Mountain area river flood dynamic rehearsal method, device and equipment based on big data fusion and medium

Abstract

The invention relates to a mountain river flood dynamic rehearsal method, device and equipment based on big data fusion and a medium. The method comprises the following steps: acquiring a multi-source fusion data set of a target mountain river; based on the multi-source fusion data set, recognizing the flood type of the target mountain river in combination with a preset threshold rule to obtain the flood type and cause parameters; according to the flood type and the cause parameter, calculating a design flood peak flow value of each node in the target mountain river; according to each design flood peak flow value, performing section-by-section trial calculation through an energy equation to obtain a design flood surface line; and performing dynamic simulation processing on the flood routing process of the target mountain river based on the designed flood water surface line, and generating a flood routing dynamic simulation report of the target mountain river. The method can accurately adapt to the characteristics of diverse causes and complex propagation process of flood in a mountain river, and remarkably improves the refinement degree and reliability of flood rehearsal.

Description

Technical Field

[0001] This invention belongs to the field of flood simulation and prediction technology, and in particular relates to a method, device, equipment and medium for dynamic flood prediction of mountain rivers based on big data fusion. Background Technology

[0002] Dynamic flood forecasting for mountain rivers is a crucial technical aspect of flood control and disaster reduction. Its core lies in comprehensively analyzing and forecasting flood formation, evolution, and inundation processes using multi-dimensional information from hydrology, meteorology, and topography. In recent years, with advancements in remote sensing technology, IoT monitoring, and meteorological forecasting capabilities, flood forecasting methods based on big data fusion have gradually become a research hotspot. This method aims to integrate multi-source heterogeneous data to construct more refined flood models, thereby improving the accuracy of predictive capabilities and decision support for flood disasters in complex mountainous environments.

[0003] However, existing flood forecasting methods still face significant challenges in practical applications. On the one hand, most methods are fragmented at the data level, failing to achieve deep coupling and dynamic assimilation between hydrological, meteorological, remote sensing, and river topographic data. This results in incomplete model input information or mismatched spatiotemporal scales, affecting the reliability of the forecasting foundation. On the other hand, existing methods rely heavily on static, fixed parameters and model structures, making it difficult to adapt to the diverse causes of floods in mountainous rivers (such as snowmelt, rainstorms, and mixed types) and the complex and ever-changing propagation processes. Consequently, the forecasting results often lack accurate reflection of flood type characteristics and dynamic evolution paths, making it difficult to effectively support the refined needs of disaster early warning and emergency response.

[0004] Therefore, there is a need for a flood prediction method for mountain rivers that can deeply integrate multi-source data and effectively couple the identification of flood causes with the dynamic evolution process. Summary of the Invention

[0005] Therefore, it is necessary to provide methods, devices, equipment, and media for dynamic flood prediction in mountainous rivers based on big data fusion to address the aforementioned technical issues.

[0006] Firstly, this application provides a method for dynamic flood prediction of mountain rivers based on big data fusion, including:

[0007] Multi-source basic data of rivers in the target mountainous area were acquired and fused to obtain a multi-source fused dataset. The multi-source data included hydrological station monitoring data, climate data, remote sensing data, and river topography data.

[0008] Based on a multi-source fusion dataset, the flood types of mountain rivers are identified by a preset threshold rule, and the flood type identification results and causal parameter set corresponding to the target mountain rivers are obtained.

[0009] Based on the multi-source fusion dataset, flood type identification results, and causal parameter set, the design peak flow corresponding to the target mountain river is calculated, and the design peak flow value of each node is obtained; the node is a key location with clear topographic features or hydrological monitoring function.

[0010] Based on the design peak flow value of each node, the design flood water surface line of the target mountain river is obtained by segmenting calculations through the energy equation.

[0011] Based on the design flood water surface line, the flood evolution process of the target mountain river is dynamically simulated to obtain a dynamic simulation report of the flood evolution of the target mountain river.

[0012] Furthermore, based on the multi-source fusion dataset, flood types in mountainous rivers are identified using preset threshold rules, yielding flood type identification results and a set of causal parameters for the target mountainous rivers, including:

[0013] Flood feature vectors are extracted from multi-source fusion datasets; the flood feature vectors include rainfall intensity, warming rate, and snowmelt rate.

[0014] Based on the feature values ​​in the flood feature vector, the flood type of mountain rivers is dynamically identified through a preset threshold rule, and the flood type identification result corresponding to the target mountain river is obtained; the flood types of mountain rivers include ablation type, rainstorm type, and rain-snow mixed type;

[0015] Based on the flood type identification results, corresponding flood causation-related parameters are extracted from the multi-source fusion dataset; these parameters are those that can directly affect the formation of the corresponding flood type.

[0016] By combining and processing various flood-related parameters, a set of causal parameters is obtained.

[0017] Furthermore, based on the multi-source fusion dataset, flood type identification results, and causal parameter set, the design peak flow corresponding to the target mountain river is calculated, and the design peak flow value for each node is obtained, including:

[0018] Based on the flood type identification results and the set of causal parameters, the corresponding frequency analysis parameters are determined;

[0019] Based on frequency analysis parameters, the flood design frequency of the target mountain river is calculated;

[0020] Extract the annual maximum flood peak flow sequence of each hydrological station from the monitoring data of the hydrological stations;

[0021] Statistical processing was performed on the annual maximum flood peak flow series to obtain the mean and standard deviation of each annual maximum flood peak flow series;

[0022] The coefficient of variation for each hydrological station was obtained based on the mean and standard deviation.

[0023] Based on the values ​​of the coefficients of variation and the flood design frequency, the initial design peak flood discharge for each hydrological station is calculated using the following formula:

[0024]

[0025] in, Let p be the initial design peak flow rate, and p be the flood design frequency. The average peak flow rate. This is the deviation coefficient corresponding to the design frequency of the flood. This represents the coefficient of variation.

[0026] The river channel attenuation characteristic parameters are obtained, and the initial design peak flow is corrected for attenuation along the course based on the river channel attenuation characteristic parameters to obtain the design peak flow value of each node.

[0027] Furthermore, based on the design peak flow values ​​at each node, the design flood water surface line of the target mountain river is obtained through segmented calculations using the energy equation, including:

[0028] The hydrological station flow measurement section downstream of the target mountain river is determined as the starting point for the design flood water surface line;

[0029] The measured water level at the corresponding flood design frequency is extracted from the monitoring data of the hydrological station as the initial water level of the starting node;

[0030] Based on the initial water level of the starting node and the design peak flow value corresponding to the starting node, the design flood level of each node is obtained by performing segment-by-segment calculations from downstream to upstream using Bernoulli's energy equation.

[0031] Connect the design flood levels of each node to generate the design flood water surface line of the target mountain river.

[0032] Furthermore, based on the initial water level of the starting node and the corresponding design peak flow value, Bernoulli's energy equation is used to perform segment-by-segment calculations from downstream to upstream to obtain the design flood level of each node, including:

[0033] Based on the initial water level of the starting node and the corresponding design peak flow value, the design flood level of each node is obtained by performing segment-by-segment calculations from downstream to upstream using Bernoulli's energy equation:

[0034]

[0035]

[0036]

[0037] in, To design flood level, Here, dv represents the design flood level of the upstream node, and dv represents the change in flow velocity. This is the kinetic energy correction factor. and Let g be the flow velocity and g be the acceleration due to gravity. Let η be the head energy loss, η be the riverbed roughness, L be the node spacing, R be the hydraulic radius, and A be the cross-sectional area of ​​the water passage.

[0038] Furthermore, based on the design flood water surface line, the flood evolution process of the target mountain river is dynamically simulated to obtain a dynamic simulation report of the flood evolution of the target mountain river, including:

[0039] Based on the design peak flow rate and river topography data at each node, the peak propagation velocity is calculated using hydraulic formulas.

[0040] Based on the propagation speed of the flood peak and the length of the river channel between the two sections, the propagation time of the flood peak is calculated.

[0041] Using the time when the flood peak arrives at the initial section as a benchmark, the flood peak propagation time of each segment is accumulated to obtain the flood peak propagation time series;

[0042] Based on the flood peak propagation time series, the design flood water surface line is spatiotemporally connected to obtain spatiotemporally connected water surface line data;

[0043] Using animation rendering technology, the spatiotemporally connected water surface data is used to generate a flood evolution animation;

[0044] By overlaying the design flood water surface line with the digital elevation model, the flood inundation depth and extent can be obtained.

[0045] A flood risk level distribution map is generated based on the flood inundation depth and extent.

[0046] Integrate flood evolution animations and flood risk level distribution maps to generate dynamic simulation reports of flood evolution for rivers in the target mountainous areas.

[0047] Furthermore, the multi-source basic data is fused to obtain a multi-source fused dataset, including:

[0048] Missing values ​​in the hydrological station monitoring data from the multi-source basic data were filled to obtain a continuous and complete hydrological data sequence.

[0049] Climate data and remote sensing data from multi-source basic data are processed into grids to obtain a raster dataset with uniform spatial resolution;

[0050] The initial multi-source fusion dataset is obtained by performing spatiotemporal registration processing on continuous and complete hydrological data sequences, raster datasets with uniform spatial resolution, and river topographic data.

[0051] The initial multi-source dataset is format-standardized to generate a multi-source fused dataset.

[0052] Secondly, this application also provides a dynamic flood prediction device for mountain rivers based on big data fusion, including:

[0053] The data acquisition module is used to acquire multi-source basic data of rivers in the target mountainous area and to fuse the multi-source basic data to obtain a multi-source fused dataset; the multi-source data includes hydrological station monitoring data, climate data, remote sensing data and river topography data;

[0054] The flood type identification module is used to identify the flood type of mountain rivers based on a multi-source fusion dataset and through preset threshold rules, so as to obtain the flood type identification result and the set of causal parameters corresponding to the target mountain river;

[0055] The design module for calculating peak flood discharge value is designed to calculate the design peak flood discharge value of the target mountain river based on the multi-source fusion dataset, flood type identification results and causal parameter set, and to obtain the design peak flood discharge value of each node; the node is a key location with clear topographic features or hydrological monitoring function.

[0056] The design flood water surface line calculation module is used to obtain the design flood water surface line of the target mountain river by performing segment-by-segment calculations based on the design peak flow value of each node and through the energy equation.

[0057] The flood evolution dynamic simulation report generation module is used to dynamically simulate the flood evolution process of mountain rivers based on the design flood water surface line, and obtain a flood evolution dynamic simulation report of the target mountain river.

[0058] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, the at least one program, the code set or instruction set is loaded and executed by the processor to implement the dynamic prediction method for mountain river floods based on big data fusion as described in any of the embodiments of this application.

[0059] Fourthly, this application also provides a computer-readable storage medium storing at least one piece of program code, which is loaded and executed by a processor to implement the dynamic prediction method for mountain river floods based on big data fusion as described in any of the embodiments of this application.

[0060] The aforementioned method, apparatus, equipment, and medium for dynamic flood prediction in mountainous rivers based on big data fusion acquire a multi-source fusion dataset of the target mountainous river; based on the multi-source fusion dataset and pre-defined threshold rules, the flood type of the mountainous river is identified to obtain a set of flood type and causal parameters; according to the flood type and causal parameter set, the design peak flow value of each node in the mountainous river is calculated; after segmental calculations using energy equations to obtain the design flood water surface line of each node, the evolution process of the mountainous river flood is dynamically simulated to generate a dynamic simulation report of the mountainous river flood evolution. This method can accurately adapt to the diverse causes and complex propagation processes of floods in mountainous rivers, significantly improving the precision and reliability of flood prediction. Attached Figure Description

[0061] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0062] Figure 1 This is a flowchart illustrating a method for dynamic flood prediction in mountainous rivers based on big data fusion in one embodiment.

[0063] Figure 2 This is a flowchart illustrating the steps of identifying flood types in mountainous rivers and obtaining the flood type identification results and causal parameter set corresponding to the target mountainous river based on a multi-source fusion dataset and a preset threshold rule in one embodiment.

[0064] Figure 3 This is a schematic diagram of a dynamic flood prediction device for mountain rivers based on big data fusion in one embodiment. Detailed Implementation

[0065] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0066] In one embodiment, a method for dynamic flood prediction of mountain rivers based on big data fusion is provided. This embodiment illustrates the application of this method to a terminal. It is understood that this method can also be applied to a server, and can also be applied to a system including a terminal and a server, and implemented through the interaction between the terminal and the server. Figure 1 As shown, in this embodiment, the method includes the following steps:

[0067] Step S101: Obtain multi-source basic data of the target mountain river and perform fusion processing on the multi-source basic data to obtain a multi-source fusion dataset; the multi-source data includes hydrological station monitoring data, climate data, remote sensing data and river topography data.

[0068] Among these, fusion processing integrates information / data / entities from multiple sources, types, or dimensions using specific methods to eliminate redundancy, compensate for the limitations of single information, and form a more comprehensive result; target mountain rivers are mountain rivers selected as objects of attention or action based on specific needs; hydrological station monitoring data refers to time-series data accumulated over a long period of observation by basic national hydrological stations within the basin (such as Bayinbuluke Station and Dashankou Station mentioned in the background information), mainly including water level, flow, and sediment content; climate data comes from meteorological monitoring station networks and reanalysis data, including a series of meteorological elements such as precipitation, temperature, wind speed, humidity, and evaporation; remote sensing data can reflect information such as land cover, snow cover range, glacier distribution, and soil moisture; geospatial data with river channel topography data as the core usually exists in the form of a Digital Elevation Model (DEM); DEM data is natural raster structure data, characterized by dividing the topographic surface of the study area into regularly arranged square or rectangular units, i.e., raster units, each raster unit corresponding to a unique geographic coordinate (such as latitude and longitude) and topographic elevation value (…). The size of the raster cells is predefined by the spatial resolution of the DEM and is an inherent property of the DEM data.

[0069] For example, basic data such as original hydrological station monitoring data, climate data, remote sensing data and river topography data are obtained from the target watershed, and these multi-source basic data with different sources, formats and spatiotemporal resolutions are fused.

[0070] Step S102: Based on the multi-source fusion dataset, the flood type of mountain river is identified by a preset threshold rule to obtain the flood type identification result and the set of causal parameters corresponding to the target mountain river.

[0071] Among them, the preset threshold rule is a judgment system pre-constructed based on the historical flood observation data, hydrological mechanism and watershed characteristics of the target mountain rivers. It sets judgment threshold ranges for different flood types for rainfall intensity, warming rate and snow melting rate, and the threshold ranges will be adaptively adjusted according to the watershed topography and seasonal characteristics such as altitude and slope.

[0072] For example, a flood feature vector is extracted from a multi-source fusion dataset. This vector contains three core features: rainfall intensity, warming rate, and snowmelt rate. Rainfall intensity, derived from climate data, reflects the amount of rainfall per unit time and is a key indicator for identifying rainstorm-type floods. The warming rate, also derived from climate data, is obtained by calculating the ratio of temperature difference to time interval over consecutive periods, characterizing the driving effect of temperature changes on snowmelt. The snowmelt rate data is derived by combining snow cover change information from remote sensing data with temperature data from climate data, analyzing the reduction in snow cover area or thickness per unit time, and is directly related to the formation potential of ablation-type floods. Based on these features and preset threshold rules, the flood types in mountainous rivers are identified, clarifying the flood type. After type identification, causal parameters closely related to the identified types are further extracted to form a causal parameter set. For example, if identified as ablation-type, glacier area, temperature gradient, etc., of the relevant region need to be extracted from the dataset.

[0073] Step S103: Based on the multi-source fusion dataset, flood type identification results and causal parameter set, calculate the design peak flow corresponding to the target mountain river and obtain the design peak flow value of each node; the node is a key location with clear topographic features or hydrological monitoring function.

[0074] For example, the appropriate flood design frequency is selected based on the flood type identification results and the set of causal parameters. This frequency reflects the recurrence probability of the flood event (e.g., once in a hundred years, once in fifty years). The coefficient of variation of each hydrological station is extracted from the multi-source fusion dataset, and the design peak discharge value of each node is calculated based on the flood design frequency and the coefficient of variation.

[0075] Step S104: Based on the design peak flow value of each node, the design flood water surface line of the target mountain river is obtained by performing segment-by-segment calculations using the energy equation.

[0076] Among them, the energy equation is a mathematical expression describing the relationship between energy conservation and transformation. Its essence is the quantitative manifestation of the law of conservation of energy, that is, energy can neither be created out of thin air nor disappear out of thin air, but can only be transformed from one form to another. The total energy remains unchanged during the transformation / transfer process. The form of the energy equation in different scenarios will be adjusted according to the type of energy involved. Segmented trial and error is a method to obtain the water level line of a river flood by segmented processing, segmented derivation and iterative verification. It is often used to solve the problem of water level distribution along the course in complex terrains such as mountain rivers.

[0077] For example, based on the design peak flow value of each node, the design flood water surface line of the target mountain river is obtained by performing segment-by-segment calculations using the energy equation.

[0078] Step S105: Based on the design flood water surface line, perform dynamic simulation processing on the flood evolution process of the target mountain river to obtain a dynamic simulation report on the flood evolution of the target mountain river.

[0079] Dynamic simulation processing refers to a complete process that uses the design flood water surface line at each node as a basis, combined with river topography and flood peak propagation patterns, to transform static flood parameters into an intuitive and continuous dynamic evolution process through multi-stage technical processing, and generate a visual report.

[0080] For example, based on the obtained design flood peak flow value and river length of each node, the time required for the flood peak to propagate from the upstream to each downstream node is calculated. Combined with the static design flood water surface line corresponding to each node, the flood of the target mountain river is dynamically simulated and processed to generate a dynamic simulation report of the flood evolution of the mountain river.

[0081] In this embodiment, deep fusion of multi-source basic data is performed; flood types are accurately identified and corresponding causal parameter sets are extracted by combining preset threshold rules; the design flood water surface line is obtained through segmented calculations using the Bernoulli energy equation; and the flood evolution process of the target mountain river is dynamically simulated based on this design flood water surface line to obtain a dynamic simulation report of flood evolution. This approach can accurately adapt to the diverse causes and complex propagation processes of floods in mountain rivers, significantly improving the precision and reliability of flood forecasting.

[0082] In one embodiment, such as Figure 2 As shown, based on a multi-source fusion dataset, flood types in mountainous rivers are identified using preset threshold rules, yielding flood type identification results and a set of causal parameters for the target mountainous rivers, including:

[0083] Step S201: Extract flood feature vectors from the multi-source fusion dataset; the flood feature vectors include rainfall intensity, warming rate and snow melting rate.

[0084] For example, the multi-source fusion dataset already contains data in a unified format and with spatiotemporal benchmarks, including hydrological, climatic, remote sensing, and topographic data. From this dataset, three key features directly related to flood formation—rainfall intensity, warming rate, and snowmelt rate—are selected. During the extraction process, the 3σ principle can be used to remove outliers and unify the time scale of the original data, ensuring that the three feature values ​​remain consistent across the time dimension, forming a structurally regular flood feature vector. The 3σ principle is based on the statistical laws of the normal distribution (Gaussian distribution) and is used to describe the central tendency and dispersion of data under a normal distribution. Its core principle is that "the vast majority of data will fall within a specific range near the mean."

[0085] Step S202: Based on the feature values ​​in the flood feature vector, the flood type of mountain river is dynamically identified by a preset threshold rule to obtain the flood type identification result corresponding to the target mountain river; the flood types of mountain river include ablation type, rainstorm type and rain-snow mixed type.

[0086] The specific identification logic of the preset threshold rule is as follows: if the snow melting rate and the warming rate in the feature vector are higher than the preset melting threshold and the rainfall intensity is lower than the preset rainfall intensity threshold, it is determined to be a melting flood, which is mainly driven by snow melting; if the rainfall intensity is higher than the preset rainfall intensity threshold and the warming rate and the snow melting rate are both lower than the corresponding thresholds, it is determined to be a rainstorm flood, which is mainly caused by heavy rainfall; if the rainfall intensity, warming rate, and snow melting rate are all in their respective intermediate threshold ranges, and the three together meet the superposition conditions for flood formation, it is determined to be a rain-snow mixed flood, which is formed by the combined effect of rainfall and snow melting.

[0087] For example, based on preset threshold rules and the feature values ​​in the extracted flood feature vector, the current flood type is dynamically identified by comparing the matching relationship between the feature values ​​and the threshold intervals in the preset threshold rules in time-by-time, thus obtaining a clear flood type identification result for the target mountain river.

[0088] Step S203: Based on the flood type identification results, extract the corresponding flood cause-related parameters from the multi-source fusion dataset; the flood cause-related parameters are parameters that can directly affect the formation of the corresponding flood type.

[0089] For example, based on the flood type identification results, parameters directly related to the causes of the current flood type are selected from the multi-source fusion dataset. Since the formation mechanisms of different types of floods are different, their corresponding cause-related parameters are also highly targeted. For example, if the identification result is an ablation flood, parameters such as snow cover rate, snow thickness, and warming duration are extracted. These parameters directly determine the total amount and rate of snow ablation, thus affecting the scale of the ablation flood. If it is a rainstorm flood, parameters such as rainfall duration, peak rainfall intensity, and average rainfall in the basin are extracted. These parameters directly reflect the intensity and duration of rainfall replenishment to the flood. If it is a rain-snow mixed flood, the core items of the above two types of parameters are extracted at the same time to cover the influencing factors of both rainfall and snow ablation. Among them, snow cover rate is taken from remote sensing data to reflect the snow distribution range in the basin; snow thickness is derived by combining remote sensing data and ground observation data; warming duration is taken from climate data and obtained by calculating the continuous period when the temperature is higher than the snow melting critical temperature; average rainfall in the basin is calculated by interpolation of meteorological station data and remote sensing rainfall data.

[0090] Step S204: Combine and process the various flood-related parameters to obtain a set of causal parameters.

[0091] For example, parameters directly related to the causes of the current flood type can be selected from the multi-source fusion dataset to form a causal parameter set.

[0092] In this embodiment, a flood feature vector composed of rainfall intensity, warming rate, and snowmelt rate is extracted from multi-source fusion data; the flood type (melting type, rainstorm type, or mixed rain and snow type) is dynamically determined based on preset threshold rules; causal parameters that directly affect the formation of different flood types are extracted; and finally, a set of causal parameters is formed. This enables deep coupling between flood type identification and causal parameter extraction, effectively improving the accuracy of flood type determination and the specificity of the parameters.

[0093] In one embodiment, based on a multi-source fusion dataset, flood type identification results, and a set of causal parameters, the design peak flow corresponding to the target mountain river is calculated to obtain the design peak flow value for each node, including:

[0094] Step S301: Based on the flood type identification results and the set of causal parameters, determine the corresponding frequency analysis parameters.

[0095] Among them, frequency analysis parameters are the core basic parameters used to calculate the design frequency of floods. Their values ​​must be highly adapted to the characteristics of flood formation. For example, if the identification result is an ablation flood, parameters such as snowmelt rate and warming duration in the causal parameter set will point to frequency analysis parameters related to snowmelt floods, such as the return period coefficient corresponding to snowmelt amount and temperature anomaly frequency parameters. If it is a rainstorm flood, the rainfall duration and peak rainfall intensity in the causal parameter set will correspond to rainstorm frequency analysis parameters, such as parameters related to the rainstorm intensity-duration-frequency curve and watershed runoff coefficient. If it is a rain-snow mixed flood, the frequency analysis parameters of the two types of floods will be integrated, and composite parameters that can reflect the superposition effect of rainfall and snowmelt will be selected.

[0096] For example, based on a pre-defined mapping table of "flood type - causal parameters - frequency analysis parameters", the corresponding frequency analysis parameters are accurately matched and determined from the mapping table by comparing the flood type identification results and causal parameter characteristics of the target mountain river. The pre-defined mapping table establishes the logical correspondence between "why floods occur (type and cause)" and "how to quantify and analyze their probability of occurrence (frequency parameters)".

[0097] Step S302: Calculate the flood design frequency for the target mountain river based on the frequency analysis parameters.

[0098] For example, frequency analysis parameters can be substituted into the probability density function of the Pearson Type III distribution. Parameter estimation determines the shape, location, and scale parameters of the distribution curve. Then, based on this distribution curve, and according to the return period requirements of flood control projects such as levee construction and reservoir scheduling, the corresponding flood design frequency can be derived. For instance, to determine the design frequency for a 100-year flood, the probability value corresponding to that return period is calculated from the distribution curve, which is the flood design frequency. The Pearson Type III distribution is an important continuous distribution commonly used in hydrology and meteorology to describe the probability distribution of extreme events such as rainstorms and flood flows. Its probability density function flexibly fits skewed data, especially right-skewed data, using the three parameters of "shape-scale-location," conforming to the characteristic of extreme events having a few large values ​​and many small values. Parameter estimation uses sample data extracted from the population to infer unknown parameters in the population, such as the population mean and population variance, solving the key problem of inferring the population from a sample.

[0099] Step S303: Extract the annual maximum flood peak flow sequence of each hydrological station from the monitoring data of the hydrological stations.

[0100] For example, the multi-source fusion dataset includes hydrological station monitoring data covering the annual time-period peak flow records of each hydrological station. First, the time range for data extraction is determined, typically at least 30 years of continuous observation data to meet the sample size requirements for hydrological frequency analysis. Then, the time-period peak flow data for each hydrological station is filtered, selecting the peak flow data with the largest flow value each year as the annual maximum peak flow value. During the extraction process, quality control is performed on the original monitoring data: outliers caused by equipment failure or human error are removed, and missing years' data are supplemented using interpolation of data from adjacent hydrological stations of the same period, forming a complete and continuous annual maximum peak flow sequence for each hydrological station. Interpolation is a mathematical method for estimating unknown values ​​between known data points. Using known, discrete observation points, a function that fits these known points is constructed, called the interpolation function, and then this function is used to calculate the unknown function value at any position between the known points.

[0101] Step S304: Perform statistical processing on the annual maximum flood peak flow series to obtain the mean and standard deviation of each annual maximum flood peak flow series.

[0102] For example, for each annual maximum flood peak flow sequence, the mean flood peak flow of the sequence is obtained by summing all the data in the sequence and then dividing by the number of samples in the sequence, i.e., the number of observation years. This mean reflects the average level of flood peak flow in the river segment where the hydrological station is located. The standard deviation is obtained by calculating the square of the difference between each data point in the sequence and the mean, summing all the squared differences, dividing by the number of samples, and then taking the square root. This indicator reflects the dispersion of the annual maximum flood peak flow sequence; the larger the standard deviation, the greater the fluctuation range of the flood peak flow over the years.

[0103] Step S305: Based on the mean and standard deviation, obtain the coefficient of variation values ​​for each hydrological station.

[0104] The coefficient of variation is a statistic that reflects the relative dispersion of the data, and its calculation formula is as follows:

[0105]

[0106] This indicator can eliminate the influence of dimensions, making it convenient for comparative analysis of the dispersion of peak flow at different hydrological stations and in different watersheds.

[0107] For example, the mean and standard deviation of each annual maximum flood peak discharge sequence are substituted into the coefficient of variation calculation formula and divided to obtain the coefficient of variation value for each hydrological station. Each time the calculation is performed, the mean and standard deviation data of the same hydrological station are retrieved to ensure that they correspond to the same time series and the same observation period. If the coefficient of variation value is large, it indicates that the annual flood peak discharge of the hydrological station fluctuates greatly relative to the mean, and the interannual differences in flood size are significant; if the coefficient of variation value is small, it indicates that the annual flood peak discharge is relatively stable.

[0108] Step S306: Based on the values ​​of each variation coefficient and the flood design frequency, the initial design peak flood discharge of each hydrological station is calculated using the following formula:

[0109]

[0110] in, Let p be the initial design peak flow rate, and p be the flood design frequency. The average peak flow rate. This is the deviation coefficient corresponding to the design frequency of the flood. This represents the coefficient of variation.

[0111] Among them, the deviation coefficient corresponding to the design frequency of floods in mountainous rivers It is obtained by querying a pre-stored table of "deviation coefficient - design frequency - variation coefficient", the data in which is generated by Pearson type III distribution.

[0112] For example, based on the current design frequency and the coefficient of variation value, the corresponding [data] is precisely matched. The initial design peak flow of each hydrological station was calculated based on the initial design peak flow calculation formula.

[0113] Step S307: Obtain the river channel attenuation characteristic parameters, and perform a path-by-path attenuation correction on the initial design peak flow based on the river channel attenuation characteristic parameters to obtain the design peak flow value for each node.

[0114] Among them, the river channel attenuation characteristic parameters are parameters that reflect the flow attenuation law during the propagation of floods in the river channel. These parameters mainly include the river channel roughness, the river channel cross-sectional morphology parameters (such as the cross-sectional width-to-depth ratio), the friction coefficient, and the watershed confluence time. These parameters directly affect the degree of flood energy dissipation. For example, the greater the river channel roughness, the stronger the flow resistance and the more significant the flow attenuation; the smaller the cross-sectional width-to-depth ratio, the weaker the lateral diffusion during flood propagation and the gentler the flow attenuation.

[0115] For example, river attenuation characteristic parameters can be obtained using a friction loss model in hydraulics. The obtained parameters are substituted into the model equations to calculate the flow attenuation as the flood propagates from the hydrological station cross-section to each key node. Subsequently, the corresponding attenuation is subtracted from the initial design peak flow to perform friction loss correction. During the correction process, the attenuation is calculated segmentally based on the river length between nodes to ensure that the design peak flow value at each node accurately reflects the actual scale of the flood propagation to that location. The friction loss model describes the flow and water level changes caused by hydraulic resistance, diffusion, and storage during the transmission of water, especially floods and river runoff. It is a key tool for river flood calculation and water resource regulation.

[0116] In this embodiment, suitable frequency analysis parameters are determined by flood type and causal parameters; the initial design peak flow is calculated by combining the hydrological frequency analysis model with the mean and standard deviation of the annual maximum peak flow sequence; and the initial values ​​are corrected by river channel attenuation characteristic parameters to obtain the actual peak flow at each node. This effectively solves the calculation error problems caused by the disconnect between parameters and flood type and the failure to consider flow attenuation along the course in traditional methods, significantly improving the calculation accuracy of design peak flow at each node of mountain rivers.

[0117] In one embodiment, based on the design peak flow value of each node, the design flood water surface line of the target mountain river is obtained by segment-by-segment calculation using the energy equation, including:

[0118] Step S401: Determine the flow measurement section of the hydrological station downstream of the target mountain river as the starting point for the design flood water surface line.

[0119] For example, the flow measurement section of the hydrological station downstream of the target mountain river is set as the starting point for the design flood water surface line. This choice is based on the natural advantages and data reliability of the downstream section. The downstream section of a mountain river is usually relatively flat, with more stable water flow. Moreover, the flow measurement section of the downstream hydrological station has been constructed and maintained for a long time, and has continuous and complete measured hydrological data such as water level, flow rate, and cross-sectional morphology data.

[0120] Step S402: Extract the measured water level corresponding to the flood design frequency from the monitoring data of the hydrological station as the initial water level of the starting node.

[0121] For example, measured water level data at the design frequency corresponding to the starting node are selected and extracted from the hydrological station monitoring data contained in the multi-source fusion dataset, and used as the initial water level for water surface line calculation. This measured water level data includes both cross-sectional water level records of actual floods at that frequency in history, and water level values ​​at the corresponding frequency derived from long-term observation data through frequency analysis methods.

[0122] Step S403: Based on the initial water level of the starting node and the design peak flow value corresponding to the starting node, perform segmented trial calculations from downstream to upstream using Bernoulli's energy equation to obtain the design flood level of each node.

[0123] Bernoulli's energy equation is an equation based on the principle of energy conservation that describes the conversion relationship between potential energy, kinetic energy, and head loss in water flow.

[0124] For example, based on the initial water level of the starting node and its corresponding design peak flow value, Bernoulli's energy equation is used to perform segmented trial calculations on each key node from downstream to upstream to obtain the design flood level of each node.

[0125] Step S404: Connect the design flood levels of each node to generate the design flood water surface line of the target mountain river.

[0126] For example, the design flood level data of each node is checked for spatiotemporal consistency to confirm that all water level values ​​correspond to the same flood design frequency and calculation period, avoiding distortion of the water surface line shape caused by misaligned data timestamps or frequency mismatches. An appropriate interpolation method is selected to connect the water levels of adjacent nodes based on the river's topographic features. For river sections with gentle terrain and linear water level changes, linear interpolation can be used; for river sections with large topographic undulations (such as sharp bends and bottleneck sections) and nonlinear water level changes, spline interpolation can be used to accurately fit the water level transition trend between adjacent nodes, generating the design flood water surface line of the target mountain river. Linear interpolation uses the linear relationship between two known data points to calculate the value at any unknown location between these two points. Essentially, it connects two points on a two-dimensional plane to form a straight line and calculates the coordinates of the intermediate point using the equation of the straight line. Spline interpolation constructs low-order polynomials, usually cubic (cubic splines), piecewise, allowing these polynomials to connect smoothly at the interpolation nodes, forming a continuous and smooth curve used to fit known discrete data points or fill data gaps.

[0127] In this embodiment, the flow measurement section of a hydrological station with reliable downstream data is selected as the starting node, and the measured initial water level of the corresponding design frequency is extracted. Based on Bernoulli's energy equation, the design flood level of each node is calculated segment by segment from downstream to upstream. The design flood water surface line is generated by connecting the design flood levels of each node through interpolation fitting. This effectively solves the problems of unreliable starting data and neglect of topographic influence in traditional water surface line calculations.

[0128] In one embodiment, based on the initial water level of the starting node and the design peak flow value corresponding to the starting node, a segmented trial calculation is performed from downstream to upstream using Bernoulli's energy equation to obtain the design flood level of each node, including:

[0129] Step S501: Based on the initial water level of the starting node and the design peak flow value corresponding to the starting node, perform segment-by-segment calculations from downstream to upstream using Bernoulli's energy equation to obtain the design flood level of each node:

[0130]

[0131]

[0132]

[0133] in, To design flood level, Here, dv represents the design flood level of the upstream node, and dv represents the change in flow velocity. This is the kinetic energy correction factor. and Let g be the flow velocity and g be the acceleration due to gravity. Let η be the head energy loss, η be the riverbed roughness, L be the node spacing, R be the hydraulic radius, and A be the cross-sectional area of ​​the water passage.

[0134] Among them, the kinetic energy correction coefficient α can be selected according to the uniformity of water flow. It is usually taken as 1.05-1.1 in slow flow scenarios and 1.1-1.2 in fast flow scenarios. The riverbed roughness value η is obtained by querying the riverbed roughness mapping table. The riverbed roughness mapping table is a standardized data table constructed based on long-term hydrological observation data of mountain rivers, riverbed material composition survey (such as sediment particle size and vegetation coverage), and indoor water flow resistance test.

[0135] For example, based on the initial water level and the design peak flow value of the starting node, the river length L between the two nodes is extracted from the river topography data of the multi-source fusion dataset, and the cross-sectional morphology parameters such as the cross-sectional width and depth of the two nodes are obtained. The parameters are then iteratively adjusted... The initial assumed values ​​are used until the energy balance at both ends of Bernoulli's energy equation is reached, and the design flood level of node 2 is determined. Then, with node 2 as the new downstream node, the above process is repeated upstream until the design flood levels of all nodes are obtained.

[0136] In this embodiment, the design flood level for each node is obtained by iteratively calculating segment by segment using Bernoulli's energy equation based on the obtained reliable initial water level and design peak flow. This effectively eliminates calculation errors caused by complex terrain and ensures the accuracy of the design flood level for each node.

[0137] In one embodiment, based on the design flood water surface line, the flood evolution process of the target mountain river is dynamically simulated to obtain a dynamic simulation report of the flood evolution of the target mountain river, including:

[0138] Step S601: Based on the design peak flow value and river topography data of each node, calculate the peak propagation speed using hydraulic formulas.

[0139] Among them, the river topographic data includes key information such as the river gradient (i.e., the ratio of the elevation difference between two points upstream and downstream of the river to the horizontal distance) and the cross-sectional shape of the river (such as the cross-sectional width and average depth); the hydraulic formula is a mathematical expression describing the laws of stillness or motion of liquids (mainly water), which revolves around the three basic principles of force balance, energy conservation, and mass conservation.

[0140] For example, based on the design peak flow values ​​of each node and the river topography data in the multi-source fusion dataset, the peak flow velocity can be calculated using hydraulic formulas. The hydraulic formula used can be the Manning formula, which calculates the flow velocity through the correlation between water flow resistance and river characteristics. The formula, after modification, can be used to derive the peak flow velocity. The formula is as follows:

[0141]

[0142] in, Let η be the peak flow velocity, η be the riverbed roughness, and R be the hydraulic radius, which is the ratio of the cross-sectional area to the wetted perimeter. For example, the cross-sectional area is determined based on the design peak flow rate and cross-sectional morphology parameters at each node, the hydraulic radius is calculated, and then substituted into the Manning formula to obtain the peak flow velocity for each river segment.

[0143] Step S602: Based on the propagation speed of the flood peak and the length of the river channel between the two sections, the propagation time of the flood peak is calculated.

[0144] For example, based on the calculated peak flow velocity and the river channel length between two adjacent cross-sections, the peak flow time for each river segment is derived. The river channel length between two cross-sections is not a straight-line distance, but rather the actual length of the river channel centerline extracted from the DEM: the actual channel length between adjacent cross-sections can be measured along the river channel centerline trajectory, avoiding straight-line distance errors caused by terrain undulations. The peak flow time is the ratio of distance to velocity; it is obtained by dividing the actual river channel length between two cross-sections by the corresponding peak flow velocity for that river segment.

[0145] Step S603: Based on the time when the flood peak arrives at the initial section, the flood peak propagation time of each segment is accumulated to obtain the flood peak propagation time series.

[0146] For example, the time when the flood peak arrives at the initial cross-section is determined as the reference time. The initial cross-section is usually a flow measurement cross-section at a designated downstream hydrological station. The reference time can be set to 0 o'clock or a specific time point according to actual flood control needs, such as the start time of a flood warning. Following the order from downstream to upstream, the flood peak propagation time of each river segment is sequentially accumulated to the reference time. This accumulation generates a flood peak propagation time series containing the arrival times of flood peaks at all nodes.

[0147] Step S604: Based on the flood peak propagation time series, the design flood water surface line is spatiotemporally connected to obtain spatiotemporally connected water surface line data.

[0148] Among them, spatiotemporal connection processing is a technology for transitioning, associating and integrating different time dimensions and spatial scenes to avoid the abruptness of spatiotemporal jumps.

[0149] For example, using the flood peak propagation time series as a time reference, the design flood water surface line is spatiotemporally linked to eliminate the data's discreteness in the time and spatial dimensions. Spatially, GIS spatial registration technology can be used to accurately correlate the water level data of each node with the coordinates of the river channel centerline, ensuring that each water level value corresponds to a unique spatial location on the river channel, avoiding distortion of the water surface line shape due to coordinate offset. Temporally, based on the flood peak propagation time series, the water level data of each node can be labeled with the corresponding flood peak arrival time, forming a three-dimensional data structure of "time-space-water level". Simultaneously, interpolation methods can be used to fit the water level-time relationship between adjacent nodes, filling the time and spatial data gaps between two nodes, ensuring the water surface line data is continuous and smooth in the spatiotemporal dimensions. The final result is spatiotemporally linked water surface line data. GIS spatial registration technology uses mathematical methods to eliminate deviations in coordinate systems, locations, scales, or shapes of multi-source spatial data, allowing them to be overlaid, analyzed, and used collaboratively within the same spatial framework.

[0150] Step S605: Using animation rendering technology, the spatiotemporally connected water surface data is used to generate a flood evolution animation.

[0151] Among them, animation rendering technology is the core technology that transforms scene data such as virtual three-dimensional or two-dimensional models, materials, lighting, and camera angles built in the computer into visual image frames through algorithm calculations, and finally strings them together to form a smooth animation.

[0152] For example, a 3D terrain model of the target mountain river is first constructed: a terrain mesh is generated based on DEM data, and surface cover information such as vegetation and buildings from remote sensing data is overlaid to recreate the real geographical scene of the river channel and surrounding area; the spatiotemporally connected water surface data is mapped to a virtual water surface in the 3D scene, and a transparent, blue gradient visual effect is set for it to simulate the physical form of flood; the position and elevation of the virtual water surface are updated frame by frame according to the time sequence to generate continuous animation frames, which can be synthesized into a complete flood evolution animation through video encoding technology. Among them, video encoding technology significantly compresses the video data volume while ensuring acceptable image quality. The unencoded original video data volume is extremely large and cannot meet the needs of storage and transmission (such as network playback). Encoding technology is used to solve this contradiction.

[0153] Step S606: Overlay the design flood water surface line with the digital elevation model to obtain the flood inundation depth and range.

[0154] Among them, overlay analysis is used to deduce the inundation depth by comparing the water surface elevation with the topographic elevation at the same spatial location.

[0155] For example, the design flood water surface data is rasterized and converted to the same coordinate system as the topographic elevation data of the DEM data to ensure a one-to-one correspondence in spatial location; the inundation depth is calculated using the following formula:

[0156]

[0157] in, To design flood water surface data, Topographic elevation data for DEM data. The inundation depth is determined. For example, the inundation depth of each raster cell is calculated. When h > 0, the raster cell is considered flooded; when h ≤ 0, it is considered unflooded. Finally, all raster cells identified as flooded are spatially aggregated to form continuous inundation area areal data, which represents the flood inundation range. Rasterization refers to using a spatial interpolation algorithm based on the spatial resolution and coordinate system of the DEM to convert discrete water surface elevation values... Fill the geographic coordinates that correspond one-to-one with the DEM raster cells to generate flood water surface raster data with the same resolution and coordinate system as the DEM. In this process, the number, size, and geographic coordinates of the raster cells are completely matched with the DEM raster cells.

[0158] Step S607: Generate a flood risk level distribution map based on the flood inundation depth and extent.

[0159] For example, based on the obtained flood inundation depth and range and predefined flood risk classification standards, each inundation grid cell is assigned to its corresponding risk level according to water depth, generating a risk level raster map. Then, vector data such as administrative divisions and transportation routes are overlaid to mark key facilities within high-risk areas, ultimately forming an intuitive flood risk level distribution map. The predefined risk classification standards are based on population density, building types, infrastructure distribution (such as roads and bridges), and flood control standards in mountainous river basins. Typically, the inundation depth is divided into multiple intervals, each corresponding to a risk level. For example, shallower inundation depths (…) (This corresponds to low risk and medium water depth) <h≤ This corresponds to medium risk, with relatively deep water (h> ( ) corresponds to high risk, and different colors can be used to indicate different risk levels.

[0160] Step S608: Integrate flood evolution animation and flood risk level distribution map to generate a dynamic simulation report of flood evolution of the target mountain river.

[0161] For example, flood evolution animations can be embedded in the report as video files, with timeline annotations and key node descriptions added to facilitate viewers' understanding of the flood evolution status at different times. Flood risk level distribution maps are inserted as high-resolution images, accompanied by legends explaining the risk level classification standards and color meanings. Furthermore, the report can include technical parameters such as design frequency, riverbed roughness value, and design peak flow at each node from the preliminary calculation process, along with explanations of data sources and uncertainty analysis of the simulation results, ultimately generating a dynamic simulation report of flood evolution for the target mountain river.

[0162] In this embodiment, a flood peak propagation time series is generated by calculating the propagation time of the flood peak between each node and superimposing these propagation times. Based on this series, the design flood water surface line is spatiotemporally connected. After the spatiotemporal connection is completed, a flood evolution animation is generated using animation rendering technology. The design flood water surface line is overlaid and analyzed with a digital elevation model to obtain the flood inundation depth and range. Based on the flood evolution animation and the flood inundation depth and range, a dynamic simulation report of the flood evolution in mountainous rivers is generated. This effectively solves the problems of insufficient dynamism, low visualization, and fragmented results presentation in traditional flood simulations.

[0163] In one embodiment, the multi-source basic data is fused to obtain a multi-source fused dataset, including:

[0164] Step S701: Perform missing value supplementation processing on the hydrological station monitoring data in the multi-source basic data to obtain a continuous and complete hydrological data sequence.

[0165] For example, hydrological station monitoring data may be affected by factors such as equipment failure and extreme weather interference, leading to interruptions or missing values, necessitating missing value supplementation. Interpolation methods can be used to repair missing values, with common methods including linear interpolation and Kriging interpolation. Linear interpolation is suitable for scenarios with short data gaps and gradual changes in adjacent monitoring values, deriving missing values ​​by calculating the linear relationship between valid monitoring values ​​before and after the missing period. Kriging interpolation, based on the principle of spatial correlation, combines concurrent monitoring data from neighboring hydrological stations to construct a spatial distribution model of regional hydrological elements, accurately estimating missing values. It is particularly suitable for situations with long missing periods or insufficient reliability of data from a single hydrological station. It first uses the 3σ principle to remove outliers from the monitoring data, such as extreme values ​​exceeding reasonable hydrological ranges, and then supplements the remaining valid data with missing values, forming a continuous and numerically reliable hydrological data sequence.

[0166] Step S702: The climate data and remote sensing data in the multi-source basic data are processed into grids to obtain a raster dataset with uniform spatial resolution.

[0167] Among them, grid processing is a technology and method that breaks down a continuous space, data or task into regular grid units such as squares and hexagons, and then solves problems through unit-based management, calculation or analysis.

[0168] For example, the original spatial resolution of climate data and remote sensing data in multi-source basic data often differs. Resampling techniques can be used for gridding to obtain a raster dataset with uniform spatial resolution and complete coverage. Resampling refers to adjusting the sampling frequency (or number of samples) of data (or signals) to adapt to different analytical needs, equipment limitations, or data format requirements. For instance, for data with a resolution higher than the target precision, the average of multiple original raster cells can be calculated to generate a single target raster cell value; for data with a resolution lower than the target precision, bilinear interpolation can be used to improve the resolution. Bilinear interpolation is an interpolation technique for numerical estimation in a two-dimensional grid. It calculates an approximate value of the target point in two steps (horizontal first, then vertical, or vice versa) using the values ​​of four known coordinate points around the target point. The target precision can be preset based on the area of ​​the mountainous river basin and the complexity of the terrain, ensuring that the raster cells accurately reflect local terrain and meteorological characteristics while avoiding excessive data volume that increases the computational burden.

[0169] Step S703 involves performing spatiotemporal registration processing on the continuous and complete hydrological data sequence, the raster dataset with uniform spatial resolution, and the river topographic data to obtain the initial multi-source fusion dataset.

[0170] Spatiotemporal registration refers to unifying data from different times and spaces into the same spatiotemporal coordinate system to eliminate spatiotemporal deviations caused by differences in data acquisition conditions.

[0171] For example, based on the obtained continuous hydrological data sequence, unified resolution raster dataset, and river topographic data from multi-source basic data, the timestamps of all data are unified, and the data is time-sliced ​​according to preset time intervals (such as hourly or daily levels) to ensure that rainfall records in climate data, flow changes in hydrological data, and surface observations in remote sensing data correspond to the same time dimension. The coordinate systems of all spatial data are unified, and spatial offsets from different data sources are corrected through coordinate transformation algorithms. Simultaneously, the hydrological data sequence is associated with the spatial locations of corresponding hydrological stations, ensuring that each hydrological monitoring value is accurately mapped to specific geographic coordinates of the DEM. Finally, the three types of data are integrated according to a three-dimensional structure of "time-space-feature" to form an initial multi-source fusion dataset. The coordinate transformation algorithm is a mathematical method for transforming spatial points between different coordinate systems. By defining the baseline differences of the coordinate systems (such as origin, coordinate axis direction, and scale), the coordinate parameter mapping relationship of the points is established.

[0172] Format standardization refers to the process of converting data from different sources and with different formats (such as text, numerical data, spatial data, etc.) into a unified and standardized format through preset rules and technical means.

[0173] Step S704: Standardize the format of the initial multi-source dataset to generate a multi-source fusion dataset.

[0174] For example, the initial multi-source dataset contains various data formats. These data can be standardized according to preset format standards. For instance, CSV-formatted hydrological data can be converted to GeoJSON (GeoJavaScript Object Notation, JavaScript being a multi-paradigm high-level interpreted programming language) format suitable for spatial analysis, ensuring its association with spatial data. Raster data and terrain data of different formats can be uniformly converted to GeoTIFF (Geographic Tagged Image File Format), and the metadata information of the data can be standardized. Finally, a multi-source fusion dataset with a unified structure and convenient access is generated. The preset format standards are a set of consistent and standardized format rules pre-defined for specific scenarios.

[0175] In this embodiment, missing values ​​in hydrological data are repaired, the spatial resolution of climate and remote sensing data is unified, and spatiotemporal registration eliminates data misalignment in time and space. Finally, the spatiotemporally registered data is standardized to form a multi-source fusion dataset. This effectively solves the pain points of fragmented multi-source data, mismatched spatiotemporal scales, and chaotic formats in traditional mountain river flood forecasting.

[0176] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0177] Based on the same inventive concept, this application also provides a device for implementing the above-mentioned method for dynamic prediction of mountain river floods based on big data fusion. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations of one or more embodiments of the device for dynamic prediction of mountain river floods based on big data fusion provided below can be found in the limitations of the method for dynamic prediction of mountain river floods based on big data fusion described above, and will not be repeated here.

[0178] In one exemplary embodiment, such as Figure 3 As shown, a dynamic flood prediction device 300 for mountain rivers based on big data fusion is provided, including:

[0179] The data acquisition module 301 is used to acquire multi-source basic data of rivers in the target mountainous area and to fuse the multi-source basic data to obtain a multi-source fused dataset; the multi-source data includes hydrological station monitoring data, climate data, remote sensing data and river topography data;

[0180] The flood type identification module 302 is used to identify the flood type of mountain rivers based on a multi-source fusion dataset and a preset threshold rule, so as to obtain the flood type identification result and the set of causal parameters corresponding to the target mountain river;

[0181] Design flood peak flow calculation module 303 is used to calculate the design flood peak flow corresponding to the target mountain river based on multi-source fusion dataset, flood type identification results and causal parameter set, and obtain the design flood peak flow value of each node; the node is a key location with clear topographic features or hydrological monitoring function.

[0182] The design flood water surface line calculation module 304 is used to obtain the design flood water surface line of the target mountain area by performing segment-by-segment calculations based on the design peak flow value of each node and the energy equation.

[0183] The flood evolution dynamic simulation report generation module 305 is used to perform dynamic simulation processing on the flood evolution process of the target mountain river based on the design flood water surface line, and obtain the flood evolution dynamic simulation report of the target mountain river.

[0184] In one embodiment, the flood type identification module 302 is further configured to:

[0185] Flood feature vectors are extracted from multi-source fusion datasets; the flood feature vectors include rainfall intensity, warming rate, and snowmelt rate.

[0186] Based on the feature values ​​in the flood feature vector, the flood type of mountain rivers is dynamically identified through a preset threshold rule, and the flood type identification result corresponding to the target mountain river is obtained; the flood types of mountain rivers include ablation type, rainstorm type, and rain-snow mixed type;

[0187] Based on the flood type identification results, corresponding flood causation-related parameters are extracted from the multi-source fusion dataset; these parameters are those that can directly affect the formation of the corresponding flood type.

[0188] By combining and processing various flood-related parameters, a set of causal parameters is obtained.

[0189] In one embodiment, the flood peak flow calculation module 303 is further configured to:

[0190] Based on the flood type identification results and the set of causal parameters, the corresponding frequency analysis parameters are determined;

[0191] Based on frequency analysis parameters, the flood design frequency of the target mountain river is calculated;

[0192] Extract the annual maximum flood peak flow sequence of each hydrological station from the monitoring data of the hydrological stations;

[0193] Statistical processing was performed on the annual maximum flood peak flow series to obtain the mean and standard deviation of each annual maximum flood peak flow series;

[0194] The coefficient of variation for each hydrological station was obtained based on the mean and standard deviation.

[0195] Based on the values ​​of the coefficients of variation and the flood design frequency, the initial design peak flood discharge for each hydrological station is calculated using the following formula:

[0196]

[0197] in, Let p be the initial design peak flow rate, and p be the flood design frequency. The average peak flow rate. This is the deviation coefficient corresponding to the design frequency of the flood. This represents the coefficient of variation.

[0198] The river channel attenuation characteristic parameters are obtained, and the initial design peak flow is corrected for attenuation along the course based on the river channel attenuation characteristic parameters to obtain the design peak flow value of each node.

[0199] In one embodiment, the flood water surface line calculation module 304 is further configured to:

[0200] The hydrological station flow measurement section downstream of the target mountain river is determined as the starting point for the design flood water surface line;

[0201] The measured water level at the corresponding flood design frequency is extracted from the monitoring data of the hydrological station as the initial water level of the starting node;

[0202] Based on the initial water level of the starting node and the design peak flow value corresponding to the starting node, the design flood level of each node is obtained by performing segment-by-segment calculations from downstream to upstream using Bernoulli's energy equation.

[0203] Connect the design flood levels of each node to generate the design flood water surface line of the target mountain river.

[0204] In one embodiment, the flood water surface line calculation module 304 is further configured to:

[0205] Based on the initial water level of the starting node and the corresponding design peak flow value, the design flood level of each node is obtained by performing segment-by-segment calculations from downstream to upstream using Bernoulli's energy equation:

[0206]

[0207]

[0208]

[0209] in, To design flood level, Here, dv represents the design flood level of the upstream node, and dv represents the change in flow velocity. This is the kinetic energy correction factor. and Let g be the flow velocity and g be the acceleration due to gravity. Let η be the head energy loss, η be the riverbed roughness, L be the node spacing, R be the hydraulic radius, and A be the cross-sectional area of ​​the water passage.

[0210] In one embodiment, the flood evolution dynamic simulation report generation module 305 is further configured to:

[0211] Based on the design peak flow rate and river topography data at each node, the peak propagation velocity is calculated using hydraulic formulas.

[0212] Based on the propagation speed of the flood peak and the length of the river channel between the two sections, the propagation time of the flood peak is calculated.

[0213] Using the time when the flood peak arrives at the initial section as a benchmark, the flood peak propagation time of each segment is accumulated to obtain the flood peak propagation time series;

[0214] Based on the flood peak propagation time series, the design flood water surface line is spatiotemporally connected to obtain spatiotemporally connected water surface line data;

[0215] Using animation rendering technology, the spatiotemporally connected water surface data is used to generate a flood evolution animation;

[0216] By overlaying the design flood water surface line with the digital elevation model, the flood inundation depth and extent can be obtained.

[0217] A flood risk level distribution map is generated based on the flood inundation depth and extent.

[0218] Integrate flood evolution animations and flood risk level distribution maps to generate dynamic simulation reports of flood evolution for rivers in the target mountainous areas.

[0219] In one embodiment, the data acquisition module 301 is further configured to:

[0220] Missing values ​​in the hydrological station monitoring data from the multi-source basic data were filled to obtain a continuous and complete hydrological data sequence.

[0221] Climate data and remote sensing data from multi-source basic data are processed into grids to obtain a raster dataset with uniform spatial resolution;

[0222] The initial multi-source fusion dataset is obtained by performing spatiotemporal registration processing on continuous and complete hydrological data sequences, raster datasets with uniform spatial resolution, and river topographic data.

[0223] The initial multi-source dataset is format-standardized to generate a multi-source fused dataset.

[0224] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the above-described method for dynamic prediction of floods in mountainous rivers based on big data fusion.

[0225] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0226] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0227] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.

Claims

1. A method for dynamic flood prediction in mountainous rivers based on big data fusion, characterized in that, The method includes: Multi-source basic data of rivers in the target mountainous area are acquired, and the multi-source basic data are fused to obtain a multi-source fused dataset; the multi-source data includes hydrological station monitoring data, climate data, remote sensing data, and river topography data; Based on the multi-source fusion dataset, the flood type of mountain river is identified by a preset threshold rule, and the flood type identification result and the set of causal parameters corresponding to the target mountain river are obtained. Based on the multi-source fusion dataset, the flood type identification results, and the causal parameter set, the design peak flow corresponding to the target mountain river is calculated, and the design peak flow value of each node is obtained; the node is a key location with clear terrain features or hydrological monitoring functions. Based on the design peak flow values ​​of each node, the design flood water surface line of the target mountain river is obtained by segment-by-segment calculation using the energy equation. Based on the design flood water surface line, the flood evolution process of the target mountain river is dynamically simulated to obtain a dynamic simulation report of the flood evolution of the target mountain river.

2. The method according to claim 1, characterized in that, Based on the multi-source fusion dataset, the flood types of mountain rivers are identified using preset threshold rules to obtain the flood type identification results and causal parameter set corresponding to the target mountain river, including: Flood feature vectors are extracted from the multi-source fusion dataset; the flood feature vectors include rainfall intensity, warming rate, and snowmelt rate. Based on the feature values ​​in the flood feature vector, the flood type of mountain rivers is dynamically identified through a preset threshold rule to obtain the flood type identification result corresponding to the target mountain river; the flood types of mountain rivers include ablation type, rainstorm type, and rain-snow mixed type; Based on the flood type identification results, corresponding flood cause-related parameters are extracted from the multi-source fusion dataset; the flood cause-related parameters are parameters that can directly affect the formation of the corresponding flood type. The flood-related parameters are combined and processed to obtain a set of causal parameters.

3. The method according to claim 1, characterized in that, The step of calculating the design peak flow corresponding to the target mountain river based on the multi-source fusion dataset, the flood type identification results, and the causal parameter set, and obtaining the design peak flow value for each node, includes: Based on the flood type identification results and the set of causal parameters, the corresponding frequency analysis parameters are determined; Based on the frequency analysis parameters, the flood design frequency of the target mountain river is calculated; Extract the annual maximum flood peak flow sequence of each hydrological station from the monitoring data of the hydrological stations; Statistical processing was performed on each of the annual maximum flood peak flow sequences to obtain the mean and standard deviation of each annual maximum flood peak flow sequence; Based on the mean and the standard deviation, the coefficient of variation values ​​for each hydrological station are obtained; Based on the aforementioned coefficient of variation values ​​and the flood design frequency, the initial design peak flood discharge for each hydrological station is calculated using the following formula: in, Let p be the initial design peak flow rate, and p be the flood design frequency. The average peak flow rate. This is the deviation coefficient corresponding to the design frequency of the flood. This represents the coefficient of variation. Obtain the river channel attenuation characteristic parameters, and perform a river-way attenuation correction process on the initial design peak flow based on the river channel attenuation characteristic parameters to obtain the design peak flow value of each node.

4. The method according to claim 1, characterized in that, The design peak flow value based on each node is used to perform segment-by-segment calculations using the energy equation to obtain the design flood water surface line of the target mountain river, including: The hydrological station flow measurement section downstream of the target mountain river is determined as the starting point for the design flood water surface line. The measured water level at the corresponding flood design frequency is extracted from the monitoring data of the hydrological station and used as the initial water level of the starting node; Based on the initial water level of the starting node and the design peak flow value corresponding to the starting node, the design flood level of each node is obtained by performing segment-by-segment calculations from downstream to upstream using Bernoulli's energy equation. Connect the design flood levels of each node to generate the design flood water surface line of the target mountain river.

5. The method according to claim 4, characterized in that, The design flood level of each node is obtained by performing segment-by-segment calculations from downstream to upstream using Bernoulli's energy equation, based on the initial water level of the starting node and the design peak flow value corresponding to the starting node, including: Based on the initial water level of the starting node and the design peak flow value corresponding to the starting node, the design flood level of each node is obtained by performing segment-by-segment calculations from downstream to upstream using Bernoulli's energy equation: in, To design flood level, Here, dv represents the design flood level of the upstream node, and dv represents the change in flow velocity. This is the kinetic energy correction factor. and Let g be the flow velocity and g be the acceleration due to gravity. Let η be the head energy loss, η be the riverbed roughness, L be the node spacing, R be the hydraulic radius, and A be the cross-sectional area of ​​the water passage.

6. The method according to claim 1, characterized in that, The process of dynamically simulating the flood evolution of the target mountain river based on the design flood water surface line yields a dynamic simulation report of the flood evolution of the target mountain river, including: Based on the design peak flow values ​​of each node and the river topography data, the peak propagation velocity is calculated using hydraulic formulas. Based on the aforementioned flood peak propagation speed and the river channel length between the two sections, the flood peak propagation time is calculated. Using the time when the flood peak arrives at the initial section as a reference, the flood peak propagation time of each segment is accumulated to obtain the flood peak propagation time series; Based on the flood peak propagation time series, the design flood water surface line is spatiotemporally connected to obtain spatiotemporally connected water surface line data; Using animation rendering technology, the spatiotemporally connected water surface data is used to generate a flood evolution animation; By overlaying the design flood water surface line with the digital elevation model, the flood inundation depth and range can be obtained. Based on the flood inundation depth and extent, a flood risk level distribution map is generated; By integrating the flood evolution animation with the flood risk level distribution map, a dynamic simulation report of the flood evolution of the target mountain river is generated.

7. The method according to claim 1, characterized in that, The process of fusing the multi-source basic data to obtain a multi-source fused dataset includes: Missing values ​​are filled into the hydrological station monitoring data in the multi-source basic data to obtain a continuous and complete hydrological data sequence. The climate data and remote sensing data in the multi-source basic data are processed into grids to obtain a raster dataset with uniform spatial resolution; The continuous and complete hydrological data sequence, the spatially unified raster dataset, and the river topographic data are spatiotemporally registered to obtain an initial multi-source fusion dataset. The initial multi-source dataset is format-standardized to generate a multi-source fusion dataset.

8. A dynamic flood prediction device for mountain rivers based on big data fusion, characterized in that, The device includes: The data acquisition module is used to acquire multi-source basic data of rivers in the target mountainous area, and to fuse the multi-source basic data to obtain a multi-source fused dataset; the multi-source data includes hydrological station monitoring data, climate data, remote sensing data and river topography data; The flood type identification module is used to identify the flood type of mountain rivers based on the multi-source fusion dataset and through preset threshold rules, so as to obtain the flood type identification result and the set of causal parameters corresponding to the target mountain river; The design flood peak discharge value calculation module is used to calculate the design flood peak discharge corresponding to the target mountain river based on the multi-source fusion dataset, the flood type identification result and the causal parameter set, and to obtain the design flood peak discharge value of each node; the node is a key location with clear terrain features or hydrological monitoring function. The design flood water surface line calculation module is used to obtain the design flood water surface line of the target mountain river by performing segment-by-segment calculations based on the design peak flow values ​​of each node and through the energy equation. The flood evolution dynamic simulation report generation module is used to perform dynamic simulation processing on the flood evolution process of the target mountain river based on the design flood water surface line, and obtain the flood evolution dynamic simulation report of the target mountain river.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images