A mountain torrent disaster risk assessment method based on hydrological hydrodynamic-disaster-bearing body coupling and double coefficient correction

Abstract

The application discloses a mountain torrent disaster risk assessment method based on hydrological hydrodynamic-disaster-bearing body coupling and double coefficient correction, and relates to the technical field of water conservancy projects. By constructing a one-dimensional river flood routing and two-dimensional submerged area expansion coupled hydrological hydrodynamic model, combining a double coefficient system (water depth increment coefficient and energy damage coefficient), quantifying a silt accumulation "geometric lifting effect" and water-sediment mixed "dynamic enhancement effect" correction model, a "disaster factor-disaster-bearing body vulnerability" double-element coupling framework is constructed, disaster parameter calculation area comprehensive risk values are substituted, risk level spatial visualization is realized in combination with a GIS platform, and a risk layer and a visualized graph of an emergency command system are output. The method can improve model simulation accuracy, solve the interference of silt and water-sediment mixture on risk assessment, and is clear, visual and intuitive in risk grading, and directly serves disaster prevention planning and emergency response, and improves the scientific nature and efficiency of mountain torrent disaster prevention and control.

Description

Technical Field

[0001] This invention relates to the field of water conservancy engineering technology, and in particular to a method for assessing the risk of flash floods based on the coupling of hydrology and hydrodynamics with the disaster-bearing body and dual-coefficient correction. Background Technology

[0002] Currently, there are several core technologies and models for flash flood disaster risk assessment: First, hydrological and hydrodynamic simulation technology, which uses one-dimensional / two-dimensional numerical models (such as the Saint-Venant equation and shallow water equation) to reproduce the formation, evolution, and inundation process of floods, and obtain disaster-causing factor parameters such as water depth, flow velocity, and inundation range. Its core advantage is that it can characterize the dynamic features of floods and is suitable for complex terrains in small and medium-sized watersheds. Second, vulnerability assessment technology for disaster-bearing bodies, which establishes damage discrimination models for facilities such as buildings, roads, and bridges based on historical disaster data or empirical criteria. Its core function is to connect disaster-causing factors with disaster losses and support risk quantification. Third, hydrodynamic models based on the clear water assumption (such as HEC-RAS and MIKE FLOOD) construct the water flow propagation path and water level changes throughout the entire process of "rainfall-runoff-confluence-inundation" by solving shallow water equations or Saint-Venant equations; fourth, quasi-static models based on DEM-filling depressions (such as FloodArea) quickly generate inundation layers through water overflow algorithms, mainly used for preliminary zoning and emergency deployment; and fifth, physical coupling models based on multiphase mechanics (such as OpenFOAM), based on the Navier-Stokes equations and multiphase flow theory, handle the complex physical interactions between fluid, particles, and structures, and are suitable for mechanism research in extreme scenarios.

[0003] The above model has the following technical limitations:

[0004] 1. The disaster-causing factors and the disaster-bearing bodies lack systematic coupling. Most studies either focus on flood evolution simulation and ignore the mechanism of facility damage, or focus on the vulnerability of disaster-bearing bodies and are detached from the actual flood dynamic process, and cannot fully cover the dynamic chain of "flood evolution-building damage-risk classification".

[0005] 2. Data and parameterization bottlenecks are prominent. Although foreign countries rely on high-resolution remote sensing (LiDAR) and big data to improve accuracy, the real-time performance is insufficient under extreme weather conditions. In China, due to the sparse monitoring network in mountainous areas and incomplete historical flood records, the model relies heavily on empirical parameters, which limits its universality.

[0006] 3. Risk classification lacks a unified standard, and quantile methods and expert threshold methods are often used. The comparability of different research results is poor, and an internationally accepted quantitative system has not been formed, making it difficult to promote across regions.

[0007] 4. The model response efficiency is out of sync with actual needs. Traditional two-dimensional hydrodynamic simulation has a long calculation cycle (serial calculation takes several hours) and low automation, which cannot meet the emergency needs of "short-term results" for sudden flash floods. Finally, there is insufficient interdisciplinary integration. The advantages of hydrology, structural engineering and GIS technology have not been fully integrated, making it difficult to cope with the characteristics of complex mountain terrain, concentrated rainfall and strong suddenness of disasters.

[0008] 5. Structural defects in clear water models: Mainstream models such as HEC-RAS are based on the "Newtonian clear water" assumption, which assumes that fluid properties do not change with sediment concentration and ignores three key effects: bed uplift and backwater effect caused by sediment deposition, kinetic energy amplification effect caused by increased fluid density, and destructive effect of particle impact on structures. This leads to an underestimation of risk by more than 20% in "high-energy sensitive zones" such as fan-shaped and canyon-shaped sections, and a false negative rate of nearly 30% in extremely high-risk sections.

[0009] 6. Insufficient practicality of physical coupling models: Models such as OpenFOAM require high-precision sediment parameters (concentration curves, particle size distribution, etc.), but data acquisition in mountainous areas is difficult; in addition, the computational cost is extremely high, with a 1-hour process simulation in a small watershed taking more than 10 hours, and the output format needs to be processed twice before it can be connected to the GIS platform, which cannot meet the needs of real-time emergency response.

[0010] 7. The risk measurement logic is too simplistic: Existing models all use "water depth + flow velocity" as the core indicators to classify risk levels, which lacks sensitivity to the coupling mechanism of "water-sand-energy". This makes it easy to misjudge high-risk scenarios with "low water but high energy" as medium or low risk, thus restricting the scientific nature of prevention and control decisions. Summary of the Invention

[0011] To address the aforementioned issues, this invention provides a method for assessing flash flood risk based on hydrological-hydrodynamic coupling and dual-coefficient correction. This method automates the entire process from flood evolution simulation to disaster damage assessment, risk level classification, and result visualization, generating a high-precision and timely flash flood risk level distribution map. Secondly, it addresses the core pain points in the field of flash flood risk assessment in high-sand-containing mountainous areas, namely, the systematic underestimation of risk by clear water models and the difficulty in engineering implementation of physically coupled models. This invention provides a dual-coefficient correction assessment method that balances accuracy, efficiency, and practicality.

[0012] This invention is implemented as follows:

[0013] A method for assessing flash flood risk based on hydrodynamic-disaster-bearing body coupling and dual-coefficient correction includes the following steps:

[0014] Step S1, Basic Data Preprocessing and Disaster Factor Simulation:

[0015] 1.1 Data Collection and Preprocessing: Acquire 30m resolution DEM data, hydrological data, disaster-bearing body data, and underlying surface data of the study area; perform outlier removal and georegistration on the collected data.

[0016] 1.2 Hydrological and hydrodynamic model construction: A coupled framework of "one-dimensional river channel model + two-dimensional inundation model" is adopted to construct a full-scenario simulation system for flash flood disaster factors;

[0017] 1.3 Model Parameter Configuration: The finite volume method is used to perform discrete calculations on the one-dimensional river channel model and the two-dimensional inundation model respectively. The flux is calculated by the Roe approximation Riemann solution. The spatial second-order accuracy is improved by combining MUSCL reconstruction. The prediction-correction scheme is used to ensure the temporal second-order accuracy. An unstructured hybrid mesh is used and the local area is refined. The friction coefficient is set according to the land use type based on the Manning formula. GPU parallel acceleration technology is introduced to improve the model calculation efficiency.

[0018] 1.4 Output of disaster-causing factors: Run the hydrological and hydrodynamic model to output the inundation range, water depth distribution, flow velocity field and flood duration, which serve as the core input parameters for damage assessment of the disaster-bearing body;

[0019] Step S2, Risk assessment of mountain floods and sediment hazards based on dual-coefficient correction:

[0020] The two-dimensional inundation model adopts a mature clear water hydrodynamic model. Running the model yields a clear water baseline parameter grid, and the baseline risk level is calculated. ;

[0021] The water depth increment coefficient is defined and calculated using the following formula:

[0022]

[0023] in, The density of the debris flow For the density of clear water, To predict the average sediment thickness, The average water depth calculated for the clear water model. This is the siltation efficiency coefficient;

[0024] Define the energy destruction factor The calculation formula is as follows:

[0025]

[0026] in, The velocity of the debris flow. For the flow rate of clean water, The impact characteristic coefficient;

[0027] Calculate risk level The unified correction formula is as follows:

[0028]

[0029] Based on a preset fixed threshold, the risk level is classified into "low-medium-high-extremely high", and a level transition list is derived to clearly record the calculation unit for upgrading from medium level to high level and from high level to extremely high level, as well as the corresponding dominant influencing factors.

[0030] Step S3, Vulnerability Characterization and Damage Assessment of the Disaster-Bearing Body:

[0031] 3.1 Classification and Parameter Calibration of Disaster-Bearing Bodies: Disaster-bearing bodies are classified according to their structural characteristics, including buildings, bridges, and roads. The critical energy absorption capacity of each type of disaster-bearing body is calibrated using historical disaster data. With the damage evolution index m;

[0032] 3.2 Calculation of cumulative destructive energy of floods: Based on the power-time integral model of the velocity-cubic relationship, a density correction factor for sediment-laden mixed fluid is introduced to construct a calculation model for the cumulative destructive energy of high sediment-laden flash floods. The density of the sediment-laden mixed fluid is calculated by the density of water, the density of sediment, and the volume concentration of sediment. The velocity is calculated by replacing the flow rate with the Manning formula.

[0033] 3.3 Calculation of Damage and Failure Probability of Disaster-Bearing Body: Constructing the energy-damage correlation, combined with the critical energy absorption capacity calibrated in step 3.1. Using the damage evolution index m, calculate the overall damage index D of the disaster-bearing body; convert the damage index D into the probability of destruction of the disaster-bearing body P using the logistic function;

[0034] Step S4, Risk Comprehensive Classification and Result Output:

[0035] 4.1 Risk Value Calculation: Combining the intensity of the disaster-causing factor output in step 1.4 and the probability of damage to the disaster-bearing body P obtained in step 3.3, the comprehensive risk value of flash flood disaster in the study area is calculated by using range standardization.

[0036] 4.2 Risk Classification: The comprehensive risk value is classified into four levels according to a unified standard: extremely dangerous, dangerous, warning, and attention, and the flood inundation characteristics and damage risk of the disaster-bearing body corresponding to each level of risk are clearly defined;

[0037] 4.3 Output Results: The risk classification results are output in GeoTIFF format, which retains the DEM projection information. Optionally, a pseudo-color risk distribution map in PNG format can be generated. The output includes a list of damaged disaster-bearing bodies containing the probability of damage to the disaster-bearing bodies and the main disaster-causing factors.

[0038] Furthermore, in step 1, the hydrological data includes hourly rainfall data, which is generated by spatial interpolation from five or more official rain gauges to form a spatial rainfall field; the disaster-bearing body data includes building structure type, bridge span / material, road grade, and latitude and longitude; the underlying surface data includes roughness coefficients corresponding to different land use types.

[0039] Furthermore, in step 1, the one-dimensional river channel model simulates the flood evolution of the main channel based on the Saint-Venant equation, and the two-dimensional inundation model simulates the inundation of the floodplain and built-up areas based on the shallow water equation. The coupling interface is set as the top line of the river levee, and the coupling trigger condition is that the calculated water level of the one-dimensional river channel is greater than or equal to the top elevation of the levee. After triggering, the data exchange between the one-dimensional and two-dimensional models is realized through overflow. The time step adopts a synchronous iteration method and satisfies the CFL stability condition, CFL<0.8, to ensure the stability of the calculation.

[0040] Furthermore, in step 3, the calculation formula for the cumulative destructive energy of the flood is as follows:

[0041]

[0042] In the formula, Accumulate destructive energy for the disaster-bearing body; The start time of the flood inundation; This refers to the end time of the flood inundation. For instantaneous destructive power; For the density of the mixed fluid; This is the drag coefficient; The instantaneous effective area of ​​the disaster-bearing body impacted by the flood; This refers to the instantaneous flood velocity.

[0043] Furthermore, in step 3, the energy-damage correlation is constructed using the following formula:

[0044]

[0045] Where D represents the overall damage index of the structure, with a value ranging from 0 to 1, where D=0 indicates no damage and D=1 indicates complete destruction. To accumulate destructive energy for the disaster-bearing body, This represents the critical energy absorption capacity of the disaster-bearing body.

[0046] Furthermore, in step 3, the probability P of the disaster-bearing body being destroyed is calculated using the following formula:

[0047]

[0048] Among them, α and β are empirical fitting parameters, with α ranging from 0.5 to 1.0 and β ranging from -3.0 to 0, obtained by fitting historical disaster data of the study area.

[0049] Furthermore, in step 4, the comprehensive risk value is calculated using the following formula:

[0050]

[0051] in, These are the standardized range values ​​for water depth, flow velocity, flood duration, and probability of damage, respectively, ranging from 0 to 1; each weight satisfies ω1+ω2+ω3+ω4=1, and the value range of a single weight is 0 to 1.

[0052] Furthermore, in step 4, the risk classification is divided into four levels according to a unified quantitative standard: Red, extremely dangerous: comprehensive risk value 0.8~1.0, corresponding to complete flooding of the danger zone, high probability of damage to resettlement sites / evacuation routes; Orange, dangerous: comprehensive risk value 0.6~0.8, corresponding to complete flooding of the danger zone; Yellow, warning: comprehensive risk value 0.4~0.6, corresponding to the danger zone beginning to flood; Blue, attention: comprehensive risk value 0.2~0.4, corresponding to flooding of the floodplains, danger zone may be flooded; when the comprehensive risk value <0.2, it is judged as no risk.

[0053] The beneficial effects of this invention are:

[0054] 1. Construct a coupled hydrological and hydrodynamic model of "one-dimensional river flood evolution + two-dimensional inundation zone expansion", and combine GPU parallel computing technology to quickly output disaster-causing factor parameters such as inundation range, water depth distribution, flow velocity field, and flood duration, so as to ensure the accuracy and real-time quantification of disaster-causing factors;

[0055] 2. Innovate building damage assessment technology based on energy time history integral, establish a cumulative damage energy model that considers the characteristics of sediment-laden water flow, combine the critical energy absorption capacity of the structure, derive the mapping relationship between damage index and failure probability, and accurately characterize the vulnerability of disaster-bearing bodies such as buildings and bridges.

[0056] 3. Construct a dual-factor coupling framework of "disaster-causing factors and vulnerability of disaster-bearing bodies", substitute the disaster-causing parameters (water depth, flow velocity, duration) output by hydrodynamic simulation into the building failure model, calculate the comprehensive risk value of the region, and avoid the limitations of assessment that "emphasizes a single factor";

[0057] 4. Establish a unified four-level risk classification standard (red, orange, yellow, and blue), combine it with a GIS platform to realize spatial visualization of risk levels, and output GeoTIFF risk layers and PNG visualization maps that can be directly connected to the emergency command system to meet the actual needs of disaster prevention planning and emergency response.

[0058] 5. Construct a dual-coefficient system with clearly defined physical meaning (water depth increment coefficient) Energy destruction coefficient The study quantifies the "geometric uplift effect" of sediment deposition and the "dynamic enhancement effect" of water-sediment mixing, breaking through the limitations of traditional clear water models that rely solely on a single path of "water depth + flow velocity," and filling the gap in characterizing the coupling mechanism of "water-sediment-energy."

[0059] 6. Enables low-intrusion modification of existing mature clean water models (such as HEC-RAS and MIKEFLOOD) without changing the model kernel structure and solution logic. Risk assessment accuracy can be improved simply by post-processing and overlaying coefficient grids, thus lowering the integration threshold with existing business platforms (GIS, emergency early warning systems).

[0060] 7. Accurately identify "critical transition units" in "key sensitive zones" such as fan-shaped openings, bridge and culvert backwaters, and rural residential areas, solving the problem of the clear water model missing high-risk scenarios of "deep water but high energy," and generating a "risk level transition list" to clarify the dominant factors amplifying the risk of each unit. leading / (Dominant / Co-dominant).

[0061] 8. Adapt to real-world scenarios where data acquisition is difficult in mountainous areas and emergency response time is tight. While ensuring assessment accuracy, control the complexity of data requirements (most parameters are derived from easily obtainable data such as drone aerial photography and historical disaster damage surveys), compress calculation time, and meet the operational needs of "rapid assessment, accurate early warning, and targeted management" of flash floods in high sandy mountainous areas.

[0062] The present invention will be explained in detail below with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0063] Figure 1 This is a flowchart of the flash flood disaster risk assessment method of the present invention;

[0064] Figure 2 This is a roadmap of the double-coefficient correction technology in Embodiment 1 of the present invention;

[0065] Figure 3 This is a schematic diagram of the simulation results of the two-dimensional shallow water equation in Embodiment 1 of the present invention;

[0066] Figure 4 This is a diagram illustrating the mechanism of the water depth increment coefficient in Embodiment 1 of the present invention;

[0067] Figure 5 This is a diagram illustrating the energy destruction coefficient mechanism in Embodiment 1 of the present invention;

[0068] Figure 6 This is a remote sensing image of the Lengshui River basin before the disaster, as shown in Embodiment 2 of the present invention.

[0069] Figure 7 This is a remote sensing image of the debris flow after it occurred in Embodiment 2 of the present invention;

[0070] Figure 8 This is a diagram showing the overall modeling of the Lengshuihe River scheme in Embodiment 2 of the present invention;

[0071] Figure 9 This is a modeling diagram of the Xintianba Reservoir breach scheme in Embodiment 2 of the present invention;

[0072] Figure 10 This is a modeling diagram of the traffic congestion solution for the Love Bridge in Langjing Village in Embodiment 2 of the present invention;

[0073] Figure 11 This is a modeling diagram of the collapse scheme of the Love Bridge in Langjing Village in Embodiment 2 of the present invention;

[0074] Figure 12 This is a risk level distribution diagram of Embodiment 2 of the present invention;

[0075] Figure 13 This is the contribution decomposition and risk level transition diagram of each unit of the double-coefficient correction in Embodiment 1 of the present invention. Detailed Implementation

[0076] Example 1:

[0077] This embodiment presents a method for assessing flash flood risk based on hydrological-hydrodynamic-disaster-bearing body coupling and dual-coefficient correction. Figure 1 As shown, it includes the following steps:

[0078] Step S1, Basic Data Preprocessing and Disaster Factor Simulation:

[0079] 1.1 Data Collection and Preprocessing:

[0080] Topographic data: Acquire 30m resolution DEM (Digital Elevation Model) data of the study area for topographic parameter extraction, including slope and hydraulic radius;

[0081] Hydrological data: Hourly rainfall data, generated from five or more official rain gauges, using spatial interpolation (such as Kriging interpolation) to create a spatial rainfall field;

[0082] Disaster-bearing body data: including building structure type, bridge span / material, road grade and latitude and longitude, removing outlier data and completing georeferencing;

[0083] Underlying surface data: roughness coefficients corresponding to land use types (e.g., 0.035 for cultivated land, 0.05 for forest land);

[0084] 1.2 Hydrological and Hydrodynamic Model Construction: A coupled framework of "one-dimensional channel model + two-dimensional inundation model" is adopted. The one-dimensional model simulates the flood evolution of the main channel based on the Saint-Venant equation (continuity equation + momentum equation), while the two-dimensional model simulates the inundation of the floodplain and built-up areas based on the shallow water equation (mass conservation + momentum conservation). The coupling interface is set as the top line of the river levee, and the coupling trigger condition is that the calculated water level of the one-dimensional channel is greater than or equal to the top elevation of the levee. After triggering, data exchange between the one-dimensional and two-dimensional models is achieved through overflow. The time step adopts a synchronous iteration method and satisfies the CFL stability condition (CFL<0.8) to ensure computational stability.

[0085] 1.3 Model Parameter Configuration: The finite volume method is used to discretize the equations, and the flux is calculated using the Roe approximation Riemann solution. Spatial second-order accuracy is improved through MUSCL reconstruction, and the prediction-correction scheme ensures temporal second-order accuracy. The computational grid adopts an unstructured hybrid grid (locally refined to 5m). The friction coefficient is set according to land use type and calculated using the Manning formula. The Manning roughness n values ​​are: forest land 0.04~0.06, cultivated land 0.02~0.04, and built-up area 0.03~0.05. GPU parallel acceleration is introduced. The numerical solution process of the two-dimensional shallow water equation is implemented in parallel using the PyTorch framework, with a time step of 1 second. Compared with CPU serial calculation at the same computational accuracy, the overall computational efficiency is improved by about 100 times.

[0086] 1.4 Disaster-causing factor output: The running model outputs the inundation range, water depth distribution h, flow velocity field u / v, and flood duration t, which serve as the core inputs for subsequent disaster-bearing body assessment.

[0087] Step S2, Risk assessment of mountain floods and sediment hazards based on dual-coefficient correction:

[0088] like Figure 2 As shown, a dual-coefficient system with clearly defined physical meaning (water depth increment coefficient) is constructed. Energy destruction coefficient This method quantifies the "geometric uplift effect" of sediment deposition and the "dynamic enhancement effect" of water-sediment mixing, breaking through the limitations of traditional clear water models that rely solely on a single path of "water depth + flow velocity," and filling the gap in characterizing the coupling mechanism of "water-sediment-energy." It also provides a low-intrusive modification to existing mature models without changing the model's core structure or solution logic. Risk assessment accuracy can be improved simply by overlaying coefficient grids in post-processing, reducing the integration threshold with existing business platforms (GIS, emergency early warning systems).

[0089] The model adopts a mature clear water hydrodynamic model (such as HEC-RAS, MIKE FLOOD), and the input basic data includes: topographic data: 30m resolution DEM and gully cross-section data;

[0090] Hydrological data: hourly rainfall observation data (spatial rainfall field generated by Kriging interpolation);

[0091] Underlying surface data: roughness coefficients corresponding to land use types (e.g., 0.035 for cultivated land, 0.05 for forest land);

[0092] Boundary conditions: Water level / discharge process at the watershed outlet.

[0093] Running the model yields a clear water baseline parameter grid: reference water depth Reference flow rate The flooding duration T is used to calculate the baseline risk level using standardized empirical formulas. The calculation formula is as follows:

[0094] (1)

[0095] Figure 3 This is a schematic diagram of the simulation results of the two-dimensional shallow water equation under clear water conditions. The color represents the reference water depth. Distribution, arrows indicate baseline flow velocity Based on the time series output of this simulation, the inundation duration T of each raster cell can be further extracted to form a complete raster dataset of clear water baseline parameters, which serves as a benchmark for risk level. The calculation provides the basic parameters.

[0096] 2.1 Obtaining and Calculating Double-Coefficient Parameters:

[0097] 2.1.1 Water depth increment coefficient Modeling of geometric lift effect:

[0098] Water depth increment coefficient Used to measure the effect of sediment deposition on water level rise and inundation expansion caused by changes in bed morphology, it is a key control term in the disaster geometry path. Extensive field observations and back-calculation analyses show that in abrupt topographic changes such as canyons and bridges, even a sediment deposition of only a few tens of centimeters can trigger a backwater rise of more than several tens of centimeters, thereby altering the static inundation boundary of the clear water model.

[0099] Figure 4 By comparing the water depth distribution of clear water and sediment-laden water, it is clearly shown that sediment-laden water significantly increases the water depth increment coefficient. This means that, under the same flow rate, the water depth of sediment-laden water is significantly higher than that of clear water, thus amplifying the range and depth of flood inundation.

[0100] To achieve a quantitative description of this impact, the water depth increment coefficient is defined as follows:

[0101] (2)

[0102] in, The unit weight of the debris flow is (kg / m³). For the density of clear water, To predict the average sediment thickness (m). The average water depth (m) calculated for the clear water model. The sedimentation efficiency coefficient is related to the abundance of sediment source, the degree of channel shrinkage, the local slope, and the backwater morphology.

[0103] This expression maps "bed surface rise of a few centimeters to tens of centimeters" to "equivalent water level rise of several centimeters" using "thickness ratio × relative weight change × backwater geometric amplification", thus compensating for the systematic underestimation of static flooding by the clear water model.

[0104] Regarding parameter acquisition, Thickness samples derived from UAV / 3D reconstruction, intra-event concentration curves, and volume fractions of effective deposition periods; Calculate by volume fraction or take empirical values ​​from similar watersheds; One-time regression calibration is performed using historical event trace elevations, measured immersion depths and thickness samples. The upper limit (0.8–1.0) is usually taken for the canyon-fan section and the backwater section of bridges and culverts, while the lower limit (0.5–0.7) is taken for open beaches.

[0105] By spatializing this coefficient, slight changes in the bed surface can be mapped to a systematic increase in the risk index, solving the problem of the clear water model's passive underestimation of "seemingly shallow but implicitly high-energy" risks. Especially in sections with dense structures or reverse slope constraints, this kind of geometric amplification effect is often the "critical point" for disaster transition.

[0106] 2.1.2 Energy Destruction Coefficient Modeling of dynamic enhancement effect:

[0107] Energy destruction coefficient It is used to reflect the enhanced kinetic energy per unit area caused by the dual increase in density and velocity of sediment mixtures, and is a key control quantity in the dynamic path of disasters. In scenarios such as debris flows in mountainous areas and floods with high sediment content, although the velocity values ​​are similar to those of clear water, the actual destructive capacity often far exceeds the prediction range of clear water models due to the increased density of the mixture, enhanced inertia, and interparticle impact.

[0108] Figure 5 By comparing the dynamic pressure distribution of clear water and sediment-laden water flow, it is clearly demonstrated that sediment-laden water flow significantly increases the energy destruction coefficient. That is, under the same conditions, the energy and destructive power of sediment-laden water flows are far greater than that of clear water, thus exacerbating the erosion and impact of flash floods on the underlying surface.

[0109] Therefore, the following expression is proposed:

[0110] (3)

[0111] in The velocity of the debris flow is (m / s). The velocity of the clean water is (m / s). Impact characteristic coefficients related to particle size distribution, gravel content, and boundary materials (concrete / masonry / soil).

[0112] This expression combines "heavier" (density-to-kinetic-energy ratio) and "faster" (velocity-to-square ratio) into a dimensionless amplification, and then uses... The differences in materials and structures are consistent with the failure mechanism and facilitate a one-time recalibration at the historical damage point.

[0113] Regarding parameter acquisition, It can be estimated by video speed measurement, cross-sectional continuity equation or radar-flow relationship, and appropriately adjusted according to the blockage / backwater situation; It is recommended to perform regression according to the zonation (coarse gravel / fine grain, well reinforced / generally reinforced). For coarse gravel, the water-facing side or unreinforced area can be 0.25–0.30, and for fine grain or well reinforced area, the value can be 0.12–0.20.

[0114] Compared to the clear water model, this coefficient significantly improves the accuracy of identifying structure failure risks under "high-speed-high-impact" scenarios, especially in areas with significant local blockage, backslope water flow, or particle impact, where the kinetic energy ratio is often the key factor dominating damage. This can be achieved through spatial deployment. The coefficient layer can make the local kinetic energy anomaly area visible, which helps to prioritize and control it in emergency response deployment.

[0115] 2.2 Dual-coefficient correction and risk generation:

[0116] Risk map of clear water Based on the risk level The unified correction formula is as follows:

[0117] (4)

[0118] in The main control is the geometric increment of "depth / width" (the equivalent water level rise and backwater expansion caused by the bed surface rise). The main control is the increase in dynamic force (due to the increased weight of the mixture and the increased destructive strength caused by particle impact). Multiplying these two factors gives the total amplification, naturally including cross-effect terms. In the critical zone where "a small amount of siltation can significantly raise the water level and the upstream components are densely packed," such as the gorge-fan area and the backwater of bridges and culverts, this intersection often determines whether the extremely high risk becomes "manifest."

[0119] 2.3 Exporting the Level Jump List:

[0120] Risk levels are mapped based on preset fixed thresholds, with the following classification standards: low (0–30), medium (31–60), high (61–90), and extremely high (>90). A list of level transitions is exported, clearly recording the calculation units that are upgraded from medium to high level and from high to extremely high level, as well as their corresponding dominant influencing factors.

[0121] Based on this, the "critical transition units" of "key sensitive zones" such as fan-shaped openings, bridge and culvert backwaters, and rural residential areas are accurately identified to solve the problem of the clear water model missing high-risk scenarios of "deep water but high energy," generating a "risk level transition list" and clarifying the dominant factors that amplify the risk of each unit. leading / (Dominant / Co-dominant).

[0122] Figure 13 The chart shows the changes in the risk index before and after the correction. The corrected risk index (orange line) has been upgraded from high to very high in several risk units. The following chart uses the contribution decomposition of the correction factor to present the changes in different risk units. (Geometric effect) The contribution ratio of (dynamic effect) and the interaction between the two provides quantitative support for determining the dominant influencing factors of each calculation unit in the grade jump list, and confirms the matching relationship between the extremely high risk area and the warning range.

[0123] Step S3, Vulnerability Characterization and Damage Assessment of the Disaster-Bearing Body:

[0124] 3.1 Classification and Parameter Calibration of Disaster-Bearing Bodies: Disaster-bearing bodies are classified according to structural characteristics (buildings: reinforced concrete / brick-concrete / timber-rammed earth; bridges: large >30m / medium 15-30m / small <15m; roads: provincial highways / rural roads / village roads / trailways). The critical energy absorption capacity of each type of disaster-bearing body is calibrated using historical disaster data. With the damage evolution index m; where typical parameters include:

[0125] reinforced concrete houses =500kJ / m², m=1.2; brick-concrete houses =300kJ / m², m=1.0; small bridges =800kJ / m², m=1.5, the parameters of other types of disaster-bearing bodies were calibrated using the same method;

[0126] 3.2 Calculation of Cumulative Damage Energy from Floods: Based on the power-time integral model of the velocity-cubic relationship, the density of the mixed fluid (including sand) is introduced, and the formula is as follows:

[0127] The cumulative destructive energy is calculated using the following formula:

[0128] (5)

[0129] In the formula: Accumulated destructive energy for the disaster-bearing body (unit: kJ); The start time of flood inundation (unit: seconds); Time of flood inundation end (unit: seconds); Instantaneous destructive power (unit: W); The density of the mixed fluid (unit: kg / m³). The drag coefficient is dimensionless and ranges from 0.6 to 1.0. The instantaneous effective area of ​​the disaster-bearing body subjected to flood impact (unit: m², calculated based on the actual projected area of ​​the disaster-bearing body subjected to impact). Instantaneous flood velocity (unit: m / s).

[0130] The density of the mixed fluid is calculated using the following formula:

[0131] (6)

[0132] in, The sediment volume concentration is dimensionless and ranges from 0 to 0.3, determined based on measured sediment concentration data in the study area.

[0133] The formula for calculating velocity substitution (Manning's formula) is as follows:

[0134] (7)

[0135] Where n is the Manning roughness coefficient; R is the hydraulic radius, in meters, calculated based on the river / surface cross-sectional morphology; and S is the riverbed / surface slope, dimensionless, extracted from DEM data.

[0136] 3.3 Calculation of the probability of damage and destruction of the disaster-bearing body:

[0137] The energy-damage relationship is constructed using the following formula:

[0138] (8)

[0139] Where D represents the overall damage index of the structure, ranging from 0 to 1, where D=0 indicates no damage and D=1 indicates complete failure; it is then converted into the failure probability using the logistic function, calculated as follows:

[0140] (9)

[0141] Among them, α and β are empirical fitting parameters, with α ranging from 0.5 to 1.0 and β ranging from -3.0 to 0, obtained by fitting historical disaster data of the study area.

[0142] Step S4, Risk classification and result output of the coupling of hydraulic and engineering damage factors:

[0143] A dual-factor coupling framework of "disaster-causing factors and vulnerability of disaster-bearing bodies" is constructed. The disaster-causing parameters (water depth, flow velocity, and duration) output by hydrodynamic simulation are substituted into the building damage model to calculate the comprehensive risk value of the region, thus avoiding the limitations of assessment that "emphasizes a single factor".

[0144] 4.1 Risk Value Calculation: Combining the intensity of disaster-causing factors (water depth, flow velocity, duration) and the probability of damage to the disaster-bearing body (P), a range standardization process is used to obtain the standardized value of each evaluation unit; then, a weighted summation method is used to calculate the comprehensive risk value, and the calculation formula is as follows:

[0145] (10)

[0146] in, These are the standardized range values ​​(ranging from 0 to 1) for water depth, flow velocity, flood duration, and probability of damage, respectively; each weight satisfies ω1+ω2+ω3+ω4=1, and the value range of each individual weight is 0 to 1.

[0147] 4.2 Risk Classification: Divided into four levels according to a unified quantitative standard---Red (Extremely Dangerous: Comprehensive risk value 0.8~1.0, corresponding to complete flooding of the danger zone, high probability of damage to resettlement sites / evacuation routes), Orange (Dangerous: Comprehensive risk value 0.6~0.8, corresponding to complete flooding of the danger zone), Yellow (Warning: Comprehensive risk value 0.4~0.6, corresponding to the beginning of flooding of the danger zone), Blue (Attention: Comprehensive risk value 0.2~0.4, corresponding to flooding of the floodplains, the danger zone may be flooded); when the comprehensive risk value is <0.2, it is judged as no risk.

[0148] 4.3 Output Results: The classification results will be output in GeoTIFF format (preserving DEM projection information). Optional pseudo-color risk distribution map (PNG format) can be generated. At the same time, a "List of Damaged Subjects" (including damage probability and main disaster-causing factors) will be output. The criteria for determining the main disaster-causing factors are: the disaster-causing factor (water depth, flow velocity, duration) with the largest standardized value or the probability of damage. If the probability of damage has the largest standardized value, then the main disaster-causing factor is the probability of damage; otherwise, the main disaster-causing factor is the disaster-causing factor with the largest standardized value.

[0149] 4.4 Results Integration: Import the GeoTIFF risk map into the GIS emergency platform, overlay the disaster-bearing body layer, and realize the linkage query of "risk level - disaster-bearing body location".

[0150] Example 2:

[0151] This embodiment presents a method for assessing flash flood disaster risk based on the coupling of hydrology and hydrodynamics with the disaster-bearing body. Figure 6 , 7 As shown, the study area is as follows: the Lengshui River basin covers an area of ​​160 km², with a total population of 68,000 (65% of whom live in the river valley). The disaster-bearing structures include 13,500 buildings (58% brick-concrete, 27% reinforced concrete, and 15% other types), 82 km of county and township roads, and 21 medium and small bridges. The average annual rainfall is 1450 mm (72% of which occurs from June to September), and the extreme peak flow exceeds 420 m³ / s.

[0152] The village and town watershed covers an area of ​​28.6 km², with a main ditch length of 9.2 km (gradient of 12.8%), 3 administrative villages (3,500 people), and is divided into 12 assessment units (covering the upstream sediment source area, gorge section, bridge and culvert area, and town area); a rainstorm on July 9, 2025 (maximum hourly rainfall intensity of 50 mm, cumulative rainfall of 180 mm) triggered a high-sediment-laden flash flood.

[0153] Step S1, Basic Data Preprocessing and Disaster Factor Simulation:

[0154] 1.1 Data collection: 30m resolution DEM (to extract slope and hydraulic radius), 20-year hourly rainfall data from 5 rain gauge stations (to generate a spatial rainfall field), historical disaster data from 2008 / 2016 (to calibrate parameters of the disaster-bearing body), and pre-disaster / post-disaster remote sensing images (to verify the inundation range);

[0155] Drone-borne aerial images after the disaster (inverting sediment thickness), data on red / orange warning areas issued by the provincial early warning platform (verification results), and output results of the HEC-RAS clear water model ( , , );

[0156] Data preprocessing: Use the rasterio library to perform DEM geographic correction, remove outliers in rainfall data that are greater than 3 standard deviations, and use ArcGIS to match the latitude and longitude of disaster-bearing bodies with the coordinates of the DEM, filtering out invalid disaster-bearing bodies at the boundary.

[0157] The boundary of the DEM was filled using the numpy.pad function, and the average sediment thickness was predicted by inversion from UAV imagery. (1.0m in town area, 1.2m in bridge and culvert backwater section, 0.3m in open beach area), take the bulk density of debris flow. =1750 kg / m³ (empirical value for high-sediment-laden flow in the Loess Plateau), debris flow velocity Based on the reference flow rate (2.5m / s) Reduced by 20%, resulting in 2.0m / s.

[0158] 1.2 Construction and Operation of Hydrological and Hydrodynamic Models:

[0159] Model construction: The model adopts a coupling of "one-dimensional Saint-Venant equation + two-dimensional shallow water equation", with one dimension covering the main channel of the river (9.2km in length) and two dimensions covering the floodplain and built-up area.

[0160] 1.3 Parameter configuration: The mesh is a mixed triangular / quadrilateral unstructured mesh (the main trench is fined to 5m, and other areas are 10m). Manning roughness coefficient: forest land 0.05, cultivated land 0.03, built area 0.04; boundary conditions: upstream is the rainfall inflow boundary, and downstream is the water level control boundary.

[0161] 1.4 Simulation Operation: The simulation was performed using the PyTorch framework to achieve GPU parallelism with a time step of 1 second. The simulation simulated a 12-hour flood process (the water depth reached 2.5m and the maximum flow velocity was 3.5m / s within 6 hours after the rainstorm). The output included the inundation range (covering 60% of the watershed) and the spatial distribution map of water depth / flow velocity. The simulation was compared with the historical inundation traces in 2016, and the error was <10%.

[0162] Step S2, Risk assessment of mountain floods and sediment hazards based on dual-coefficient correction:

[0163] 2.1 Double coefficient calculation and calibration:

[0164] 2.1.1 Calculate unit by unit according to the formula :

[0165] Backwater section:

[0166] Open beach area: ;

[0167] generate Grid (0.17-0.56).

[0168] 2.1.2 Calculate unit by unit according to the formula :

[0169] Water-facing side of bridge and culvert: =0.28×1750 / 1000×(4² / 5²)≈0.25;

[0170] Village and town areas: =0.22×1.75×(3.5² / 4²)≈0.19;

[0171] generate Grid (0.15-0.25).

[0172] 2.1.3 Coefficient Spatialization: Using ArcGIS, the 12 units were spatialized... , Generate a raster map to ensure accurate mapping of coefficients for sensitive areas such as bridges, culverts, and towns.

[0173] 2.2 Risk Correction and Results Verification:

[0174] 1) Risk Correction Calculation

[0175] according to = ×(1+ )×(1+ Calculation per grid:

[0176] Backwater section of bridge and culvert: =70×(1+0.56)×(1+0.25)≈136 (extremely high risk);

[0177] Open beach area: =40×(1+0.17)×(1+0.15)≈53 (medium risk).

[0178] 2.3 Risk Level Classification and Jump List:

[0179] Level classification: Extremely high risk (>90) is concentrated in the backwater sections of bridges and culverts and in town areas, which completely matches the scope of the provincial red alert;

[0180] Transition List: Extract 3 "High → Extremely High" transition units and label the bridge / culvert section as " + "Collaboration-led", with the town area as " leading".

[0181] Step S3, Vulnerability Characterization and Damage Assessment of the Disaster-Bearing Body:

[0182] 3.1 Classification and Parameter Calibration of Disaster-Bearing Bodies: After classifying by structure, parameters were calibrated using 2016 disaster data---reinforced concrete houses. =500kJ / m², m=1.2; brick-concrete houses =300kJ / m², m=1.0; small bridges =800kJ / m², m=1.5; α=0.8, β=-2.5;

[0183] 3.2 Calculation of cumulative destructive energy: Taking the volume concentration of sediment from this rainstorm as an example. =0.2,

[0184] Given Manning's roughness coefficient n = 0.035, hydraulic radius R = 1.2 m, and slope S = 0.012, we get... 2.8 m / s; according to , ( =0.6, =10m²), calculated to =410kJ;

[0185] Damage and destruction probability: For reinforced concrete houses, D=(410 / 500)^1.2≈0.85, P=1 / (1+e^(-(0.8×0.85-2.5)))=0.23; for brick-concrete houses, D=(410 / 300)^1.0≈1.37, P=0.68, which matches the actual damage rate in 2016 (65% for brick-concrete houses) by more than 95%.

[0186] Step S4, Risk Comprehensive Classification and Result Output:

[0187] Figures 8 to 11 This demonstrates a complete simulation of the evolution of flash floods in river channels: Figure 8 It shows the evolution of the flood in the main channel and floodplain, with the flood advancing downstream along the main channel and causing flooding in some areas; Figure 9 It shows the water depth distribution in the early stage of the flood's evolution. At this time, the water depth is generally low and mainly concentrated in the main channel. The floodplains have not yet been submerged on a large scale. The flow field shows that the flood is slowly advancing from upstream to downstream. Figure 10 It depicts the interaction between flood and underlying surface, and how engineering measures such as dikes constrain the lateral spread of flood, thereby changing the flow field structure and inundation range; Figure 11 It shows the spatial distribution of water depth at the final moment of the flood's evolution, with the water depth gradually increasing from upstream to downstream, intuitively quantifying the level of inundation risk; Figure 12 The system uses color-coded levels to display the inundation extent and risk level of flood events. Red, green, and yellow areas correspond to different levels of inundation, providing comprehensive technical support for flood control planning and disaster risk management at the watershed scale.

[0188] 4.1 Risk Value Calculation: The comprehensive risk value is obtained by summing the standardized values ​​of water depth (weight 0.4), flow velocity (weight 0.3), and probability of damage (weight 0.3).

[0189] 4.2 Classification Results: The red extremely dangerous zone is concentrated in the valley settlements (water depth > 1.5m, brick-concrete houses P > 0.6) and around small bridges (flow velocity > 3m / s); the orange dangerous zone is the edge of the valley (water depth 1.0-1.5m); the yellow warning zone is the farmland near the valley (water depth 0.5-1.0m); and the blue concern zone is the highlands (water depth < 0.5m).

[0190] 4.3 Output: Output GeoTIFF format risk map (retaining UTM projection) and PNG pseudo-color map, and output "damage risk list" (including 5 high-risk bridges out of 21 bridges and 1200 high-risk buildings out of 13500 buildings), which matches the actual disaster situation in 2016 by more than 90%.

[0191] 4.4 Verification results: The accuracy of identifying extremely high-risk areas was improved by 28% compared with the clear water model, and all two bridge and culvert damage points that were missed were made explicit.

[0192] Step S5: Matching Risk Assessment Results with Governance:

[0193] Results Integration: Import GeoTIFF risk maps into the GIS emergency platform, overlay disaster-bearing body layers, and enable linked queries of "risk level - disaster-bearing body location";

[0194] Matching governance measures: targeting In the main area (downstream backwater area), dredging and desilting works will be deployed (removing 0.5-1.0m of silt). In the primary area (township area), construct energy dissipation structures (such as stilling basins); in coordination with the primary area (bridge and culvert sections), implement water-facing reinforcement (thicken bridge foundations and add anti-collision piers).

[0195] Effectiveness verification: After the measures were implemented, under the same rainstorm scenario, the probability of damage in the town area decreased from 0.68 to 0.21, and the risk level of the bridge and culvert area decreased from red to orange, verifying the effectiveness of the measures.

[0196] The core function of this invention:

[0197] 1. Collaborative assessment of disaster cause and disaster bearing capacity: It achieves deep coupling between hydrological and hydrodynamic simulation and building energy damage assessment, and can quantify the mapping relationship of "flood energy-structural damage-risk level", covering the entire disaster chain;

[0198] 2. Precise correction of clear water model: Corrects traditional clear water model in a "low-intrusion" manner without reconstructing the kernel. It only makes up for the lack of sediment geometry and dynamic effects by using dual coefficients, and is compatible with existing business platforms.

[0199] 3. Rapid response and visualization: By introducing GPU parallel acceleration, the entire process (data input - result output) takes only 15 minutes, and the correction process takes less than 5 minutes; the output results are compatible with GIS and can generate pseudo-color maps and grade transition lists to intuitively support decision-making.

[0200] 4. Unified grading and traceability: Establish a unified four-level standard of red / orange / yellow / blue, and at the same time, achieve traceability of risk sources and matching of governance measures through coefficient decomposition and damage factor labeling.

[0201] The main advantages of this invention are:

[0202] 1. Solve the problem of underestimation by traditional models: The accuracy of identifying extremely high-risk areas is improved by 25% compared with the clear water model, and the underreporting rate of engineering damage points is reduced by 25 percentage points. It can capture the "critical transition" risk of sensitive areas such as fan-shaped openings and bridges and culverts.

[0203] 2. Balancing accuracy and operational feasibility: It does not rely on high-precision data from complex models such as OpenFOAM, but uses readily available DEM, rainfall, and UAV data, and can be directly integrated with existing emergency platforms (such as flash flood disaster survey and assessment platforms), making integration difficult.

[0204] 3. Strong regional adaptability: parameters (α, λ, (etc.) can be calibrated using historical data to adapt to different mountainous landforms such as the Qinling-Bashan Mountains and the Loess Plateau. The correction range is small in low-sensitivity areas (gentle slopes, open beaches), avoiding excessive warnings.

[0205] 4. Supporting precise governance: through the decomposition of dominant factors (such as town areas) (Leading role, bridge and culvert area coordinated leadership), which can be matched with "dredging and siltation" ( Dominant area), energy dissipation structure ( Differentiated measures such as "leading area" and "water-facing reinforcement (coordination area)" are used to form an "assessment-governance" closed loop.

[0206] This invention achieves full-process system optimization from flood evolution simulation, disaster-bearing body damage assessment, sediment-laden flow correction to risk classification visualization, meeting the needs of rapid, accurate, practical and operational emergency assessment of flash flood disasters in high sediment-laden mountainous areas.

[0207] The above description is only used to illustrate the technical solutions of the present invention and is not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention (such as the application of various formulas, the order of steps, etc.) without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for assessing flash flood disaster risk based on hydrological-hydrodynamic-disaster-bearing body coupling and dual-coefficient correction, characterized by comprising the following steps: Step S1, Basic Data Preprocessing and Disaster Factor Simulation: 1.1 Data Collection and Preprocessing: Acquire 30m resolution DEM data, hydrological data, disaster-bearing body data, and underlying surface data of the study area; perform outlier removal and georegistration on the collected data. 1.2 Hydrological and hydrodynamic model construction: A coupled framework of "one-dimensional river channel model + two-dimensional inundation model" is adopted to construct a full-scenario simulation system for flash flood disaster factors; 1.3 Model Parameter Configuration: The finite volume method is used to perform discrete calculations on the one-dimensional river channel model and the two-dimensional inundation model respectively. The flux is calculated by the Roe approximation Riemann solution. The spatial second-order accuracy is improved by combining MUSCL reconstruction. The prediction-correction scheme is used to ensure the temporal second-order accuracy. An unstructured hybrid mesh is used and the local area is refined. The friction coefficient is set according to the land use type based on the Manning formula. GPU parallel acceleration technology is introduced to improve the model calculation efficiency. 1.4 Output of disaster-causing factors: Run the hydrological and hydrodynamic model to output the inundation range, water depth distribution, flow velocity field and flood duration, which serve as the core input parameters for damage assessment of the disaster-bearing body; Step S2, Risk assessment of mountain floods and sediment hazards based on dual-coefficient correction: The two-dimensional inundation model adopts a mature clear water hydrodynamic model. Running the model yields a clear water baseline parameter grid, and the baseline risk level is calculated. ; The water depth increment coefficient is defined and calculated using the following formula: in, The density of the debris flow For the density of clear water, To predict the average sediment thickness, The average water depth calculated for the clear water model. This is the siltation efficiency coefficient; Define the energy destruction factor The calculation formula is as follows: in, The velocity of the debris flow. For the flow rate of clean water, The impact characteristic coefficient; Calculate risk level The unified correction formula is as follows: Risk levels are classified into "low-medium-high-extremely high" based on preset fixed thresholds, and a level transition list is derived to clearly record the calculation units for upgrading from medium level to high level and from high level to extremely high level, as well as the corresponding dominant influencing factors. Step S3, Vulnerability Characterization and Damage Assessment of the Disaster-Bearing Body: 3.1 Classification and Parameter Calibration of Disaster-Bearing Bodies: Disaster-bearing bodies are classified according to their structural characteristics, including buildings, bridges, and roads. The critical energy absorption capacity of each type of disaster-bearing body is calibrated using historical disaster data. With the damage evolution index m; 3.2 Calculation of cumulative destructive energy of floods: Based on the power-time integral model of the velocity-cubic relationship, a density correction factor for sediment-laden mixed fluid is introduced to construct a calculation model for the cumulative destructive energy of high sediment-laden flash floods. The density of the sediment-laden mixed fluid is calculated by the density of water, the density of sediment, and the volume concentration of sediment. The velocity is calculated by replacing the flow rate with the Manning formula. 3.3 Calculation of Damage and Failure Probability of Disaster-Bearing Body: Constructing the energy-damage correlation, combined with the critical energy absorption capacity calibrated in step 3.

1. Using the damage evolution index m, calculate the overall damage index D of the disaster-bearing body; convert the damage index D into the probability of destruction of the disaster-bearing body P using the logistic function; Step S4, Risk Comprehensive Classification and Result Output: 4.1 Risk Value Calculation: Combining the intensity of the disaster-causing factor output in step 1.4 and the probability of damage to the disaster-bearing body P obtained in step 3.3, the comprehensive risk value of flash flood disaster in the study area is calculated by using range standardization. 4.2 Risk Classification: The comprehensive risk value is classified into four levels according to a unified standard: extremely dangerous, dangerous, warning, and attention, and the flood inundation characteristics and damage risk of the disaster-bearing body corresponding to each level of risk are clearly defined; 4.3 Output Results: The risk classification results are output in GeoTIFF format, which retains the DEM projection information. Optionally, a pseudo-color risk distribution map in PNG format can be generated. The output includes a list of damaged disaster-bearing bodies containing the probability of damage to the disaster-bearing bodies and the main disaster-causing factors.

2. The flash flood disaster risk assessment method based on hydrological-hydrodynamic-disaster-bearing body coupling and dual-coefficient correction as described in claim 1, characterized in that, In step 1, the hydrological data includes hourly rainfall data, which is generated by spatial interpolation from five or more official rain gauges to form a spatial rainfall field; the disaster-bearing body data includes building structure type, bridge span / material, road grade, and latitude and longitude; the underlying surface data includes roughness coefficients corresponding to different land use types.

3. The flash flood disaster risk assessment method based on hydrological-hydrodynamic-disaster-bearing body coupling and dual-coefficient correction as described in claim 1, characterized in that, In step 1, the one-dimensional river channel model simulates the flood evolution of the main channel based on the Saint-Venant equation, and the two-dimensional inundation model simulates the inundation of the floodplain and built-up areas based on the shallow water equation. The coupling interface is set as the top line of the river levee, and the coupling trigger condition is that the calculated water level of the one-dimensional river channel is greater than or equal to the top elevation of the levee. After triggering, the data exchange between the one-dimensional and two-dimensional models is realized through overflow. The time step adopts a synchronous iteration method and satisfies the CFL stability condition, CFL<0.8, to ensure the stability of the calculation.

4. The flash flood disaster risk assessment method based on hydrological-hydrodynamic-disaster-bearing body coupling and dual-coefficient correction as described in claim 1, characterized in that, In step 3, the cumulative destructive energy of the flood is calculated using the following formula: In the formula, Accumulate destructive energy for the disaster-bearing body; The start time of the flood inundation; This refers to the end time of the flood inundation. For instantaneous destructive power; For the density of the mixed fluid; This is the drag coefficient; The instantaneous effective area of ​​the disaster-bearing body impacted by the flood; This refers to the instantaneous flood velocity.

5. The flash flood disaster risk assessment method based on hydrological-hydrodynamic-disaster-bearing body coupling and dual-coefficient correction as described in claim 1, characterized in that, In step 3, the energy-damage correlation is constructed using the following formula: Where D represents the overall damage index of the structure, with a value ranging from 0 to 1, where D=0 indicates no damage and D=1 indicates complete destruction. To accumulate destructive energy for the disaster-bearing body, This represents the critical energy absorption capacity of the disaster-bearing body.

6. The flash flood disaster risk assessment method based on hydrological-hydrodynamic-disaster-bearing body coupling and dual-coefficient correction as described in claim 1, characterized in that, In step 3, the probability P of the disaster-bearing body being destroyed is calculated using the following formula: Among them, α and β are empirical fitting parameters, with α ranging from 0.5 to 1.0 and β ranging from -3.0 to 0, obtained by fitting historical disaster data of the study area.

7. The flash flood disaster risk assessment method based on hydrological-hydrodynamic-disaster-bearing body coupling and dual-coefficient correction as described in claim 1, characterized in that, In step 4, the comprehensive risk value is calculated using the following formula: in, These are the standardized range values ​​for water depth, flow velocity, flood duration, and probability of damage, respectively, ranging from 0 to 1; each weight satisfies ω1+ω2+ω3+ω4=1, and the value range of a single weight is 0 to 1.

8. The flash flood disaster risk assessment method based on hydrological-hydrodynamic-disaster-bearing body coupling and dual-coefficient correction as described in claim 1, characterized in that, In step 4, the risk classification is divided into four levels according to a unified quantitative standard: Red, extremely dangerous: comprehensive risk value 0.8~1.0, corresponding to the entire danger zone being flooded, and a high probability of damage to resettlement sites / evacuation routes; Orange, dangerous: comprehensive risk value 0.6~0.8, corresponding to the entire danger zone being flooded; Yellow, warning: comprehensive risk value 0.4~0.6, corresponding to the danger zone beginning to be flooded; Blue, attention: comprehensive risk value 0.2~0.4, corresponding to flooding and the danger zone possibly being flooded. When the overall risk value is less than 0.2, it is considered as risk-free.

Images