Flood risk assessment method and system based on multi-dimensional data integration analysis

Abstract

The invention relates to the technical field of flood management, in particular to a flood risk assessment method and system based on multi-dimensional data integration analysis. The method comprises the following steps: acquiring a multi-dimensional data set including geographic information data and hydro meteorological data; determining a water flow direction by using the geographic information data, and identifying a catchment area based on the water flow direction; carrying out drainage basin runoff production calculation according to the hydro-meteorological data, and recording runoff volume; executing runoff collection simulation in the catchment area by utilizing runoff volume to form a runoff process line; calculating flood peak flow based on the runoff hydrograph; evaluating the river water level according to the flood peak flow, and calculating the river overflow according to the river water level; generating a submerging depth distribution map according to the river overflow quantity; and performing flood risk prediction based on the submerging depth distribution map, and generating a flood management risk level. Fine evaluation and dynamic prediction of the flood risk are realized based on the flood management technology, and the flood prediction accuracy and the flood control decision efficiency are improved.

Description

Technical Field

[0001] This invention relates to the field of flood management technology, and in particular to a flood risk assessment method and system based on multidimensional data integration analysis. Background Technology

[0002] Existing flood risk assessment and hydrological management technologies typically rely on a single type of data source for risk prediction, such as using only historical hydrological and meteorological records, river level monitoring, or traditional hydrodynamic models to simulate floods. While this approach can largely achieve flood risk prediction and flood control decision support in small watersheds or with simple river network structures, it struggles to accurately reflect flood evolution and potential risk distribution in complex watersheds or scenarios involving parallel analysis of multi-source information. This is due to the variable spatial distribution of rainfall, complex river network structures, and dynamic nonlinearity of hydrological processes. While existing studies can generate flood inundation prediction maps using topographic data or hydrodynamic models, they generally suffer from the following problems: First, they only present static or simplified water level distribution information, lacking comprehensive analysis of rainfall intensity, soil infiltration characteristics, and watershed runoff dynamics; second, they cannot achieve detailed simulation of catchment area delineation and runoff processes based on multi-source data, resulting in a lack of spatial resolution and temporal accuracy in risk assessment; third, current flood simulation and assessment processes largely rely on manual model construction and parameter calibration, which are costly and time-consuming, and lack the ability to automatically integrate, intelligently analyze, and visualize decision support based on multi-dimensional data. Summary of the Invention

[0003] Therefore, it is necessary for the present invention to provide a flood risk assessment method and system based on multidimensional data integration analysis to solve at least one of the above-mentioned technical problems.

[0004] To achieve the above objectives, a flood risk assessment method based on multidimensional data ensemble analysis includes the following steps: Step S1: Obtain a multidimensional dataset, including geographic information data and hydrological and meteorological data; use the geographic information data to determine the direction of water flow, and identify the catchment area based on the direction of water flow; Step S2: Calculate watershed runoff based on hydrological and meteorological data and record the runoff volume; use the runoff volume to perform runoff collection simulation in the catchment area to form a runoff process curve; Step S3: Calculate the peak flow based on the runoff hydrograph; assess the river level based on the peak flow, and calculate the river overflow based on the river level; Step S4: Generate an inundation depth distribution map based on the river overflow; predict flood risk based on the inundation depth distribution map and generate a flood management risk level.

[0005] Preferably, this specification also provides a flood risk assessment system based on multidimensional data integration analysis, used to perform the flood risk assessment method based on multidimensional data integration analysis as described above, wherein the flood risk assessment based on multidimensional data integration analysis includes: The catchment area identification module is used to acquire multidimensional datasets, including geographic information data and hydrological and meteorological data; it uses geographic information data to determine the direction of water flow and identifies the catchment area based on the direction of water flow. The runoff collection simulation module is used to calculate watershed runoff based on hydrological and meteorological data and record runoff volume; it also uses the runoff volume to perform runoff collection simulation in the catchment area to form a runoff process curve. The peak flow calculation module is used to calculate the peak flow based on the runoff process curve; assess the river level based on the peak flow; and calculate the river overflow based on the river level. The flood risk prediction module is used to generate an inundation depth distribution map based on the river overflow; to predict flood risk based on the inundation depth distribution map; and to generate a flood management risk level.

[0006] The beneficial effects of this invention are as follows: (1) By integrating and processing multidimensional data, including geographic information data, hydrological and meteorological data and river network structure information, combined with rasterization processing, confluence path analysis and elevation gradient calculation, a detailed characterization of the watershed spatial structure and catchment area was achieved, ensuring the accuracy and completeness of flood risk assessment input data.

[0007] (2) In the process of identifying the catchment area and simulating the runoff process, the method of water flow direction coding, downstream location fast lookup table and isochronous sub-region division is adopted to dynamically advance the confluence path and calculate the sub-region weight, thereby realizing the intelligent reconstruction of the watershed runoff generation and confluence process and the accurate representation of the time-space distribution, which improves the realism and operability of flood process simulation.

[0008] (3) In the process of calculating flood peak flow and assessing river level and overflow, flood peak flow and overflow information are generated based on runoff process line and river cross-sectional characteristics. Combined with time series data, the river overflow is dynamically mapped, realizing intuitive visualization of flood inundation depth and risk area, which facilitates flood control decision-making and risk management.

[0009] (4) In the process of flood risk prediction and dynamic optimization, the risk level and spatial distribution are adaptively adjusted by using river overflow and inundation depth data. The runoff contribution is updated by combining the peak delay degree and sub-region weight correction factor. This realizes intelligent optimization and dynamic expression of flood risk assessment, providing scientific basis and decision support for basin flood control management, emergency dispatch and disaster prevention and mitigation. Attached Figure Description

[0010] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a schematic diagram of the steps of a flood risk assessment method based on multidimensional data integration analysis according to the present invention; Figure 2 This is a schematic diagram of the flood risk assessment system based on multidimensional data integration analysis according to the present invention; Figure 3 This is a schematic diagram of the flooding depth distribution map in this invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0011] The technical method of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this invention.

[0012] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0013] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0014] To achieve the above objectives, please refer to Figures 1 to 3 This invention provides a flood risk assessment method based on multidimensional data integration analysis, the method comprising the following steps: Step S1: Obtain a multidimensional dataset, including geographic information data and hydrological and meteorological data; use the geographic information data to determine the direction of water flow, and identify the catchment area based on the direction of water flow; In one embodiment, multi-source data is collected for the study watershed, including a high-resolution digital elevation model (DEM), land use data, river network information, and hydrological and meteorological data, such as rainfall intensity sequences and soil permeability distribution. First, the DEM data is rasterized, and the elevation value is extracted for each raster cell. The elevation difference between adjacent raster cells is calculated to determine the direction of the flow gradient. Based on the elevation gradient direction, the lowest descent path is generated, and the upstream raster positions flowing into the raster cells are counted, connecting them to form confluence channels. Subsequently, a flow direction coding field (using D8 discrete direction coding) is established based on the confluence channels, and a fast lookup table for downstream locations is constructed. Using the cumulative confluence count as the initial value, confluence path advancement is performed to form a complete confluence path network. Finally, the catchment areas are identified through the confluence path network, providing basic data and spatial reference for subsequent watershed runoff calculations and flood peak simulations.

[0015] In another embodiment, it is assumed that the total area of ​​the study watershed is approximately 120 km², the DEM grid size is 20 m × 20 m, and the land use types include forest, farmland, and urban areas, accounting for 40%, 35%, and 25%, respectively. The rainfall intensity sequence is assumed to be 1–50 mm / h, and the soil permeability distribution ranges from 0.2–0.5 mm / h. The flow direction coding field contains approximately 15,000 grids, of which approximately 1,500 grids have coding conflict anomalies, requiring local smoothing correction. Ultimately, five main catchment areas were identified, with a total of approximately 14,800 grids. The flow direction and runoff path network of each area are clearly labeled, providing a fine spatial reference for subsequent runoff generation and flood peak calculations.

[0016] Step S2: Calculate watershed runoff based on hydrological and meteorological data and record the runoff volume; use the runoff volume to perform runoff collection simulation in the catchment area to form a runoff process curve; In one embodiment, based on the identified catchment area boundaries, an area-weighted average is applied to the rainfall intensity sequence to generate the time-varying areal rainfall for each sub-region. A dynamic infiltration threshold is set based on soil permeability distribution; when the time-varying areal rainfall exceeds the infiltration threshold, watershed runoff calculation is performed to record the runoff volume for each grid. Then, the catchment area is divided into multiple sub-regions according to isochrones, the area proportion of each sub-region is calculated, a sub-region weight sequence is generated, and the runoff volume is linearly allocated according to the sub-region weights to obtain the runoff contribution value. The runoff volumes of each sub-region at different time steps are sequentially concatenated to form a complete runoff process curve, providing input for peak flow calculation and accurately reflecting the contribution of different sub-regions to the total runoff volume.

[0017] In another embodiment, the catchment area is assumed to be divided into 10 isochronous sub-regions with area proportions of 0.08, 0.12, 0.10, 0.11, 0.09, 0.12, 0.10, 0.11, 0.09, and 0.08, respectively. The assumed areal rainfall (mm / h) for each time period are 3.5, 4.2, 3.8, 4.0, 3.6, 4.1, 3.9, 4.3, 3.7, and 4.0, respectively. The infiltration threshold is set to 0.35 mm / h, and runoff generation is calculated for sub-regions exceeding this threshold. The runoff contribution values ​​(m³ / s) after linear allocation are 0.85, 1.02, 0.95, 1.0, 0.90, 1.03, 0.97, 1.05, 0.92, and 1.0, respectively, with a maximum total runoff of approximately 10.9 m³ / s. This simulation can clearly show the runoff contribution and time series changes of each sub-region, providing a quantitative basis for subsequent flood peak analysis.

[0018] Step S3: Calculate the peak flow based on the runoff hydrograph; assess the river level based on the peak flow, and calculate the river overflow based on the river level; In one embodiment, peak flow is calculated for each cross-section of the river using the formed runoff process curve. Combining river cross-sectional parameters, including river width, slope, and Manning roughness coefficient, the river level is calculated using the Manning formula or a numerical hydraulic model. The overflow is then assessed based on whether the water level exceeds the design height of the river. This calculation process can simultaneously output peak flow time series, river level change curves, and overflow length, providing key parameter inputs for inundation depth distribution maps and flood risk prediction. Furthermore, it can simulate river responses for different design flow rates or extreme rainfall events, assisting in flood control scheduling and emergency decision-making.

[0019] In another embodiment, the average cross-sectional width of the river channel is assumed to be 10m, the hydraulic gradient to be 0.002, and the Manning roughness coefficient to be 0.035. The maximum runoff volume is 10.9 m³ / s, corresponding to a peak flood level of approximately 2.8m. With a design river level of 2.5m, the overflow is estimated at approximately 0.4 m³ / s, and the overflow length is approximately 150m. The calculation time step is 1 hour, and the peak flood is assumed to occur 6 hours after the start of rainfall. The river overflow is mainly concentrated in two downstream cross-sections, with overflow widths between 10 and 15m. These values ​​provide a clear quantitative basis for flood peak and overflow characteristics, enabling inundation simulation and risk level classification.

[0020] Of particular importance, step S3, which involves calculating the peak flow based on the runoff hydrograph, includes: Read the runoff flow sequence over time periods from the runoff process curve and arrange them in ascending order to form ordered flow points; use the ordered flow points to compare the runoff flow in adjacent time periods and record the local maxima. In one embodiment, the time-period runoff sequence of the catchment area is read, and the runoff volumes of each time period are arranged in ascending order to form an ordered flow point sequence. Subsequently, using this ordered flow point sequence, the runoff volumes of adjacent time periods are compared in chronological order. Flow points in adjacent time periods that are greater than those in the preceding and following time periods are recorded as local maxima, thus identifying potential peak flow locations. This information can be used for subsequent peak flow extraction and flood response analysis.

[0021] In another embodiment, assume the following 12 consecutive time periods of runoff (in m³ / s) are read: [0.42, 0.51, 0.65, 0.62, 0.78, 0.74, 0.91, 0.88, 0.67, 0.59, 0.48, 0.45]. After sorting in ascending order, comparing adjacent time periods reveals local maxima at times 3, 5, and 7, corresponding to flows of 0.65, 0.78, and 0.91 m³ / s, respectively. These local maxima are potential flood peak locations and are marked for subsequent flood peak analysis.

[0022] Extract the peak flow rate from the local maxima; determine the peak time based on the extracted peak flow rate; perform non-maximum suppression based on the time period runoff sequence, and determine the peak flow rate.

[0023] In one embodiment, based on the extracted local maxima, the flow rate corresponding to each local maxima is selected as the initial flow peak, and its time position is recorded to obtain the peak time series. Subsequently, non-maximum suppression (NMS) is performed on the time-period flow series to remove non-peak interference points, and finally the flood peak flow rate and its corresponding time are determined for flood simulation, flood peak scheduling, and risk assessment.

[0024] In another embodiment, assume the local maximum flow rate is [0.65, 0.78, 0.91] m³ / s, corresponding to times [t3, t5, t7]. Through non-maximum suppression, values ​​of 0.65 m³ / s and 0.78 m³ / s are removed, retaining the highest value of 0.91 m³ / s, corresponding to time t7. The final peak flow rate is determined to be 0.91 m³ / s, and the peak occurs during the 7th time period. This result can be directly used in subsequent basin flood evolution simulations or flood control scheduling decisions.

[0025] Step S4: Generate an inundation depth distribution map based on the river overflow; predict flood risk based on the inundation depth distribution map and generate a flood management risk level.

[0026] In one embodiment, the overflow area is rasterized based on the calculated river overflow, and flood inundation simulation is performed using DEM data to generate an inundation depth distribution map. Flood risk levels (low, medium, and high) are classified according to inundation depth and regional importance, and a flood risk level map is output. This map provides an intuitive basis for flood control management, emergency dispatch, and public early warning. It can also be overlaid with a Geographic Information System (GIS) for analysis, enabling spatial risk distribution visualization and supporting decision-making departments in formulating regional flood control measures and emergency plans.

[0027] In another embodiment, it is assumed that the total number of grid cells in the overflow area is 1,500, of which approximately 800 cells are inundation depths of 0.1–0.3 m, approximately 500 cells are inundation depths of 0.3–0.6 m, and approximately 200 cells are inundation depths of >0.6 m. Risk levels are categorized by inundation depth: low risk accounts for 53%, medium risk for 33%, and high risk for 14%. Simulation results show that high-risk areas are mainly distributed in the downstream of the river and low-lying farmland, medium-risk areas cover the urban fringe and along roads, and low-risk areas are mainly in high-altitude and hilly areas. The resulting flood risk level distribution map can provide data support for flood control command, emergency plans, river patrols, and public safety alerts.

[0028] Of particular importance, step S4, which generates an inundation depth distribution map based on the river overflow, includes: An overflow diffusion area is generated based on the river overflow; water level is estimated based on the overflow diffusion area, and a dynamic water depth map is generated; the ground elevation is calculated based on the dynamic water depth map, and the inundation depth is calculated based on the ground elevation. In one embodiment, an overflow diffusion area is first generated based on the river overflow volume. The diffusion process of the overflow on the riverbank and surrounding terrain is simulated using a fluid dynamics model or grid cells to obtain water level information for each grid cell. Subsequently, based on the water level distribution in the overflow diffusion area, water level is estimated for each grid cell to generate a dynamic water depth map. Furthermore, geographic information data (such as DEM elevation data) is used to calculate the ground elevation of each grid cell, and the dynamic water depth is combined with the ground elevation to calculate the inundation depth value, providing basic data for flood inundation risk assessment.

[0029] In another embodiment, assuming a 10×10m grid is selected in a certain river section, corresponding to a river overflow of [5.2, 6.1, 7.3, 6.8, 5.9] m³ / s, a dynamic water depth map (in meters) is obtained through grid diffusion simulation: [[0.0, 0.1, 0.2, 0.3, 0.4], [0.0, 0.2, 0.3, 0.5, 0.6], [0.0, 0.3, 0.5, 0.6, 0.7], [0.1, 0.4, 0.6, 0.8, 0.9], [0.2, 0.5, 0.7, 0.9]. Combined with DEM ground elevation data (assumed to be 0.0–0.5m), the inundation depth matrix (unit: m) is obtained as follows: [[0.0,0.0,0.0,0.0,0.0],[0.0,0.0,0.0,0.2,0.3],[0.0,0.1,0.2,0.3,0.4],[0.0,0.2,0.4,0.6,0.7],[0.0,0.3,0.5,0.7,0.8]]. This inundation depth matrix can be used for subsequent flood impact analysis or flood control emergency response.

[0030] Depth levels are set based on the inundation depth values, and a depth level distribution field is generated. Depth contour lines are extracted based on the depth level distribution field, and an inundation depth distribution map is generated based on the depth contour lines.

[0031] In one embodiment, based on the obtained inundation depth value, a preset depth level threshold (e.g., 0–0.2m, 0.2–0.5m, 0.5–1.0m) is used to divide each grid cell into different depth levels, generating a depth level distribution field. Subsequently, based on the depth level distribution field, depth contour lines for each level are extracted, and these contour lines are used to generate an inundation depth distribution map to visually display the inundation degree of different areas, providing a reference for flood control planning or evacuation route design.

[0032] In another embodiment, assuming the flooding depth value matrix is ​​as described above, and setting the depth level thresholds as 0–0.2m, 0.2–0.5m, 0.5–0.8m, and 0.8–1.0m, the distribution areas obtained by classifying by level are as follows: the grid cells for 0–0.2m include [0,1], [0,2], [1,1], [2,1], etc., totaling 8 cells; the grid cells for 0.2–0.5m include [1,3], [1,4], [2,3], [2,4], [3,2], [...]. There are a total of 9 grid units, including [3,3], [3,4], [4,2], and [4,3]. The 0.5–0.8m grid units include 3 units, including [3,4], [4,3], and [4,4]. The 0.8–1.0m grid unit is [4,4]. Based on the above distribution, a depth contour map is generated, and the 0.2m, 0.5m, and 0.8m depth lines are drawn on the map to intuitively show the inundation range and depth differences of each grid unit, providing visualization support for subsequent flood response.

[0033] It should be noted that you should refer to [link / reference]. Figure 3 The map uses different colors to distinguish water depth ranges of 0.01-0.15m, 0.15-0.30m, 0.30-0.50m, 0.50-1.00m, and >1.00m. It includes road networks, water areas, and regional boundaries, with a directional indicator in the upper right corner (N for north) and a scale representing 1000 meters in the ground at the bottom. It can intuitively show the differences in the inundation range and depth of different areas, providing data support for flood risk assessment, flood control planning, and emergency response.

[0034] Preferably, determining the water flow direction using geographic information data in step S1 includes: Generate raster cells using geographic information data; extract elevation values ​​from the raster cells; calculate the elevation difference between raster cells based on the elevation values; and mark the elevation gradient direction based on the elevation difference. In one embodiment, high-resolution digital elevation model (DEM) data of the study watershed and related geographic information data (such as land use type and river network information) are acquired. The DEM data is rasterized, each raster cell is assigned a unique index, and the elevation value corresponding to that cell is extracted and recorded. ,in , This indicates the row and column number of the grid. Then, for each grid cell, the elevation difference between it and its eight adjacent directional cells is calculated. The elevation change rate in each direction is obtained. Based on the elevation difference, the elevation gradient direction of each grid cell is marked using the D8 discrete direction encoding method, pointing to the direction of the neighboring cell with the fastest elevation decrease. The elevation gradient field of the entire grid cell can be used for subsequent minimum descent path analysis, providing basic spatial information for the identification of confluence channels.

[0035] In another embodiment, assuming the total area of ​​the watershed under study is approximately 50 km², and the grid size is 20m × 20m, a total of 62,500 grid cells are generated. Elevation values The range is assumed to be 50–250m, and the elevation difference between neighboring areas is... _ , The elevation gradient ranges from -15m to +20m. After calculating the elevation difference, approximately 62,000 cells clearly indicate the elevation gradient direction. The remaining cells require local smoothing due to the flat surrounding elevation. The resulting elevation gradient direction marking matrix visually displays the potential direction of water flow on the grid, providing precise spatial information for determining subsequent confluence paths and flow directions.

[0036] The lowest descent path is determined using the elevation gradient direction; along the lowest descent path, the grid positions of the inflowing grid cells are counted and connected to form a confluence channel; the flow direction is determined based on the confluence channel.

[0037] In one embodiment, the water flow path is traced grid-by-grid using the marked elevation gradient direction to determine the lowest descent path. For each grid cell, downstream cells are iteratively visited along the gradient direction from that cell until the channel outlet or watershed boundary is reached. After each iteration, the positions of upstream cells flowing into the current grid are counted to form a cumulative inflow count. Subsequently, consecutive inflow grids are connected to form a complete confluence channel, and the downstream direction of each grid within the confluence channel is taken as the water flow direction of that grid cell. The formation of the confluence channel takes into account topographic slope, catchment area, and watershed boundary conditions, so that each channel can reflect the optimal path of water flow under real topographic conditions, while providing basic information for sub-watershed division and catchment area identification.

[0038] In another embodiment, it is assumed that the watershed contains 62,500 grid cells, and the elevation gradient direction has been determined. Through iterative path tracing, approximately 3,200 minimum descent paths are generated, each with an average length of approximately 50–120 grid cells. The cumulative number of inflow grid cells ranges from 1 to 150, with the maximum cumulative value corresponding to the main river channel. The final determined confluence channels cover approximately 4,800 grid cells, with paths mainly concentrated in low-lying valleys and river valleys. The longest channel is approximately 120 grid cells, corresponding to an actual river channel length of approximately 2.4 km. In this way, the flow direction, inflow count, and confluence channel information for each grid cell are clearly quantified, providing spatial foundational data for subsequent watershed runoff calculations, flood peak simulations, and flood risk assessments.

[0039] Preferably, step S1, which identifies the catchment area based on the direction of water flow, includes: A water flow direction coding field is generated based on the water flow direction; a downstream location fast lookup table is constructed using the water flow direction coding field, and the cumulative flow count of the downstream location fast lookup table is initialized to 1; In one embodiment, a water flow direction coding field is generated based on watershed DEM data and rasterized geographic information, with each raster cell F(i,j) assigned its steepest descent direction. Subsequently, using this water flow direction coding field, a fast lookup table for downstream locations is constructed for each raster cell. This table records the indices of neighboring cells directly to which the water flow originates, and the cumulative confluence count in the table is initialized to 1 for subsequent cumulative calculations. The fast lookup table can be implemented using a matrix index or a hash table, ensuring that each raster cell can access its downstream cells in constant time, thereby improving the efficiency of large-scale watershed confluence path calculations. This step ensures the integrity of the confluence information in the initial stage, providing foundational data for subsequent iterative accumulation and catchment area identification.

[0040] In another embodiment, assuming a total watershed area of ​​approximately 50 km², a grid size of 20 m × 20 m, and a total of 62,500 grid cells are generated. After calculating the flow direction coding field using the D8 method, a downstream index table for each cell records the position of the steepest descending cell within its eight neighboring areas, with the initial cumulative confluence count set to 1. Assuming an average slope of 5° and a maximum slope of approximately 15° within the watershed, after constructing a fast downstream lookup table, approximately 62,000 cells can quickly find their downstream cells. The remaining 500 cells, located in flat areas, require local fine-tuning to ensure that the downstream direction is not lost. This data structure provides efficient and quantifiable initial conditions for confluence path advancement and catchment area identification.

[0041] The downstream location is used to quickly look up a table to determine the downstream pointing relationship; the confluence path is advanced according to the downstream pointing relationship to generate a confluence path network; and the catchment area is identified based on the confluence path network.

[0042] In one embodiment, a generated downstream location lookup table is used to perform confluence path advancement for each grid cell. This involves progressively accumulating the number of inflow cells along the downstream direction and recording the grid cells traversed by the path as confluence channels. Path advancement can employ an iterative algorithm, iteratively visiting the source grid cells downstream until reaching the outlet or boundary. Subsequently, based on the confluence path network of the entire watershed, the cumulative inflow count for each grid cell is counted. When the cumulative count reaches a preset threshold (e.g., 20 upstream grid cells flowing in), the connected region containing that grid cell is identified as a catchment area. The entire catchment area identification process can combine river network, topographic relief, and watershed boundary conditions to obtain a continuous catchment area surface for subsequent runoff calculations and flood simulations.

[0043] In another embodiment, it is assumed that the watershed contains 62,500 grid cells, with an initial cumulative runoff count of 1. Through iterative path advancement, approximately 3,200 runoff paths are generated, with an average path length of approximately 80 grid cells and a maximum cumulative inflow count of 150 cells, corresponding to the main river channel. Based on a threshold setting (cumulative inflow ≥ 20), approximately 4,800 grid cells are identified as catchment areas, forming approximately 15 main catchment sub-regions. The average area of ​​each catchment area is approximately 0.9 km², with a maximum area of ​​approximately 3 km². This runoff path network and catchment area delineation can provide accurate spatial basis data for subsequent hydrological simulation, runoff calculation, and flood risk assessment, supporting quantitative analysis and dynamic simulation.

[0044] Preferably, generating a water flow direction coding field based on the water flow direction includes: Read the angle value of the water flow direction and normalize it to a continuous interval of 0° to 360°; divide the angle value into D8 direction intervals; determine the discrete direction based on the D8 direction intervals; In one embodiment, the water flow direction angle of each grid cell is obtained using a digital elevation model (DEM) or hydrogeographic data. The angles, ranging from 0° to 360°, are normalized to ensure a smooth transition within the continuous range. These continuous angles are then divided into D8 directional intervals, corresponding to eight main directions (east, southeast, south, southwest, west, northwest, north, and northeast), each interval spanning approximately 45°. Based on this division, the angles are mapped to discrete direction markers for subsequent confluence path advancement and flow direction encoding. This step ensures that the flow direction data retains continuity while facilitating discretization, enabling rapid generation of the flow direction encoding field.

[0045] In another embodiment, it is assumed that there are 50,000 grid cells in the watershed, and the angle value of each grid cell is... After normalization, the data is distributed between 0° and 360°, with an average interval of approximately 0.007°. This is divided into eight D8 directional intervals, each 45° wide; for example, 0°–45° represents east, 45°–90° represents southeast, and so on. After discretization, each raster cell is assigned a discrete directional number (1–8). During this process, approximately 500 raster cells exhibit uncertain directions due to flat slopes or angles close to interval boundaries, requiring further marking as "points to be smoothed" to provide a basis for subsequent encoding corrections. This division ensures that the directional data for the entire watershed is both discretized and reflects local slope variations.

[0046] The power-law encoding values ​​are assigned according to discrete directions, and a water flow direction encoding field is generated based on the power-law encoding values. During the process of generating the water flow direction encoding field based on the power-law encoding values, encoding conflict anomalies are identified, and local smoothing correction is performed based on the encoding conflict anomalies.

[0047] In one embodiment, a power-law encoding value is assigned to each grid cell based on discrete direction markers (e.g., 2^0 = 1 for east, 2^1 = 2 for southeast, and so on), generating a flow direction encoding field. This encoding field can be used for rapid querying of downstream cells and cumulative confluences. Subsequently, during the generation process, each grid cell is checked for encoding conflict anomalies, such as inconsistencies between the flow direction encoding of the same grid cell and its surrounding neighbors, or the formation of a cyclic flow direction. For anomalies, a local smoothing correction is performed: the average encoding direction of the grid cell and its eight neighboring cells is calculated, and the grid cell's encoding value is updated to ensure the overall directional consistency and the physical plausibility of the flow. In this way, the generated encoding field ensures both rapid computational performance and hydrological plausibility.

[0048] In another embodiment, assuming a watershed contains 50,000 raster cells, after initial discrete direction power-law coding, the coded values ​​are distributed as [1, 2, 4, 8, 16, 32, 64, 128], averaging approximately 6,250 cells per direction. During the coding process, approximately 1,200 raster cells were found to have conflict anomalies due to gentle terrain or angles close to directional boundaries. Through local smoothing correction, the coded values ​​of these anomalies were adjusted to the neighborhood average, and the corrected flow direction was calculated using weights, with the maximum correction deviation not exceeding one direction (45°). After correction, the flow direction in the coded field is continuous and free of cyclical conflicts, ensuring the accuracy and stability of downstream path advancement and catchment area analysis, while providing reliable input data for subsequent dynamic flood simulation.

[0049] Preferably, in the process of generating the water flow direction coding field based on the power-order coding value, identifying coding conflict anomalies includes: In the process of generating the water flow direction coding field based on the power-law coding value, the coordinate offset is calculated using the power-law coding value; and the downstream target position is determined based on the coordinate offset. In one embodiment, in the generated water flow direction coding field, the power-law coding value of each grid cell is read (e.g., the coding value for east is 2^0=1, for southeast is 2^1=2, and so on). The corresponding two-dimensional coordinate offset is calculated using the correspondence between the coding value and the D8 direction. , Applying the offset to the current raster coordinates yields the downstream target location coordinates. In this process, the offset is corrected by combining DEM elevation information and raster resolution to ensure that the calculated downstream target location is located in the downhill direction of the water flow and does not exceed the watershed boundary. In this way, each raster cell can quickly determine its potential downstream confluence location, providing basic data for confluence path advancement.

[0050] In another embodiment, it is assumed that the watershed is divided into 50,000 grid cells, each with a side length of 10m. The offset is generated based on the power-law encoded value. , For example, eastward offset grid, Southwest grid, Calculations show that approximately 48,000 of the 50,000 grids have downstream target locations within neighboring grids, while about 2,000 grids require offset adjustments to prevent them from exceeding boundaries due to proximity to boundaries or flat terrain. Ultimately, the average X and Y coordinates of the downstream target locations are approximately 25,000 ± 12,500 grids, ensuring the flow continuity and hydraulic rationality of the entire coding field.

[0051] Determine the downstream grid inflow direction based on the downstream target location; identify coding conflict anomalies based on the downstream target location and the downstream grid inflow direction.

[0052] In one embodiment, for each grid cell, the inflow direction of the downstream grid is calculated based on its downstream target location, i.e., it is determined whether the upstream flow direction of the downstream grid points to the current grid cell. If the inflow direction is consistent with the expected flow direction, it is marked as a normal point; if it is found that the downstream target location of a grid cell points to multiple inflow paths in the neighborhood or forms a loop, it is determined as an anomaly point of coding conflict. After an anomaly is detected, local correction can be performed using the neighborhood average flow direction or weighted direction to make the inflow direction of the grid cell continuous and consistent with the surrounding water flow direction. This step can ensure the physical rationality and computational reliability of the flow direction information before generating the confluence path.

[0053] In another embodiment, assuming that among 50,000 raster cells in the watershed, approximately 46,500 raster cells have normal inflow directions calculated from the downstream target location, while the remaining 3,500 raster cells exhibit coding conflict anomalies. For these anomalies, the average power-law coding direction of their surrounding 8 neighboring raster cells is calculated, and the flow direction of the anomaly raster cells is corrected according to a weight of 0.6:0.4. For example, an eastward conflict point is corrected to a southeast-southeast direction, with an average offset angle correction of 22.5°. After correction, all raster inflow directions form a coherent network without loops or multiple confluence conflicts, providing reliable input data for subsequent confluence path advancement and catchment area delineation.

[0054] Preferably, in step S2, the watershed runoff is calculated based on hydrological and meteorological data, and the recorded runoff includes: Read the rainfall intensity sequence and soil permeability distribution from hydrometeorological data; identify the catchment boundaries of the catchment area; use the catchment boundaries to perform an area-weighted average on the rainfall intensity sequence to generate the inter-time areal rainfall; In one embodiment, hydrological and meteorological data within the target watershed are first acquired, including hourly or minute-by-minute rainfall intensity sequences (in mm / h) and soil permeability distributions (in mm / h) for each sub-region within the watershed. The catchment boundaries and sub-catchment boundaries of the watershed are then identified using a digital elevation model (DEM) and surface feature data. Finally, an area-weighted average of the rainfall intensity within each sub-catchment is calculated to generate a time-varying areal rainfall sequence (in mm / time period) for that sub-region.

[0055] In another embodiment, it is assumed that the target watershed is divided into 10 sub-catchments, the rainfall observation period is 12 hours, and the rainfall intensity sequence (unit mm / h) of each sub-region is as follows: Sub-region 1: [2.0,5.1,0.0,0.0,1.2,3.5,4.0,0.5,0.0,0.0,0.0,0.0,0.0], Sub-region 2: [1.5,4.0,0.0,0.0,0.8,2.9,3.5,0.4,0.0,0.0,0.0,0.0], and so on up to sub-region 10. The area of ​​each sub-region is assumed to be between 1.2 km² and 3.0 km². The rainfall intensity of each sub-region is weighted by area to obtain the time-period isal rainfall sequence. For example, sub-region 1 is [2.2, 5.0, 0.0, ..., 0.0], and sub-region 2 is [1.8, 4.1, 0.0, ..., 0.0]. Finally, the time-period isal rainfall input for the entire watershed is formed, which provides a basis for subsequent infiltration and runoff calculations.

[0056] A dynamic infiltration threshold is set based on the soil permeability distribution; the dynamic infiltration threshold is compared with the inter-period rainfall. If the inter-period rainfall exceeds the dynamic infiltration threshold, watershed runoff is calculated to record the runoff volume.

[0057] In one embodiment, a dynamic infiltration threshold (unit: mm / time period) is set for each sub-region based on the soil permeability distribution. This threshold can be dynamically adjusted according to soil moisture or previous rainfall conditions. Then, the time-period areal rainfall of each sub-region is compared with the corresponding dynamic infiltration threshold. If the time-period areal rainfall is less than or equal to the threshold, the rainfall mainly infiltrates; if the time-period areal rainfall exceeds the threshold, the excess rainfall is converted into runoff. The runoff of the sub-region is calculated using a hydrological model and recorded in the time-series runoff sequence.

[0058] In another embodiment, assuming the dynamic infiltration thresholds for the 10 sub-regions are as follows: Sub-region 1: 3.0, Sub-region 2: 2.5, Sub-region 3: 4.0, Sub-region 4: 3.2, Sub-region 5: 2.8, Sub-region 6: 3.5, Sub-region 7: 3.0, Sub-region 8: 2.9, Sub-region 9: 3.3, Sub-region 10: 3.5 mm / time period, the runoff is obtained by comparing the time period areal rainfall sequence generated in step S1 with the threshold. For example, in Sub-region 1, the time period areal rainfall in the first hour is 2.2 mm ≤ 3.0 mm → no runoff, and in the second hour it is 5.0 mm > 3.0 mm → runoff 2. 0 mm. In sub-region 2, the areal rainfall in the 6th hour was 2.9 mm > 2.5 mm, resulting in a runoff of 0.4 mm. The runoff sequence for each sub-region over 12 hours was calculated sequentially. The cumulative total runoff was 10.5 mm for sub-region 1, 7.2 mm for sub-region 2, 12.1 mm for sub-region 3, 8.7 mm for sub-region 4, 6.9 mm for sub-region 5, 11.3 mm for sub-region 6, 9.5 mm for sub-region 7, 7.8 mm for sub-region 8, 10.0 mm for sub-region 9, and 12.4 mm for sub-region 10, forming the runoff sequence for the entire watershed, providing input for subsequent runoff analysis.

[0059] Preferably, the area-weighted average of the rainfall intensity sequence is performed using the catchment boundary to generate the time-limited areal rainfall, including: The center coordinates are determined using the catchment boundary, and an effective set of rainfall grids is generated based on the center coordinates; rainfall intensity values ​​are read according to the rainfall intensity sequence to form a rainfall distribution field for a given period. In one embodiment, the catchment boundaries of the watershed are determined using a digital elevation model (DEM) and surface feature data, and the center coordinates of each sub-catchment are calculated. Based on the center coordinates, an effective rainfall grid set for each sub-catchment is generated. Subsequently, the hourly rainfall intensity sequence (unit: mm / h) from the hydrometeorological data is read, and the rainfall intensity values ​​are mapped to each effective rainfall grid to form a time-period rainfall distribution field, which is used for subsequent watershed runoff or runoff calculations.

[0060] In another embodiment, it is assumed that the target watershed is divided into 8 sub-catchments, and the number of effective rainfall grids generated in each sub-catchment is as follows: Sub-catchment 1: 25 grids, Sub-catchment 2: 30 grids, Sub-catchment 3: 28 grids, Sub-catchment 4: 22 grids, Sub-catchment 5: 35 grids, Sub-catchment 6: 27 grids, Sub-catchment 7: 33 grids, Sub-catchment 8: 29 grids; the rainfall observation period is 12 hours, and the rainfall intensity values ​​(in mm / h) of each grid are as follows: Sub-catchment 1, Grid 1: [0.0, 1.2, [3.5,0.0,0.0,…,0.5], sub-region 1 grid 2: [0.1,0.8,3.0,0.0,0.0,…,0.3], sub-region 2 grid 1: [0.2,1.0,2.8,0.0,0.0,…,0.4], and so on, forming a complete time-period rainfall distribution field. The average rainfall value of all grids in each sub-region can be calculated in each time period, providing input for the subsequent area-weighted average generation of time-period areal rainfall.

[0061] Based on the effective rainfall raster set and the area-weighted average of the rainfall distribution field during the time period, the isal rainfall during the time period is generated.

[0062] In one embodiment, the generated effective rainfall grid set is averaged with the corresponding time-period rainfall distribution field by area weighting: for each sub-region, the rainfall of each grid is multiplied by the grid area weight, the sum is divided by the total area of ​​the sub-region to obtain the areal rainfall sequence (unit mm / time period) for each sub-region. Then, the areal rainfall of each sub-region is combined to form the time-period areal rainfall sequence of the entire watershed, providing basic data for subsequent infiltration, runoff generation or confluence calculations.

[0063] In another embodiment, assuming that the area of ​​each grid in the eight sub-catchments is between 0.05 and 0.12 km², the rainfall intensity of each grid in each time period is weighted by area. For example, the rainfall values ​​of each grid in the first hour of sub-area 1 are [0.0, 0.1, 0.2, 0.0, 0.3, ..., 0.0], and the weighted average is 0.05 mm of areal rainfall; the grid rainfall values ​​in the second hour are [1.2, 0.8, 1.0, ..., 1.0], and the weighted average is 0.95 mm of areal rainfall; the areal rainfall sequence for 12 hours is calculated as [0.05, 0.95, 2.8, 0.0, 0.0, 1.2, 0.8, 0.3, 0.0, 0.0, 0.1, 0.0]. The 12-hour areal rainfall sequence for sub-region 2 is [0.08, 0.92, 2.5, 0.0, 0.0, 1.0, 0.7, 0.4, 0.0, 0.0, 0.2, 0.0]. The other sub-regions follow the same pattern, ultimately forming 12-hour areal rainfall sequences for all 8 sub-regions of the entire watershed. This provides quantifiable data input for subsequent infiltration threshold determination and watershed runoff simulation.

[0064] Preferably, step S2 involves performing runoff collection simulation in the catchment area using runoff volume to form a runoff process curve, including: Runoff concentration simulation is performed in the catchment area using runoff volume, and the distribution of isochrones is read; the catchment area is divided into multiple isochrone sub-regions based on the isochrones. In one embodiment, runoff collection simulation is first performed on the target catchment area to obtain isochrones. Based on these isochrones, the catchment area is divided into several isochrone sub-areas, each representing a region with similar isochrone times from the boundary to the confluence. Subsequently, parameters such as catchment area, average slope, and soil permeability within each sub-area are recorded to provide input for subsequent sub-area weight calculations and runoff contribution analysis.

[0065] In another embodiment, assuming the total catchment area is 15 km², it is divided into 5 isochronous sub-regions by isochronous lines: sub-region 1 (3.2 km²), sub-region 2 (2.8 km²), sub-region 3 (3.5 km²), sub-region 4 (2.5 km²), and sub-region 5 (3.0 km²). Runoff time series (in m³ / s) for each sub-region are obtained through runoff collection simulation. For example, the isochronous runoff series for sub-region 1 is [0.12, 0.25, 0.38, 0.50, 0.47, 0.30, 0.15], and for sub-region 2 it is [0.08, 0.18, 0.30, 0.42, 0.40, 0.28, 0.12]. The remaining sub-regions are analyzed similarly to obtain the complete isochronous sub-region runoff time series, providing data support for subsequent linear allocation and confluence analysis.

[0066] The area proportion is determined based on multiple isochronous sub-regions, and a sub-region weight sequence is generated; the runoff is linearly allocated based on the sub-region weight sequence and recorded as the runoff contribution value; the time period runoff is calculated based on the runoff contribution value; the time period runoff is connected in series according to its time step to form a runoff process line.

[0067] In one embodiment, a weight sequence is calculated based on the area proportions and topographic parameters of multiple isochronous sub-regions. The total runoff of the catchment area is then linearly distributed to each sub-region based on this sequence, and recorded as runoff contribution values. Subsequently, the runoff contribution values ​​of each sub-region are summed over time to obtain the time-period runoff. By sequentially concatenating all time-period runoff values, a complete runoff hydrograph can be formed for subsequent flood risk assessment or hydrological simulation.

[0068] In another embodiment, assuming the sub-region weights are calculated based on area ratios, the sequence is: Sub-region 1: 0.21, Sub-region 2: 0.18, Sub-region 3: 0.23, Sub-region 4: 0.17, Sub-region 5: 0.21. If the total runoff measured in the total watershed at a certain time period is 2.5 m³ / s, then the runoff contribution values ​​for each sub-region are obtained through linear allocation: Sub-region 1: 0.525 m³ / s, Sub-region 2: 0.45 m³ / s, Sub-region 3: 0.575 m³ / s, Sub-region 4: 0.425 m³ / s, Sub-region 5: 0.525 m³ / s. Based on the time step, the production sequence of each sub-region is as follows: Sub-region 1 [0.12,0.25,0.38,0.50,0.47,0.30,0.15], Sub-region 2 [0.08,0.18,0.30,0.42,0.40,0.28,0.12], Sub-region 3 [0.10,0.22,0.35,0.48,0.46,0.32,0.18], Sub-region 4 [0.07,0.16,0.28,0.40,0.38,0.26,0.11], Sub-region 5 [0.11,0.23,0.36,0.50,0.44,0.32,0.18]. The runoff contribution values ​​of the five sub-regions at each time period are summed to obtain the runoff sequence [0.48, 1.04, 1.67, 2.30, 2.15, 1.48, 0.74] m³ / s for that time period. These are then connected in series according to the time step to form a complete runoff process line, which can be used for subsequent flood evolution simulation or risk analysis.

[0069] Preferably, the linear allocation of runoff based on the sub-region weight sequence is recorded as the runoff contribution value, including: Based on the sub-region weight sequence, a proportional multiplication is performed to generate instantaneous distributed flow; the instantaneous distributed flow is used to plot the time-series flow generation contribution curve; and the time-series flow generation contribution curve is used to calculate the peak delay. In one embodiment, based on the generated sub-region weight sequence, the total runoff of the catchment sub-regions is multiplied proportionally to obtain the instantaneous distribution flow sequence of each sub-region, and a time-series runoff contribution curve is plotted accordingly. Subsequently, by analyzing the time-series runoff curve of each sub-region, the difference between the time of the flood peak and the time of the total basin flood peak is determined, and the peak lag is calculated to assess the time contribution of each sub-region to the total basin runoff response. This information can be used for optimized scheduling, flood warning, and basin management.

[0070] In another embodiment, it is assumed that the catchment area is divided into 5 sub-regions, with a sub-region weight sequence of [0.21, 0.18, 0.23, 0.17, 0.21]. After performing a proportional multiplication on a total runoff of 2.5 m³ / s for a certain period, the instantaneous distribution flow sub-region sequence is obtained: Sub-region 1 [0.52, 0.85, 1.12, 1.35, 1.28, 0.98, 0.62] m³ / s, Sub-region 2 [0.45, 0.72, 0.95, 1.15, 1.10, 0.84, 0.53] m³ / s, Sub-region 3 […]. [0.58, 0.95, 1.25, 1.50, 1.42, 1.05, 0.65] m³ / s, sub-region 4 [0.42, 0.68, 0.90, 1.12, 1.05, 0.78, 0.48] m³ / s, sub-region 5 [0.52, 0.87, 1.15, 1.37, 1.30, 0.97, 0.61] m³ / s. Based on the instantaneous distribution flow curves of each sub-region, the peak delay of each sub-region can be calculated: sub-region 12h, sub-region 21.5h, sub-region 32.2h, sub-region 41.8h, sub-region 52.1h. The difference in arrival time of the peak relative to the total basin peak can be plotted as a complete time-series runoff contribution curve, which intuitively reflects the time contribution characteristics of each sub-region to the peak response.

[0071] The weight correction factor is determined based on the peak delay; the sub-region weight sequence is linearly scaled based on the weight correction factor and recorded as the runoff contribution value.

[0072] In one embodiment, the delay characteristics of each sub-region are analyzed using the peak delay, and a weight correction factor is determined to adjust the original sub-region weight sequence. By applying a linear scaling to the original weight sequence, i.e., multiplying it by the weight correction factor, the corrected sub-region weight sequence is obtained and recorded as the final runoff contribution value for subsequent runoff process simulation or flood risk analysis.

[0073] In another embodiment, assuming the calculated peak delay sequence is [2.0, 1.5, 2.2, 1.8, 2.1]h, the corresponding determined weight correction factor is [0.95, 1.05, 0.92, 1.08, 0.97]. The original sub-region weight sequence [0.21, 0.18, 0.23, 0.17, 0.21] is multiplied term by term by the correction factor to obtain the linearly scaled weight sequence [0.1995, 0.189, 0.2116, 0.1836, 0.2037], which is recorded as the runoff contribution value. Using this runoff contribution value, the actual runoff of each sub-region at different time steps can be further calculated, forming a complete runoff contribution curve, providing a quantitative basis for flood evolution simulation and flood control scheduling.

[0074] Preferably, this specification also provides a flood risk assessment system based on multidimensional data integration analysis, used to perform the flood risk assessment method based on multidimensional data integration analysis as described above, wherein the flood risk assessment based on multidimensional data integration analysis includes: The catchment area identification module 101 is used to acquire a multidimensional dataset, including geographic information data and hydrological and meteorological data; to determine the direction of water flow using the geographic information data; and to identify the catchment area based on the direction of water flow. The runoff collection simulation module 102 is used to calculate watershed runoff based on hydrological and meteorological data and record runoff volume; it also uses the runoff volume to perform runoff collection simulation in the catchment area to form a runoff process line. The peak flow calculation module 103 is used to calculate the peak flow based on the runoff process line; assess the river level based on the peak flow; and calculate the river overflow based on the river level. The flood risk prediction module 104 is used to generate an inundation depth distribution map based on the river overflow; to predict flood risk based on the inundation depth distribution map; and to generate a flood management risk level.

[0075] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0076] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A flood risk assessment method based on multidimensional data ensemble analysis, characterized in that, Includes the following steps: Step S1: Obtain a multidimensional dataset, including geographic information data and hydrological and meteorological data; Geographic information data is used to determine the direction of water flow, and the catchment area is identified based on the direction of water flow; Step S2: Calculate watershed runoff based on hydrological and meteorological data and record the runoff volume; use the runoff volume to perform runoff collection simulation in the catchment area to form a runoff process curve; Step S3: Calculate the peak flow based on the runoff hydrograph; assess the river level based on the peak flow, and calculate the river overflow based on the river level; Step S4: Generate an inundation depth distribution map based on the river overflow; predict flood risk based on the inundation depth distribution map and generate a flood management risk level.

2. The flood risk assessment method based on multidimensional data integration analysis according to claim 1, characterized in that, Step S1, which uses geographic information data to determine the direction of water flow, includes: Generate raster cells using geographic information data; extract elevation values ​​from the raster cells; calculate the elevation differences between raster cells based on the elevation values; and mark the elevation gradient direction based on the elevation differences. The lowest descent path is determined using the elevation gradient direction; along the lowest descent path, the grid positions of the inflowing grid cells are counted and connected to form a confluence channel; the flow direction is determined based on the confluence channel.

3. The flood risk assessment method based on multidimensional data integration analysis according to claim 1, characterized in that, Step S1, which identifies the catchment area based on the direction of water flow, includes: A water flow direction coding field is generated based on the water flow direction; a downstream location fast lookup table is constructed using the water flow direction coding field, and the cumulative flow count of the downstream location fast lookup table is initialized to 1; The downstream location is used to quickly look up a table to determine the downstream pointing relationship; the confluence path is advanced according to the downstream pointing relationship to generate a confluence path network; and the catchment area is identified based on the confluence path network.

4. The flood risk assessment method based on multidimensional data integration analysis according to claim 3, characterized in that, The generation of the water flow direction encoding field based on the water flow direction includes: Read the angle value of the water flow direction and normalize it to a continuous interval of 0° to 360°; divide the angle value into D8 direction intervals; determine the discrete direction based on the D8 direction intervals; The power-law encoding values ​​are assigned according to discrete directions, and a water flow direction encoding field is generated based on the power-law encoding values. During the process of generating the water flow direction encoding field based on the power-law encoding values, encoding conflict anomalies are identified, and local smoothing correction is performed based on the encoding conflict anomalies.

5. The flood risk assessment method based on multidimensional data integration analysis according to claim 4, characterized in that, In the process of generating the water flow direction coding field based on the power-law coding value, identifying coding conflict anomalies includes: In the process of generating the water flow direction coding field based on the power-law coding value, the coordinate offset is calculated using the power-law coding value; and the downstream target position is determined based on the coordinate offset. Determine the downstream grid inflow direction based on the downstream target location; identify coding conflict anomalies based on the downstream target location and the downstream grid inflow direction.

6. The flood risk assessment method based on multidimensional data integration analysis according to claim 1, characterized in that, In step S2, watershed runoff is calculated based on hydrological and meteorological data, and the recorded runoff includes: Read the rainfall intensity sequence and soil permeability distribution from hydrometeorological data; identify the catchment boundaries of the catchment area; use the catchment boundaries to perform an area-weighted average on the rainfall intensity sequence to generate the inter-time areal rainfall; A dynamic infiltration threshold is set based on the soil permeability distribution; the dynamic infiltration threshold is compared with the inter-period rainfall. If the inter-period rainfall exceeds the dynamic infiltration threshold, watershed runoff is calculated to record the runoff volume.

7. The flood risk assessment method based on multidimensional data integration analysis according to claim 6, characterized in that, By applying an area-weighted average to the rainfall intensity sequence using the catchment boundary, the time-limited areal rainfall includes: The center coordinates are determined using the catchment boundary, and an effective set of rainfall grids is generated based on the center coordinates; rainfall intensity values ​​are read according to the rainfall intensity sequence to form a rainfall distribution field for a given period. Based on the effective rainfall raster set and the area-weighted average of the rainfall distribution field during the time period, the isal rainfall during the time period is generated.

8. The flood risk assessment method based on multidimensional data integration analysis according to claim 1, characterized in that, Step S2 involves performing runoff collection simulation in the catchment area using runoff volume to generate a runoff process curve, including: Runoff concentration simulation is performed in the catchment area using runoff volume, and the distribution of isochrones is read; the catchment area is divided into multiple isochrone sub-regions based on the isochrones. The area proportion is determined based on multiple isochronous sub-regions, and a sub-region weight sequence is generated; the runoff is linearly allocated based on the sub-region weight sequence and recorded as the runoff contribution value; the time period runoff is calculated based on the runoff contribution value; the time period runoff is connected in series according to its time step to form a runoff process line.

9. The flood risk assessment method based on multidimensional data integration analysis according to claim 8, characterized in that, Linear allocation of runoff based on sub-region weight sequences, recorded as runoff contribution values, includes: Based on the sub-region weight sequence, a proportional multiplication is performed to generate instantaneous distributed flow; the instantaneous distributed flow is used to plot the time-series flow generation contribution curve; and the time-series flow generation contribution curve is used to calculate the peak delay. The weight correction factor is determined based on the peak delay; the sub-region weight sequence is linearly scaled based on the weight correction factor and recorded as the runoff contribution value.

10. A flood risk assessment system based on multidimensional data ensemble analysis, characterized in that, For executing the flood risk assessment method based on multidimensional data integration analysis as described in claim 1, the flood risk assessment system based on multidimensional data integration analysis comprises: The catchment area identification module is used to acquire multidimensional datasets, including geographic information data and hydrological and meteorological data; it uses geographic information data to determine the direction of water flow and identifies the catchment area based on the direction of water flow. The runoff collection simulation module is used to calculate watershed runoff based on hydrological and meteorological data and record runoff volume; it also uses the runoff volume to perform runoff collection simulation in the catchment area to form a runoff process curve. The peak flow calculation module is used to calculate the peak flow based on the runoff process line; assess the river level based on the peak flow; and calculate the river overflow based on the river level. The flood risk prediction module is used to generate an inundation depth distribution map based on the river overflow; to predict flood risk based on the inundation depth distribution map; and to generate a flood management risk level.

Images