area.

4. The method for constructing hydraulically consistent surface data for urban stormwater flooding simulation according to claim 1, characterized in that, The water mask consists of a vector water surface and a vector water system line, wherein the water system line generates a buffer zone based on the water level or river width attribute, and merges with the water surface to form a continuous water mask.

5. The method for constructing hydraulically consistent surface data for urban stormwater flooding simulation according to claim 1, characterized in that, The low envelope elevation within the segment is adopted using z road_i =min{z(x)|x∈M road_i, And z(x) is the effective DSM elevation calculated using the formula, where M road_i Let z(x) be the road mask for the i-th road segment, and z(x) be the DSM elevation of the raster cell x. The median elevation within the segment is represented by z. river_j =median{z(x)|x∈M river_j , and z(x) is the effective DSM elevation} calculated using the formula, where M river_j The water body is a water body mask for the j-th river segment; the water body elevation adopts the rule of covering the road elevation, and is used for the areas where the road intersects with the river, the road crosses the river, and the bridge and viaduct cover the river, so that the river corridor in the two-dimensional model is not misinterpreted as a complete water-blocking structure by the visible bridge surface or viaduct surface.

6. The method for constructing hydraulically consistent surface data for urban stormwater flooding simulation according to claim 1, characterized in that, The constrained depression filling method is a priority flood depression filling or equivalent connectivity correction method that limits the maximum depression filling depth. It is used to remove artificial non-physical depressions while retaining real local depression storage space with hydrological significance.

7. The method for constructing hydraulically consistent surface data for urban stormwater flooding simulation according to claim 1, characterized in that, The distributed surface parameter grid includes: Manning roughness coefficient, impermeability, maximum deficit, initial deficit, and constant loss rate. These parameters are assigned by land cover category, and the original built-up area is further reclassified into building category and road category through building outlines and road vectors.

8. The method for constructing hydraulically consistent surface data for urban stormwater flooding simulation according to claim 1 or claim 7, characterized in that, The distributed surface parameter grid is expressed using the Deficiency and Constant infiltration scheme. The impermeability of building categories is set to 100%, and the maximum deficit, initial deficit, and constant loss rate are set to 0. The impermeability, maximum deficit, initial deficit, and constant loss rate of roads and other categories are determined based on the road paving materials, road grades, green belt configuration, permeable paving ratio, drainage facility density, and historical water accumulation calibration results of the target city to accommodate the differences in surface infiltration capacity of roads in different cities.

9. The method for constructing hydraulically consistent surface data for urban stormwater flooding simulation according to claim 1, characterized in that, The hydraulically consistent DSM and model parameter grid are used as inputs to flood simulation models of two-dimensional stormwater models, two-dimensional shallow water equation models, or Rain-on-Grid models, for urban rainstorm flood simulation, historical event review, risk assessment, or subsequent deep learning flood prediction model training.

Images