60 results about "Flash flood" patented technology

The invention discloses a rainstorm torrential flood disaster-causing element prediction method and system based on modal decomposition and deep learning, and solves the problems that a traditional model is insufficient in non-linear time sequence feature capture, and a physical model depends on complex data and is weak in generalization ability. Comprising the steps of collecting flow, flow velocity and water level data of an upstream site as input, and taking downstream disaster point data as output; preprocessing the data; a frost ice optimization algorithm is adopted to optimize variational mode decomposition parameters; a Fourier transform high and low frequency feature enhanced attention network is constructed, low-frequency and high-frequency components are divided, trend features are extracted through a fluctuation enhancement module, dynamic changes are captured through a multi-path difference calculation unit, self-perception attention is introduced to achieve feature weighted fusion and long-term memory, and a downstream hydrological state prediction result is output. By optimizing a modal decomposition and deep learning cooperation mechanism, the precision, robustness and generalization ability of sudden mountain torrent prediction are significantly improved, and the method is suitable for disaster emergency management in complex scenes of small and medium watersheds.

The invention discloses a deep learning-based less-data small-watershed water level forecasting method. The method comprises the following steps of: acquiring historical hourly rainfall data of a plurality of upstream rainfall stations of a small watershed and historical hourly water level data of a downstream water level station; preprocessing the acquired data, and screening rainfall sessions with the rainfall capacity of over 15mm per hour within 24 hours and corresponding water level data; the method comprises the following steps: performing data enhancement through a double-generation GAN model, generating simulated rainfall data by a first GAN model by taking white noise as input, generating simulated water level data by a second GAN model by taking rainfall data as input, combining historical data and simulated data, performing normalization, and dividing the data into a training set, a verification set and a test set according to a proportion; constructing a CNN-LSTM hybrid prediction model, training the model by taking a Nash coefficient as an optimization target, and verifying the prediction model by adopting a progressive verification strategy; and deploying the model for real-time water level prediction to obtain a water level prediction result. The method can break through the bottleneck of few data, and achieves the high-precision mountain torrent early warning of the small watershed water level.

A flash flood simulation method includes: obtaining a water storage in a vegetation root zone and obtaining an actual evaporation at any point in the vegetation root zone within a target watershed; obtaining a total infiltration rate in an unsaturated zone based on an infiltration model; obtaining a saturation overland flow and an interflow based on a vertical distance from a wetting front to a slope surface; updating an average saturated groundwater table depth based on the interflow and the total infiltration rate of the unsaturated zone; constructing a flash flood model for the target watershed based on the water storage of the vegetation root zone, the actual evaporation, the total infiltration rate of the unsaturated zone, the saturation overland flow, the interflow, and the average saturated groundwater table depth; and simulating an event-based flood hydrograph of the target watershed based on the flash flood model for the target watershed.

The invention discloses a method, a product and equipment for generating a mountain torrent disaster dynamic plan based on causal inference, and belongs to the technical field of mountain torrent disaster monitoring. The method comprises the following steps: S1, acquiring multi-source data, and extracting causal features of the multi-source data; during extraction, dual robust data deviation elimination is realized based on a tendency scoring model and a result regression model; s2, performing dynamic causal graph modeling based on the extracted causal features and a historical disaster data set, and updating the weight to obtain a causal graph with a time sequence weight; s3, determining a current disaster situation state according to the causal diagram information, and performing anti-fact plan generation in combination with a preset intervention measure set to obtain an optimal plan set; and S4, performing multi-department collaborative execution based on the optimal plan set. According to the method, the pain point problems of passive response, lack of causal support and cross-department collaboration low efficiency of a traditional mountain torrent plan can be solved.

The invention discloses a mountain torrent early warning rainfall threshold dynamic adjustment method and system, and relates to the technical field of hydrological early warning, and the method comprises the steps: obtaining basic rainfall data and a basic early warning threshold of a target region; according to the basic rainfall data, performing calculation to obtain a multi-factor correction factor; the multi-factor correction factors comprise an early-stage wettability correction factor, a rainfall intensity correction factor and a rain pattern mode correction factor; calculating to obtain a comprehensive correction factor according to the multi-factor correction factor; according to the comprehensive correction factor and the basic early warning threshold value, calculating to obtain a dynamic early warning threshold value; the dynamic early warning threshold value comprises a ready transfer dynamic early warning threshold value and an immediate transfer dynamic early warning threshold value; and determining a rainfall early warning level according to the dynamic early warning threshold and the real-time rainfall. According to the invention, the early warning threshold can be dynamically adjusted according to the real-time working condition, and the accuracy and timeliness of mountain torrent early warning are improved.

This invention discloses a flash flood and debris flow monitoring device based on phased array radar technology, comprising a base, a support plate, and a mounting plate. A column is rotatably connected to the top of the base, and a fixing groove is provided on the top of the base, within which a fixing mechanism corresponding to the column is fixedly connected. A support rod is slidably inserted into the top of the column, and the top of the support rod is fixedly connected to the bottom of the support plate. A support block is fixedly connected to one side of the column, and the support block is connected to the support plate via turnbuckles. A fixing rod is rotatably connected to the top of the support plate via a rotating rod, and the top of the fixing rod is fixedly connected to the bottom of the mounting plate. A sliding groove is provided on the top of the support plate, within which a screw is rotatably connected. A slider is threaded onto the screw, and the slider is connected to the mounting plate via a transmission rod. This invention, through the setting of multiple adjustment mechanisms, allows the monitoring device to be freely adjusted according to actual installation needs, thereby effectively improving the installation and usage effect of the monitoring device.

This invention belongs to the field of flash flood and debris flow disaster prevention and mitigation, and provides a method for simulating flash flood and debris flow processes based on rheological evolution characteristics. The method includes the following steps: establishing a generalized numerical model of the target channel; using a CFD-DEM coupled model to simulate the transport process of sediment particles in the generalized numerical model under the action of water flow or debris flow slurry; and, based on the semi-empirical relationship between the drag coefficient and Reynolds number of sediment particles carried by flash floods and debris flows, derived in this invention, realizing real-time adjustment of rheological parameters during the simulated flash flood and debris flow transport process through the bidirectional coupling mechanism of the CFD-DEM coupled model. This invention's method can more realistically simulate the spatial distribution of sediment particles during flash flood and debris flow transport, increasing the accuracy of flash flood and debris flow process simulation, and solving the problem that existing technologies struggle to simulate the rheological evolution phenomenon of flash floods and debris flows caused by changes in volumetric sediment content.

The invention discloses a mountain torrent disaster early warning method, device and equipment and a medium, and relates to the technical field of hydrological monitoring and disaster early warning. The method comprises the steps of determining theoretical confluence time according to static attributes of a target drainage basin, determining an observation window of a rainfall event according to the theoretical confluence time, and performing sliding window analysis on a rainfall process of the rainfall event by using the observation window to obtain accumulated rainfall in the observation window, according to runoff reduction analysis, the accumulated rainfall is converted into effective peak flow, the effective peak flow is compared with disaster-causing flow, and early warning is given out according to a comparison result. According to the early warning method, the observation window is associated with the flow concentration time of the drainage basin, and the peak rainfall related to the flow concentration time of the drainage basin is captured based on the accumulated rainfall counted by the observation window, so that the rainfall process is prevented from being split by statistics in a fixed time period, and the early warning effectiveness is improved.

The application discloses a mountain torrent induced landslide disaster risk early warning method and device and a storage medium, and relates to the technical field of geological disaster meteorological risk early warning.

The invention relates to the technical field of disaster early warning processing, in particular to a mountain torrent disaster early warning system based on deep learning and air-space-ground integrated perception. The system comprises the following modules: a space-air-ground sensing module, an edge calculation preprocessing module, a deep learning prediction module and an intelligent decision and early warning module. Wherein the air-space-ground sensing module is used for collecting multi-source data; wherein the multi-source data comprises earth surface deformation data obtained through a spaceborne synthetic aperture radar, high-precision topographic data obtained through an unmanned aerial vehicle laser radar, hydrological sensing data obtained through a ground internet-of-things distributed sensor, and meteorological numerical forecasting data obtained through a meteorological data interface. Through air-space-ground integrated perception, physical constraint deep learning and edge intelligence technologies, mountain torrent multi-source data collaborative acquisition and accurate mountain torrent disaster prediction are realized, so that the comprehensiveness, reliability and timeliness of mountain torrent disaster early warning are improved.

The invention discloses a mountain torrent disaster risk avoiding guidance and emergency command system and method based on edge computing, and relates to the technical field of mountain torrent disaster dispatch, in the method, an edge disaster sensing and gathering module gathers disaster dynamic data and positioning information on a target monitoring mountainous area map, and a comprehensive disaster situation map is constructed; an intelligent risk-avoiding path planning module automatically plans a risk-avoiding path by using a dynamic cost grid algorithm and an A path-finding algorithm according to the comprehensive disaster situation map, and generates a path instruction; a self-adaptive emergency command decision module automatically matches a preset emergency plan to generate a command instruction according to the real-time submerging range, the early warning level and the affected population density of each disaster area; and the multi-mode guide instruction issuing module sends a path instruction and a command instruction. According to the invention, the efficiency and rationality of mountain torrent disaster risk avoiding guidance and emergency command can be improved.

This application discloses a method, apparatus, and equipment for classifying flash flood disasters, relating to the field of disaster monitoring technology. The method includes: acquiring a geographic feature dataset of multiple watershed nodes in a target area, and constructing a feature matrix based on the geographic feature dataset; inputting the feature matrix into a flash flood disaster classification model, and outputting the classification probability of each watershed node belonging to each flash flood disaster category; the flash flood disaster classification model is used to: calculate the edge weight matrix of the graph structure formed by the feature matrix and the watershed nodes in the target area using a multi-head attention mechanism; perform sparsification processing on the edge weight matrix, and perform graph convolution operation based on the feature matrix and the sparsified edge weight matrix to obtain aggregated features; perform global feature enhancement on the aggregated features to obtain enhanced features; and output the classification probability of each watershed node belonging to each flash flood disaster category based on the enhanced features. This application can improve the accuracy of flash flood disaster classification.

This application discloses a dynamic decision-making method and system for flash floods and debris flows based on multi-agent and LLM (Limited Language Model), belonging to the field of emergency management and intelligent decision-making technology. This method constructs an intelligent decision-making architecture based on a large model and multi-agent collaboration. It solves the execution challenges of complex disaster tasks through the division of labor and cooperation among multiple agents, and overcomes the decision-making limitations of traditional systems by leveraging the deep reasoning capabilities of a large language model. Specifically, the multi-agent collaborative framework decomposes disaster prevention and control decisions into sub-tasks such as intent parsing, contingency plan retrieval, spatial analysis, data acquisition, and analysis and prediction. Each functional agent focuses on its area of ​​expertise, improving the efficiency and accuracy of task execution. The agent communication protocol ensures efficient information interaction and collaborative scheduling between modules, realizing fully automated processing from task input to data output.

The application discloses a kind of mountain torrent risk prediction big model's partition classification parallel training method, the method includes the following steps: step 1, mountain torrent disaster sample big data construction;Step 2, mountain torrent disaster is classified and input-output strategy is formulated;Step 3, partition classification carries out big model parallel training;Step 4, big model input strategy adjustment;Step 5, interactive automatic optimization big model parameter.The method disclosed in the application improves the generalization ability and prediction accuracy of mountain torrent risk prediction big model by partition classification of sample big data;Proposed mountain torrent risk prediction big model parallel training method improves training efficiency;Using the characteristics of big model back propagation, calculate the relative contribution of input elements to mountain torrent disaster, guide to simplify input elements, on the basis of ensuring prediction accuracy, light weight mountain torrent risk prediction big model.

This invention relates to the field of cultivation technology for specific crops or plants not mentioned above, and discloses a method for creating vegetation on alluvial and rockfall riverbanks. The method involves identifying weak erosion zones less affected by unstable landslides and alluvial deposits left by flash floods through exploration. Perennial vines are then planted at these locations to cover the remaining riverbank area, preventing damage to the vegetation from alluvial deposits and fallen rocks. Microenvironmental wells are dug within the weak erosion zones, and the perennial vines are planted inside to block flash floods from damaging the vegetation and surrounding soil. The soil within the wells is rotated, with topsoil containing a large amount of sand and gravel, being infertile and unable to retain water, being turned to the bottom of the well, thus no longer hindering vegetation growth and facilitating groundwater infiltration. The deeper soil, free of sand and gravel, relatively fertile, and with better water retention, is turned to the top layer, allowing the vegetation within the well to grow better. The combination of these methods overcomes the damage and stress to vegetation, enabling the creation of vegetation on riverbanks where normal growth is impossible.

An intelligent early warning and automatic triggering type emergency linkage system for small watershed mountain torrents of the present invention belongs to the technical field of mountain torrent disaster monitoring and early warning, and comprises a data acquisition and processing module, a risk assessment module, an early warning judgment module and an emergency response triggering module. According to the method, early warning can be given out earlier, precious time is won for disaster prevention and reduction, it is ensured that emergency resources are put to the most needed place with the optimal path in the shortest time, the response efficiency is remarkably improved, the system can continuously learn from practice, the early warning and response performance of the system is continuously improved, and the continuous self-evolution ability is achieved.

The invention relates to the technical field of mountain torrent monitoring, in particular to a mountain torrent disaster real-time monitoring method and system integrating radar and video. Comprising the following steps: respectively acquiring a water velocity signal and a mountain torrent region image sequence through radar equipment and video equipment, and aligning the frequency difference of acquired data through a timestamp calibration mechanism to form a multi-source data set with preliminarily aligned time and space; motion feature vectors and boundary contour features are extracted from the data set, filtering smoothing processing is carried out on the motion features, and noise is removed; correcting timestamp offset, ensuring data time unification, and associating motion features with boundary contour features through space mapping to generate a fusion feature map; calculating the corresponding relation between the water velocity and the boundary contour through the fusion feature map, generating a dynamic hydrological parameter distribution map, marking a high-risk disaster range, generating a comprehensive monitoring report, and determining the mountain torrent disaster state in real time. According to the invention, the real-time monitoring precision and early warning capability of mountain torrent disasters are effectively improved.

The invention discloses a mountain torrent disaster comprehensive risk prediction method and system based on multi-source data, and relates to the technical field of mountain torrent disaster risk prediction.The method comprises the steps of multi-source data acquisition, multi-source data analysis and mountain torrent disaster prediction feedback, the multi-source data is accurately acquired, then the state corresponding to a region is evaluated, and the risk prediction accuracy is improved. And predicting whether the mountain torrent disaster occurs in the region and the disaster level, analyzing to obtain early warning maintenance of the mountain torrent disaster corresponding to the region, executing negative hidden danger cost pre-simulation of the adopted maintenance set under the mountain torrent disaster risk level, and evaluating to obtain the cost of the adopted early warning maintenance of the mountain torrent disaster corresponding to the region. And adaptive early warning maintenance under the corresponding mountain torrent disaster of the region is obtained through screening, so that accurate prediction of the corresponding mountain torrent disaster risk of the region is realized, effective measures are taken for prevention of a prediction result in time, and dangerous influence and risk loss caused by the mountain torrent disaster are reduced.

This invention discloses a method for early warning of flash floods in important villages based on small reservoir hydrological and rainfall monitoring. The method first collects data from the study area, constructing a multi-source heterogeneous data fusion system and a hydrological-topographic-socioeconomic correlation network for the study area. Then, it constructs spatial location and water flow path correlations between small reservoirs and important villages, generating inundation range maps and delineating risk zones based on the topography, building layout, and set water level thresholds of the important villages. A reservoir-river-village hydrodynamic coupling model is constructed to calculate the correlation between small reservoir monitoring indicators and village disaster events, and this is input into a disaster-causing water level inversion model to obtain the probability distribution of disaster-causing water levels. Finally, a flash flood disaster early warning model is constructed, extracting the probability distribution of disaster-causing water levels as model input to perform multi-timescale risk forecasting and classify early warning levels, developing differentiated early warning schemes for different levels. This improves the scientific rigor and effectiveness of flash flood disaster prevention.

The invention relates to the technical field of watershed hydrological model parameter calibration, in particular to a hydrological model parameter calibration method for a watershed lacking hydrological data, which comprises the following steps: determining a watershed A lacking data and a watershed B controlled by an adjacent hydrological station (reference station) with data, and collecting hydrometeorological and underlying surface feature data of the two watersheds; calculating the design flood of the drainage basin A and the drainage basin B by using a rainstorm runoff survey algorithm; comparing the design flood calculated by the rainstorm runoff of the watershed B close to the data with the design flood calculated by the actually measured flow data, and determining a flood correction coefficient kappa for correcting the design flood calculated by the watershed A lacking the data by using the rainstorm runoff calculation algorithm; inputting the model parameters of the data-lacking watershed A and the design surface rainstorm into a hydrological model; simulating an outlet section flood; and comparing a simulation result with a corrected rainstorm runoff checking result, analyzing parameter sensitivity, and adjusting and optimizing model parameters. The method is suitable for parameter calibration of the data-lacking small-watershed mountain torrent forecasting model, and a new means is provided for flood control and disaster reduction.

The invention discloses a traffic engineering premises mountain torrent risk early warning analysis method based on multi-source perception, and relates to the technical field of mountain torrent disaster monitoring and emergency management, and the method comprises the steps: carrying out the mountain torrent audio feature recognition of sound pressure signals based on a pre-trained AI model, and carrying out the water body form mutation recognition of video streams; in the far-end server, multi-source sensor data and weather forecast data are fused, weights are distributed to the sensors according to the topographic and geological risk levels of the river reach where the sensors are located, and the comprehensive risk score of the upstream river reach is calculated; predicting a flood routing path, a flood routing speed and a flood range based on the upstream comprehensive risk score and a hydraulic model; dynamically planning an optimal escape path according to the prediction result and the real-time acquired position information of the stationed personnel; according to the method, the upstream fusion decision model based on distance attenuation and river reach weight is adopted, so that the downstream station can sense the threat from the upstream early and accurately.

The invention discloses a mountain torrent disaster early warning information broadcasting terminal, and relates to the technical field of mountain torrent disaster early warning, and the terminal comprises an information receiving unit which analyzes, collects and verifies multi-source data; the AI operation processing unit classifies and sorts the data, predicts a trend, judges an early warning level and optimizes pushing; the display output unit converts the data into three-dimensional content and displays the three-dimensional content in multiple forms; the sound-light alarm unit adjusts sound and light and broadcasts according to grades; the user interaction unit records behavior data and supports personalized setting, and all the units cooperate to achieve intelligent early warning; according to the invention, multi-source data is processed through the information receiving unit, and the energy quantification and coupling prediction model of the AI unit is combined, so that the dependence of a traditional single data source is broken through, a full-chain intelligent decision is formed, and the prediction accuracy and the grade division scientificity are improved; meanwhile, the early warning form is adjusted through an energy threshold adaptation algorithm, three-dimensional rendering and grading acousto-optic technologies are integrated, the understanding threshold is reduced through multi-modal interaction, the risk avoiding decision time is shortened, and the disaster prevention efficiency is improved.

This invention discloses a deep learning-based automatic flash flood identification and real-time early warning system, comprising: a data acquisition and quality control module for acquiring precipitation and topographic data to generate basic observation data; a precipitation correction and fusion module for generating a rasterized precipitation field and uncertainty metric; a watershed topology construction module for filling depressions and connecting depressions in the topographic data to construct the watershed topology; a spatiotemporal feature generation module for aligning the rasterized precipitation field with the watershed topology to obtain a spatiotemporal feature sequence; a depth prediction module for inputting the spatiotemporal feature sequence into a spatiotemporal deep learning model to generate depth prediction results; a hydrodynamic simulation module for setting initial and boundary conditions on an unstructured grid and outputting hydrodynamic simulation results; and a coupling correction and early warning release module for updating predictions and boundaries, generating tiered early warnings, and archiving them. This invention achieves automatic flash flood identification and real-time early warning.

The present disclosure relates to a flash flood disaster refuge transfer path planning method, device and system. The method comprises: obtaining to-be-transferred position information and target transfer position information; determining a to-be-planned path region based on the to-be-transferred position information and the target transfer position information, and taking a node corresponding to the to-be-transferred position information as a starting point and a node corresponding to the target transfer position information as an ending point; performing grid division on the to-be-planned path region to obtain a plurality of grids and a plurality of nodes corresponding to the plurality of grids respectively; the plurality of nodes include the starting point and the ending point; based on an A* algorithm, a preset cost estimation function is used to calculate a cost estimation value of a candidate node in the plurality of nodes; and a planning path is generated by using the cost estimation value of each candidate node, the starting point and the ending point. The present scheme can quickly and accurately complete the planning of the transfer refuge path.

The utility model relates to the technical field of mountain torrent monitoring, in particular to a mountain torrent monitoring device which comprises a communication module and a monitoring module electrically connected with the communication module, and further comprises a fixing rod and a water storage tank fixedly connected to the fixing rod, the fixing rod is vertically fixed to the ground, and the water storage tank is provided with a rain inlet with an upward opening. The monitoring module is fixedly connected to the fixing rod and located on the upper side of the rain inlet, and the monitoring module monitors the change of the water level in the water storage tank; according to the mountain torrent monitoring device, through cooperation of the water storage tank and the monitoring module, the rainfall within a period of time is monitored; by monitoring the operation state of the monitoring plate, comprehensive monitoring of the rainfall and the river flow state is further formed.

This invention discloses a method for simulating complex slope runoff based on multi-source data fusion. It includes: fusing UAV LiDAR, ground-penetrating radar, and remote sensing images, and considering physical coupling relationships such as vegetation root systems, to generate a high-precision and physically consistent initial surface parameter field; identifying micro-topographic depressions using graph cutting algorithms and constructing a dynamically evolving slope overflow network based on a depression-filling-overflow mechanism; establishing a dynamic feedback control mechanism of overflow-scour-infiltration: the scour intensity calculated based on overflow data is used to correct the soil saturated hydraulic conductivity in real time, and the corrected parameters are then fed back into an improved Green-Ampt infiltration model; and outputting the spatiotemporal distribution of slope runoff through multi-timescale algorithm coupling, which can trigger early warning and invert critical rainfall conditions. This invention can simulate the dynamic feedback process of the underlying surface, improve simulation reliability, and provide technical support for flash flood early warning.

This invention provides a structure for flood control, aiming to address the challenges of impact damage, energy dissipation and deceleration, and smooth transition between mountain flood interception ditches and urban drainage systems. The invention mainly includes a cascading water system, a gradually widening section, an energy dissipation pool, a grating, a retaining wall, a slope toe, connecting pipes, ground level, inspection wells, and an energy dissipation sill. Through a three-stage energy dissipation system (cascading water system + energy dissipation pool + energy dissipation sill), the water flow velocity is significantly reduced. The grating intercepts debris and filters wastewater before it is channeled into the urban pipe network. A liquid level sensor triggers an alarm, providing sufficient emergency response time for wastewater treatment plants. This modular structure is suitable for narrow mountain sites, with a short construction period and high practicality.

The application discloses a mountain torrent disaster early warning method and system, and the method comprises the following steps: acquiring cloud image data; processing the cloud image data based on a pre-trained deep learning model to obtain a predicted rainfall sequence; constructing a hydrological and hydrodynamic model and performing scenario simulation to demarcate a disaster threshold rainfall; and combining the predicted rainfall sequence and the disaster threshold rainfall to perform early warning. The system comprises an image acquisition module, a rainfall prediction module, a threshold determination module and an early warning module. By using the application, mountain torrent disaster early warning can be performed in advance, important disaster relief and risk avoidance time can be obtained for mountain torrent disasters, and the prediction accuracy is well guaranteed due to the combination of the rainfall prediction module and the atmospheric pressure observation and the cloud rainfall estimation method based on deep learning. The application can be widely applied in the field of natural disaster early warning.

This invention provides a dynamic prediction method for flash floods based on satellite remote sensing multi-source data fusion and a spatiotemporal Transformer, including multi-source data acquisition and fusion, feature extraction and dimensionality reduction, model construction and training, and flash flood risk assessment and prediction. This invention deeply integrates advanced data fusion algorithms and a spatiotemporal Transformer network architecture to efficiently process multi-source heterogeneous data such as rainfall distribution, topography, and vegetation cover acquired by satellite remote sensing. The spatiotemporal Transformer network, with its powerful attention mechanism, can accurately capture the coupling relationships and dynamic evolution patterns of various influencing factors in the spatiotemporal dimension during the occurrence and development of flash floods. Compared with existing technologies based on satellite remote sensing and deep learning, this method significantly improves the prediction accuracy of flash floods, providing more reliable data support for disaster early warning and emergency decision-making.

This invention discloses a method for analyzing the formation and disaster-causing mechanism of flash floods based on disaster-prone environments, comprising the following steps: Step 1, constructing a set of historical flash flood disaster attributes and environmental factors for the study area; Step 2, determining the zoning of flash flood disaster-prone environments and their characteristics; Step 3, determining the explanatory power and uncertainty range of environmental categories and their environmental factors for changes in flash flood disaster attributes in each zoning; Step 4, determining the dominant environmental factors in each disaster-prone environment zoning; Step 5, analyzing the impact of changes in dominant environmental factors on the flash flood formation and disaster-causing process. The method of this invention fully considers the spatiotemporal heterogeneity of meteorological, geographical, and socioeconomic factors in the affected watershed, and separates and quantifies the contributions and uncertainties of different environmental factors such as meteorology, geography, and socioeconomic factors to flash flood formation and disaster-causing factors. It can improve the comprehensive prevention and control level of flash flood disasters, has strong applicability, and provides a scientific basis for flash flood disaster prevention and management.