A method and system for determining dynamic critical rainfall in mountainous watersheds

Abstract

This invention relates to the field of natural disaster prediction and assessment technology, and discloses a method and system for determining dynamic critical rainfall in mountainous watersheds. The method includes the following steps: acquiring basic data, constructing and calibrating a runoff generation and confluence model, and outputting the actual soil moisture content; normalizing soil wetting status; extracting time-series critical boundary point sets; generating an explicit formula for dynamic critical rainfall; using a two-dimensional hydrodynamic model to calculate water depth and flow velocity values; performing fixed-increment iterative correction to determine the minimum effective critical rainfall; writing the boundary fitting formula parameters and the minimum effective critical rainfall values ​​corresponding to different wetting levels and different rainfall durations into a dynamic critical rainfall threshold sample library, and assessing the risk of the target disaster-bearing objects. This invention can effectively improve the consistency between critical rainfall results and actual disaster-causing conditions.

Description

Technical Field

[0001] This invention relates to the field of natural disaster prediction and assessment technology, specifically to a method and system for determining dynamic critical rainfall in mountainous watersheds. Background Technology

[0002] Mountainous watersheds are characterized by undulating terrain, steep river slopes, and rapid flow, making them prone to flash floods when affected by short-duration heavy rainfall. These flash floods threaten settlements, roads, bridges, and riverside facilities. Existing flash flood early warning technologies mostly rely on fixed critical rainfall, water level, or flow thresholds for identification. Among these, the critical rainfall method is widely used due to its ease of calculation and practical application.

[0003] However, existing methods for determining critical rainfall typically employ empirical statistical methods or fixed threshold methods, failing to adequately consider the impact of prior soil wetting conditions on watershed runoff generation and confluence processes. This results in difficulties in accurately characterizing the disaster-causing response corresponding to the same rainfall event under different initial wetting conditions. Furthermore, existing technologies generally lack mechanisms for progressively verifying and correcting critical rainfall using flood inundation results when determining critical rainfall, making it difficult to ensure consistency between the threshold results and actual water depth, flow velocity, and disaster-causing conditions. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for determining dynamic critical rainfall in mountainous watersheds. This method can construct explicit formulas for dynamic critical rainfall under different rainfall durations based on the current soil moisture state of the watershed, and use two-dimensional hydrodynamic simulation results to incrementally correct the minimum effective critical rainfall to determine the minimum effective critical rainfall, thereby improving the consistency between the critical rainfall results and the actual disaster-causing conditions.

[0005] This invention is implemented as follows: A method for determining the dynamic critical rainfall in a mountainous watershed includes acquiring rainfall, evaporation, underlying surface and historical flood data of the study watershed, constructing and calibrating a runoff generation and confluence model, and outputting the actual soil moisture content of the watershed at the current moment. Acquire rainfall, evaporation, underlying surface, and historical flood data for the study watershed, perform preprocessing, and establish a basic database for the study watershed; A runoff generation and runoff model was constructed based on the basic database of the watershed, and the parameters of the constructed runoff generation and runoff model were calibrated. The actual soil moisture content of the watershed at the current time was calculated using the calibrated runoff generation and runoff model. The runoff generation and confluence model is a dual-supermodel, which simultaneously considers the super-permeability runoff generation process and the saturation runoff generation process, and outputs the actual soil moisture content.

[0006] S2. Based on the soil saturation water content, the actual soil water content of the watershed at the current moment is normalized to obtain the soil water saturation. The soil water saturation is then divided into multiple humidity levels. Each humidity level is used to index the corresponding dynamic critical rainfall explicit formula and minimum effective critical rainfall. Soil saturated water content is predetermined based on soil type parameters and underlying surface parameters of the study watershed, and assigned values ​​according to watershed calculation units.

[0007] S3. Construct historical samples exceeding warning levels and historical samples not exceeding warning levels, and based on these samples, extract critical boundary point sets for each of the multiple preset rainfall durations. Historical samples exceeding warning levels are those that reached the preset warning state during historical flood events, while historical samples not exceeding warning levels are those that did not reach the preset warning state during historical flood events. The critical boundary point sets are obtained by using a boundary screening model to screen the historical samples exceeding warning levels and historical samples not exceeding warning levels.

[0008] S4. Fit the critical boundary point set for each rainfall duration to generate the explicit formula for the dynamic critical rainfall for each rainfall duration, and calculate the initial dynamic critical rainfall for different rainfall durations. S5. When the real-time rainfall or forecast rainfall reaches the initial dynamic critical rainfall for the corresponding rainfall duration, the corresponding rainfall process is input into the two-dimensional hydrodynamic model to simulate the flood propagation process in the target area and calculate the maximum water depth and maximum flow velocity of each calculation unit in the target area. S6. Compare the output results of the two-dimensional hydrodynamic model with the preset disaster criteria. When the current simulation results do not meet the disaster criteria, the dynamic critical rainfall under the current rainfall duration is incrementally corrected by a fixed rainfall increment. S7. Repeat the two-dimensional hydrodynamic simulation to form an incremental sequence consisting of an initial dynamic critical rainfall and several fixed rainfall increments, until the disaster criterion changes from unsatisfied to satisfied for the first time in the incremental sequence. The corresponding current round rainfall value is determined as the minimum effective critical rainfall under the current soil moisture saturation and the current rainfall duration. In other words, the minimum effective critical rainfall is the rainfall value that makes the disaster criterion change from unsatisfied to satisfied for the first time in the incremental sequence.

[0009] S8. Write the boundary fitting formula parameters and minimum effective critical rainfall value corresponding to different humidity levels and different rainfall durations into the dynamic critical rainfall threshold sample library. Based on the dynamic critical rainfall threshold sample library and real-time rainfall data, determine the risk of the target disaster-bearing object.

[0010] The dynamic critical rainfall threshold sample library is used to directly call the corresponding boundary fitting formula parameters and minimum effective critical rainfall under the same humidity level and the same rainfall duration.

[0011] The target disaster-affected areas may include at least one of the following: residential areas, roads, bridges, river floodplains, and low-lying, flood-prone areas.

[0012] Furthermore, the method for calibrating the fixed-production confluence model in step S1 includes: constructing a penalty function to describe the degree of violation of constraints by the parameter combination, and constructing a fitness function to optimize the parameter combination. The smaller the penalty function, the closer the parameter combination is to the feasible region; the smaller the fitness function, the closer the simulated flow process corresponding to the set of parameters is to the target flood process.

[0013] Furthermore, the method for extracting the critical boundary point set for each rainfall duration from multiple preset rainfall durations includes: For each preset rainfall duration, extract the boundary point set to output the critical boundary point set for the corresponding rainfall duration; Boundary point set extraction is performed using a boundary screening model, which is a support vector machine model.

[0014] Furthermore, in step S4, the least squares linear fitting method is used to fit the critical boundary point set for each rainfall duration.

[0015] Furthermore, in step S5, the governing equations of the two-dimensional hydrodynamic model include the continuity equation and the momentum equation. The two-dimensional hydrodynamic model uses the continuity equation and the momentum equation to simulate the flood propagation process in the target area and outputs the maximum water depth and maximum flow velocity values ​​of each calculation unit within a preset time period.

[0016] Furthermore, in step S6, the disaster criterion may include: the maximum water depth in the target area is greater than or equal to a preset water depth threshold; the maximum flow velocity in the target area is greater than or equal to a preset flow velocity threshold; or the product of the maximum water depth and the maximum flow velocity in the target area is greater than or equal to a preset combined threshold.

[0017] Furthermore, in step S6, the fixed rainfall increment is an integer multiple of the preset rainfall resolution for the corresponding rainfall duration.

[0018] Furthermore, in step S8, the dynamic critical rainfall threshold sample library is indexed according to the four-element mapping relationship of "humidity level - rainfall duration - boundary fitting formula parameters - minimum effective critical rainfall". When the watershed is at the same humidity level and corresponds to the same rainfall duration, the boundary fitting formula parameters and minimum effective critical rainfall in the dynamic critical rainfall threshold sample library are directly called to improve the efficiency of real-time early warning.

[0019] Furthermore, in step S8, the risk assessment results include the minimum effective critical rainfall value, the distribution information of dangerous areas, and the risk level information of the target disaster-bearing objects.

[0020] A system for determining dynamic critical rainfall in mountainous watersheds includes: The runoff generation and confluence model construction unit is used to acquire rainfall, evaporation, underlying surface and historical flood data of the study watershed, construct and calibrate the runoff generation and confluence model, and output the actual soil moisture content of the watershed at the current moment; The wet state normalization unit is used to calculate the soil moisture saturation based on the actual soil moisture content and soil saturation moisture content output by the runoff generation and confluence model. The boundary point set extraction unit is used to construct historical over-alert samples and historical non-over-alert samples respectively, and extract the critical boundary point set for each rainfall duration based on the historical over-alert samples and historical non-over-alert samples for multiple preset rainfall durations respectively; The explicit formula generation unit is used to fit the critical boundary point set for each rainfall duration, generate the explicit formula for the dynamic critical rainfall for each rainfall duration, and calculate the initial dynamic critical rainfall for different rainfall durations. The incremental iterative correction unit is used to call the two-dimensional hydrodynamic model when the rainfall reaches the dynamic critical rainfall amount under the corresponding rainfall duration, and to incrementally correct the dynamic critical rainfall amount with a fixed rainfall increment according to the preset disaster criterion, until the disaster criterion changes from not being met to being met for the first time in the incremental sequence. The threshold sample library storage unit is used to store different humidity levels, different rainfall durations, boundary fitting formula parameters, and minimum effective critical rainfall values; The disaster-bearing object risk assessment unit is used to assess the risk of target disaster-bearing objects based on a dynamic critical rainfall threshold sample library and real-time rainfall data.

[0021] Compared with the prior art, the beneficial effects of the present invention are: This invention introduces soil moisture saturation to achieve a normalized characterization of the early-stage wet state of the watershed, enabling the critical rainfall to be dynamically adjusted according to changes in the wet state. This invention achieves direct calculation of dynamic critical rainfall for different durations by extracting critical boundary point sets for different rainfall durations and fitting explicit formulas for each set. This invention uses a two-dimensional hydrodynamic model and a fixed rainfall increment sequence to gradually correct the dynamic critical rainfall, so that the final determined minimum effective critical rainfall is more consistent with the actual disaster-causing conditions. This invention constructs a dynamic critical rainfall threshold sample library, which allows for direct retrieval of corresponding results under the same humidity level and rainfall duration, thereby improving the efficiency of real-time business applications. This invention can construct explicit formulas for dynamic critical rainfall under different rainfall durations based on the current soil moisture state of the watershed, and determine the minimum effective critical rainfall by incrementally correcting the two-dimensional hydrodynamic simulation results with a fixed rainfall increment, thereby improving the consistency between the critical rainfall results and the actual disaster-causing conditions. Attached Figure Description

[0022] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to represent selected embodiments of the invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] Please see Figure 1 A method for determining the dynamic critical rainfall in a mountainous watershed includes the following steps: S1. Obtain rainfall, evaporation, underlying surface and historical flood data for the study watershed, construct and calibrate the runoff generation and confluence model, and output the actual soil moisture content of the watershed at the current moment; Acquire digital elevation models, land use data, soil type data, rainfall data, evaporation data, historical flood data, river cross-section data, and historical disaster data for the study watershed. Perform depression filling, flow direction analysis, and runoff accumulation analysis on the digital elevation model to extract watershed boundary, river network structure, and slope information; perform time-stepping unification and outlier verification on rainfall, water level, and flow data to establish a basic database for the study watershed.

[0026] This invention employs a dual-super-super-model to describe and study the runoff generation and confluence processes in a watershed. The dual-super-super-model simultaneously considers both infiltration-excess runoff generation and storage-saturation runoff generation, thus effectively reflecting the flood formation mechanism under torrential rainfall conditions in small mountain watersheds.

[0027] To improve the accuracy of model parameter calibration, an improved genetic algorithm is used for parameter optimization. A penalty function is constructed to describe the degree to which parameter combinations violate constraints:

[0028] In the formula: X is the vector of parameters to be optimized; This is an inequality constraint function; is the equality constraint function; m is the total number of constraints; q is the number of inequality constraints.

[0029] The smaller the value, the closer the parameter combination is to the feasible region.

[0030] To further optimize parameter combinations, a fitness function is constructed:

[0031] In the formula: The degree to which the target flow range is violated; To simulate the deviation of the flow rate from the lower limit of the interval; This is to simulate the deviation of the flow rate from the upper limit of the interval.

[0032] The smaller the value, the closer the simulated flow process corresponding to this set of parameters is to the target flood process.

[0033] In the selection operator, the sum of individual fitness is:

[0034] The probability of each individual being selected is:

[0035] In the formula: M is the population size; Let be the fitness value of the i-th individual; The probability of being selected for the next generation.

[0036] Through the above calibration, the parameters of the dual-super-water model suitable for the study watershed were obtained, and the actual soil moisture content at the current time was output using the calibrated dual-super-water model. .

[0037] S2. Based on the soil saturation water content, the actual soil water content of the watershed at the current moment is normalized to obtain the soil water saturation, and the soil water saturation is divided into multiple humidity levels. To uniformly characterize the pre-wet state of the watershed, soil moisture saturation was used as a normalized index, and its expression is as follows:

[0038] In the formula, Soil moisture saturation; This represents the current actual soil moisture content; This represents the soil saturation water content. The soil saturation water content is predetermined based on soil type parameters and underlying surface parameters of the study watershed, and is assigned a value according to the watershed calculation unit.

[0039] Based on a preset range, the soil moisture saturation is divided into multiple humidity levels for subsequent indexing and retrieval of dynamic critical rainfall formulas and minimum effective critical rainfall results.

[0040] S3. Construct historical samples of exceeding the warning level and historical samples of not exceeding the warning level respectively, and extract the critical boundary point set for each rainfall duration based on the historical samples of exceeding the warning level and historical samples of not exceeding the warning level respectively; The critical boundary point set is obtained by screening historical over-alarm samples and historical non-over-alarm samples using a boundary screening model.

[0041] For multiple preset rainfall durations, sample sets are constructed, consisting of historical samples exceeding warning levels and historical samples not exceeding warning levels. The historical samples exceeding warning levels are those that reached the preset warning state during historical flood events, while the historical samples not exceeding warning levels are those that did not reach the preset warning state during historical flood events.

[0042] Critical boundary point sets are extracted for each rainfall duration. Preferably, a boundary screening model, such as a support vector machine model, is used to extract the boundary point sets. An independent boundary screening model is established for each preset rainfall duration to output the critical boundary point set for that rainfall duration.

[0043] S4. Fit the critical boundary point set for each rainfall duration to generate the explicit formula for the dynamic critical rainfall for each rainfall duration, and calculate the initial dynamic critical rainfall for different rainfall durations. After obtaining the critical boundary point set for each rainfall duration, a dynamic critical rainfall explicit formula is constructed for each rainfall duration using the least squares linear fitting method:

[0044] In the formula, Rainfall duration The dynamic critical rainfall; and The fitting coefficients are the values ​​corresponding to the duration of the rainfall.

[0045] For example, if the preset rainfall duration includes 1 hour, 3 hours, and 6 hours, the corresponding explicit formulas for dynamic critical rainfall are as follows: P1h = -9.159S + 60.966 P3h = -48.305S + 170.255 P6h = -32.782S + 232.836 Therefore, based on the current soil moisture saturation... The initial dynamic critical rainfall for different rainfall durations can be directly obtained.

[0046] S5. When the real-time rainfall or forecast rainfall reaches the initial dynamic critical rainfall for the corresponding rainfall duration, the corresponding rainfall process is input into the two-dimensional hydrodynamic model to calculate the maximum water depth and maximum flow velocity of each calculation unit in the target area. When the real-time or forecasted rainfall reaches the initial dynamic critical rainfall for the corresponding rainfall duration, the corresponding rainfall process is input into a two-dimensional hydrodynamic model to simulate the flood propagation process in the target area. The governing equations of the two-dimensional hydrodynamic model include the continuity equation and the momentum equation.

[0047] The continuity equation is:

[0048] The momentum equation is:

[0049] In the formula: For water depth; and Let be the velocity components in two orthogonal directions; q be the lateral inflow term; and g be the gravitational acceleration. The turbulent viscosity coefficient; The coefficient of friction at the bottom; is the Coriolis coefficient.

[0050] The two-dimensional hydrodynamic model outputs the maximum water depth and maximum flow velocity values ​​for each calculation unit within a preset time period.

[0051] S6. Compare the output results of the two-dimensional hydrodynamic model with the preset disaster criteria. When the current simulation results do not meet the disaster criteria, the dynamic critical rainfall under the current rainfall duration is incrementally corrected by a fixed rainfall increment. S7. Repeat the two-dimensional hydrodynamic simulation to form an incremental sequence consisting of the initial dynamic critical rainfall and several fixed rainfall increments, until the disaster criterion changes from not being met to being met for the first time in the incremental sequence. The corresponding current round of rainfall value is determined as the current soil moisture saturation and the minimum effective critical rainfall under the current rainfall duration. The output results are compared with preset disaster-causing criteria. Disaster-causing criteria may include: the maximum water depth in the target area is greater than or equal to a preset water depth threshold; the maximum flow velocity in the target area is greater than or equal to a preset flow velocity threshold; or the product of the maximum water depth and the maximum flow velocity in the target area is greater than or equal to a preset combined threshold.

[0052] When the results of the current simulation do not meet the criteria for disaster occurrence, a fixed rainfall increment is used. An incremental correction is applied to the dynamic critical rainfall amount under the current rainfall duration, and a new two-dimensional hydrodynamic simulation is performed. The rainfall increment is fixed. It is an integer multiple of the preset rainfall resolution for the corresponding rainfall duration.

[0053] During the incremental correction process, an incremental sequence is formed, consisting of the initial dynamic critical rainfall and several fixed rainfall increments superimposed.

[0054] When the disaster criterion changes from being unmet to being met for the first time in the increasing sequence, the corresponding current round of rainfall value is determined as the minimum effective critical rainfall under the current soil moisture saturation and the current rainfall duration.

[0055] In other words, the minimum effective critical rainfall is the rainfall amount that, for the first time in an increasing sequence, changes the disaster criterion from being unsatisfied to being satisfied.

[0056] S8. Write the boundary fitting formula parameters and minimum effective critical rainfall value corresponding to different humidity levels and different rainfall durations into the dynamic critical rainfall threshold sample library. Based on the dynamic critical rainfall threshold sample library and real-time rainfall data, determine the risk of the target disaster-bearing object.

[0057] The dynamic critical rainfall threshold sample library is used to directly call the corresponding boundary fitting formula parameters and minimum effective critical rainfall under the same humidity level and the same rainfall duration.

[0058] The boundary fitting formula parameters and minimum effective critical rainfall values ​​corresponding to different humidity levels and rainfall durations are entered into a dynamic critical rainfall threshold sample library. The sample library is indexed according to a four-element mapping relationship of "humidity level - rainfall duration - boundary fitting formula parameters - minimum effective critical rainfall".

[0059] In subsequent operational applications, when watersheds are at the same humidity level and correspond to the same rainfall duration, the boundary fitting formula parameters and minimum effective critical rainfall from the threshold sample library can be directly invoked to improve real-time early warning efficiency. Based on the dynamic critical rainfall threshold sample library and real-time rainfall data, risk assessment is performed on target disaster-prone objects. The target disaster-prone objects may include at least one of the following: residential areas, roads, bridges, river floodplains, and low-lying flood-prone areas.

[0060] The risk assessment results include the minimum effective critical rainfall value, information on the distribution of dangerous areas, and information on the risk level of the target disaster-bearing objects.

[0061] A system for determining dynamic critical rainfall in mountainous watersheds includes: The runoff generation and confluence model construction unit is used to acquire rainfall, evaporation, underlying surface and historical flood data of the study watershed, construct and calibrate the runoff generation and confluence model, and output the actual soil moisture content of the watershed at the current moment; The wet state normalization unit is used to calculate the soil moisture saturation based on the actual soil moisture content and soil saturation moisture content output by the runoff generation and confluence model. The boundary point set extraction unit is used to construct historical over-alert samples and historical non-over-alert samples respectively, and extract the critical boundary point set for each rainfall duration based on the historical over-alert samples and historical non-over-alert samples for multiple preset rainfall durations respectively; The explicit formula generation unit is used to fit the critical boundary point set for each rainfall duration, generate the explicit formula for the dynamic critical rainfall for each rainfall duration, and calculate the initial dynamic critical rainfall for different rainfall durations. The incremental iterative correction unit is used to call the two-dimensional hydrodynamic model when the rainfall reaches the dynamic critical rainfall amount under the corresponding rainfall duration, and to incrementally correct the dynamic critical rainfall amount with a fixed rainfall increment according to the preset disaster criterion, until the disaster criterion changes from not being met to being met for the first time in the incremental sequence. The threshold sample library storage unit is used to store different humidity levels, different rainfall durations, boundary fitting formula parameters, and minimum effective critical rainfall values; The disaster-bearing object risk assessment unit is used to assess the risk of target disaster-bearing objects based on a dynamic critical rainfall threshold sample library and real-time rainfall data.

[0062] This invention introduces soil moisture saturation to achieve a normalized characterization of the early-stage wet state of the watershed, enabling the critical rainfall to be dynamically adjusted according to changes in the wet state. This invention achieves direct calculation of dynamic critical rainfall for different durations by extracting critical boundary point sets for different rainfall durations and fitting explicit formulas for each set. This invention uses a two-dimensional hydrodynamic model and a fixed rainfall increment sequence to gradually correct the dynamic critical rainfall, so that the final determined minimum effective critical rainfall is more consistent with the actual disaster-causing conditions. This invention constructs a dynamic critical rainfall threshold sample library, which allows for direct retrieval of corresponding results under the same humidity level and rainfall duration, thereby improving the efficiency of real-time business applications. This invention can construct explicit formulas for dynamic critical rainfall under different rainfall durations based on the current soil moisture state of the watershed, and determine the minimum effective critical rainfall by incrementally correcting the two-dimensional hydrodynamic simulation results with a fixed rainfall increment, thereby improving the consistency between the critical rainfall results and the actual disaster-causing conditions.

[0063] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the invention by those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the invention should be included within the scope of protection of the invention.

Claims

1. A method for determining the dynamic critical rainfall in a mountainous watershed, characterized in that: Includes the following steps: S1. Obtain rainfall, evaporation, underlying surface and historical flood data for the study watershed, construct and calibrate the runoff generation and confluence model, and output the actual soil moisture content of the watershed at the current moment; S2. Based on the soil saturation water content, the actual soil water content of the watershed at the current moment is normalized to obtain the soil water saturation, and the soil water saturation is divided into multiple humidity levels. S3. Construct historical samples of exceeding the warning level and historical samples of not exceeding the warning level respectively, and extract the critical boundary point set for each rainfall duration based on the historical samples of exceeding the warning level and historical samples of not exceeding the warning level respectively; S4. Fit the critical boundary point set for each rainfall duration to generate the explicit formula for the dynamic critical rainfall for each rainfall duration, and calculate the initial dynamic critical rainfall for different rainfall durations. S5. When the real-time rainfall or forecast rainfall reaches the initial dynamic critical rainfall for the corresponding rainfall duration, the corresponding rainfall process is input into the two-dimensional hydrodynamic model to calculate the maximum water depth and maximum flow velocity of each calculation unit in the target area. S6. Compare the output results of the two-dimensional hydrodynamic model with the preset disaster criteria. When the current simulation results do not meet the disaster criteria, the dynamic critical rainfall under the current rainfall duration is incrementally corrected by a fixed rainfall increment. S7. Repeat the two-dimensional hydrodynamic simulation to form an incremental sequence consisting of the initial dynamic critical rainfall and several fixed rainfall increments, until the disaster criterion changes from not being met to being met for the first time in the incremental sequence. The corresponding current round of rainfall value is determined as the current soil moisture saturation and the minimum effective critical rainfall under the current rainfall duration. S8. Write the boundary fitting formula parameters and minimum effective critical rainfall value corresponding to different humidity levels and different rainfall durations into the dynamic critical rainfall threshold sample library. Based on the dynamic critical rainfall threshold sample library and real-time rainfall data, determine the risk of the target disaster-bearing object.

2. The method for determining the dynamic critical rainfall in a mountainous watershed according to claim 1, characterized in that, The method for calibrating the fixed production confluence model in step S1 includes: constructing a penalty function to describe the degree of violation of the constraints by the parameter combination, and constructing a fitness function to optimize the parameter combination.

3. The method for determining the dynamic critical rainfall in a mountainous watershed according to claim 1, characterized in that, Methods for extracting the critical boundary point set for each rainfall duration from multiple preset rainfall durations include: For each preset rainfall duration, extract the boundary point set to output the critical boundary point set for the corresponding rainfall duration; Boundary point set extraction is performed using a boundary screening model, which is a support vector machine model.

4. The method for determining the dynamic critical rainfall in a mountainous watershed according to claim 1, characterized in that, In step S4, the least squares linear fitting method is used to fit the critical boundary point set for each rainfall duration.

5. The method for determining the dynamic critical rainfall in a mountainous watershed according to claim 1, characterized in that, In step S5, the governing equations of the two-dimensional hydrodynamic model include the continuity equation and the momentum equation.

6. The method for determining the dynamic critical rainfall in a mountainous watershed according to claim 1, characterized in that, In step S6, the disaster criterion may include: the maximum water depth in the target area is greater than or equal to a preset water depth threshold; the maximum flow velocity in the target area is greater than or equal to a preset flow velocity threshold; or the product of the maximum water depth and the maximum flow velocity in the target area is greater than or equal to a preset combined threshold.

7. The method for determining the dynamic critical rainfall in a mountainous watershed according to claim 1, characterized in that, In step S6, the fixed rainfall increment is an integer multiple of the preset rainfall resolution for the corresponding rainfall duration.

8. The method for determining the dynamic critical rainfall in a mountainous watershed according to claim 1, characterized in that, In step S8, the dynamic critical rainfall threshold sample library is indexed according to the four-element mapping relationship of "humidity level - rainfall duration - boundary fitting formula parameters - minimum effective critical rainfall". When the watershed is at the same humidity level and has the same rainfall duration, the boundary fitting formula parameters and minimum effective critical rainfall in the dynamic critical rainfall threshold sample library are directly called.

9. The method for determining the dynamic critical rainfall in a mountainous watershed according to claim 1, characterized in that, In step S8, the risk assessment results include the minimum effective critical rainfall value, the distribution information of dangerous areas, and the risk level information of the target disaster-bearing objects.

10. A system for determining dynamic critical rainfall in mountainous watersheds, characterized in that, include: The runoff generation and confluence model construction unit is used to acquire rainfall, evaporation, underlying surface and historical flood data of the study watershed, construct and calibrate the runoff generation and confluence model, and output the actual soil moisture content of the watershed at the current moment; The wet state normalization unit is used to calculate the soil moisture saturation based on the actual soil moisture content and soil saturation moisture content output by the runoff generation and confluence model. The boundary point set extraction unit is used to construct historical over-alert samples and historical non-over-alert samples respectively, and extract the critical boundary point set for each rainfall duration based on the historical over-alert samples and historical non-over-alert samples for multiple preset rainfall durations respectively; The explicit formula generation unit is used to fit the critical boundary point set for each rainfall duration, generate the explicit formula for the dynamic critical rainfall for each rainfall duration, and calculate the initial dynamic critical rainfall for different rainfall durations. The incremental iterative correction unit is used to call the two-dimensional hydrodynamic model when the rainfall reaches the dynamic critical rainfall amount under the corresponding rainfall duration, and to incrementally correct the dynamic critical rainfall amount with a fixed rainfall increment according to the preset disaster criterion, until the disaster criterion changes from not being met to being met for the first time in the incremental sequence. The threshold sample library storage unit is used to store different humidity levels, different rainfall durations, boundary fitting formula parameters, and minimum effective critical rainfall values; The disaster-bearing object risk assessment unit is used to assess the risk of target disaster-bearing objects based on a dynamic critical rainfall threshold sample library and real-time rainfall data.

Images