Regional composite disaster chain risk assessment method under influence of tropical cyclones

Abstract

The invention relates to a regional composite disaster chain risk assessment method under the influence of a tropical cyclone, and the method comprises the steps: obtaining and preprocessing the historical multi-source observation data of the tropical cyclone, and constructing a multi-dimensional disaster element data set; analyzing non-linear change characteristic parameters of each disaster element based on the multi-dimensional disaster element data set, and constructing a multi-dimensional disaster element quantitative model and a comprehensive disaster intensity index based on the non-linear change characteristic parameters; modeling the historical change trajectory of the comprehensive disaster intensity index, extracting an abnormal mode and time-space characteristic parameters, and constructing a composite disaster chain spatio-temporal evolution prediction model; and acquiring real-time multi-source observation data of the tropical cyclones, inputting the real-time multi-source observation data into the composite disaster chain spatio-temporal evolution prediction model to predict a composite disaster chain spatio-temporal evolution trend, performing quantitative analysis on the composite disaster chain spatio-temporal evolution trend, and determining risk levels of different regions. According to the method, quantitative description and evaluation of the tropical cyclone composite disaster chain spatio-temporal evolution process are realized by integrating multi-source heterogeneous data.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of disaster assessment, in particular to a regional compound disaster chain risk assessment method under the influence of tropical cyclones. BACKGROUND

[0002] Tropical cyclones, as one of the most destructive natural disasters in the world, cause huge life and property losses in coastal areas every year. With the intensification of global climate change and the acceleration of urbanization in coastal areas, the phenomenon of compound disaster chain triggered by tropical cyclones is becoming increasingly prominent, and its risk assessment has become a core research direction in the field of disaster prevention and reduction. Current risk assessment methods mainly analyze single disaster types, lacking in-depth understanding of the interaction mechanism between multiple secondary disasters triggered by tropical cyclones. Existing assessment systems often treat disasters such as heavy rain, flood, and wind disaster separately, which cannot accurately reflect the overall risk characteristics of compound disaster chain, resulting in significant deviation between the assessment results and actual disaster losses.

[0003] The key challenge in tropical cyclone compound disaster chain risk assessment lies in the comprehensive quantification of multi-dimensional disaster elements. The tropical cyclone system contains multiple physical elements such as cyclone intensity, rainfall distribution, and wind field structure. These elements exhibit highly complex nonlinear variation characteristics in space and time scales, making it difficult for traditional single-index evaluation methods to fully depict their dynamic evolution process. The difficulty in quantifying multi-dimensional elements further leads to technical bottlenecks in multi-source data fusion. Remote sensing monitoring data, ground observation data, and numerical simulation results from different sources have significant differences in spatial resolution, accuracy standards, and data formats. The lack of an effective data integration framework makes it impossible to establish a unified evaluation benchmark. The problem of data fusion directly restricts the accurate identification of the spatio-temporal evolution characteristics of the disaster chain. Since the quantitative correlation between various secondary disasters cannot be established, existing methods cannot accurately assess the risk level of each link in the compound disaster chain. Therefore, the present application proposes a regional compound disaster chain risk assessment method under the influence of tropical cyclones. SUMMARY

[0004] The purpose of the present application is to provide a regional compound disaster chain risk assessment method under the influence of tropical cyclones, which builds a comprehensive evaluation system covering cyclone intensity classification, rainfall distribution, wind field simulation, and flood inundation range, and realizes quantitative description of the spatio-temporal evolution process of tropical cyclone compound disaster chain by effectively integrating multi-source heterogeneous data.

[0005] To achieve the above purpose, the present application provides the following solutions:

[0006] A regional compound disaster chain risk assessment method under the influence of tropical cyclones, comprising:

[0007] Acquire historical multi-source observation data of tropical cyclones and preprocess, construct multi-dimensional disaster element data set;

[0008] Based on the multi-dimensional disaster element data set, analyze the nonlinear change characteristic parameters of each disaster element, and based on the nonlinear change characteristic parameters, construct a multi-dimensional disaster element quantization model and a comprehensive disaster intensity index;

[0009] Model the historical change trajectory of the comprehensive disaster intensity index, extract abnormal patterns and spatio-temporal characteristic parameters, and construct a complex disaster chain spatio-temporal evolution prediction model;

[0010] Acquire real-time multi-source observation data of tropical cyclones, input the complex disaster chain spatio-temporal evolution prediction model to predict the spatio-temporal evolution trend of the complex disaster chain, and quantitatively analyze the spatio-temporal evolution trend of the complex disaster chain to determine the risk level of different regions, and complete the risk assessment of the complex disaster chain under the influence of tropical cyclones.

[0011] Optionally, acquire historical multi-source observation data of tropical cyclones and preprocess, construct multi-dimensional disaster element data set, including:

[0012] Acquire historical multi-source observation data of tropical cyclones from satellite remote sensing, ground meteorological stations and ocean buoys, including cyclone intensity data, rainfall distribution data and wind field structure data;

[0013] After spatio-temporal registration, interpolation and correction processing of the historical multi-source observation data, the characteristic parameters of the cyclone intensity, rainfall distribution and wind field structure are extracted respectively, and a characteristic parameter set is obtained;

[0014] Based on the characteristic parameter set, the cyclone intensity data and wind field structure data are comprehensively analyzed by using random forest algorithm to determine the core influence area and intensity level distribution of the cyclone;

[0015] Through the core influence area and intensity level distribution of the cyclone, combined with the rainfall distribution data, spatial superposition analysis is carried out to obtain a risk distribution layer and fuse disaster element data, and the multi-dimensional disaster element data set is constructed.

[0016] Optionally, based on the multi-dimensional disaster element data set, the nonlinear change characteristic parameters of each disaster element are analyzed, including:

[0017] The multi-dimensional disaster element data set is analyzed by using a nonlinear change analysis method to obtain preliminary nonlinear change characteristic parameters and dynamic change description of each disaster element;

[0018] Based on the dynamic change description, the change rate value of cyclone intensity is calculated, and the mutation event is determined through the change rate value to obtain the mutation event identification of cyclone intensity;

[0019] Based on the mutation event identification, the spatial pattern of rainfall distribution is analyzed in combination with the spatial pattern of rainfall distribution at the time of the mutation event, and a nonlinear variation parameter of rainfall distribution is obtained;

[0020] Based on the nonlinear variation parameter, the change characteristics of the wind field structure under the influence of the mutation event are extracted in combination with the evolution trajectory of the wind field structure, and a dynamic evolution mode of the wind field structure is obtained;

[0021] The correlation influence degree of each disaster element is determined by the mutation event identification, the nonlinear variation parameter, and the dynamic evolution mode, and the preliminary nonlinear variation characteristic parameter of each disaster element is integrated and processed based on the correlation influence degree, and a final nonlinear variation characteristic parameter is obtained.

[0022] Optionally, a multi-dimensional disaster element quantification model and a comprehensive disaster intensity index are constructed based on the nonlinear variation characteristic parameter, including:

[0023] The nonlinear variation characteristic parameter is standardized to obtain a disaster element matrix, i.e. the multi-dimensional disaster element quantification model;

[0024] The disaster element matrix is weighted and fused to obtain a comprehensive disaster intensity value and to determine a comprehensive disaster intensity grade distribution, and if the comprehensive disaster intensity grade distribution exceeds a preset grade threshold, the weight coefficient is adjusted and optimized, the comprehensive disaster intensity value is recalculated until the preset threshold range is met, and a comprehensive disaster intensity index is obtained.

[0025] Optionally, the historical change trajectory of the comprehensive disaster intensity index is modeled, and an abnormal mode and a spatiotemporal characteristic parameter are extracted to construct a complex disaster chain spatiotemporal evolution prediction model, including:

[0026] Long-term recorded data of the comprehensive disaster intensity index is obtained, and a time series analysis method is used to extract the trajectory change characteristics of the comprehensive disaster intensity, to determine key time nodes and spatial distribution rules;

[0027] Based on the key time nodes and spatial distribution rules, the deviation value of the current time comprehensive disaster intensity index from the historical same period data is calculated to obtain a deviation distribution result;

[0028] If the deviation distribution result exceeds a preset statistical threshold, the current time comprehensive disaster intensity index is determined to be an abnormal mode, and the spatiotemporal characteristic parameter of the abnormal mode is obtained;

[0029] The complex disaster chain spatiotemporal evolution prediction model is constructed in combination with the spatiotemporal characteristic parameter and the long-term recorded data.

[0030] Optionally, real-time multi-source observation data of the tropical cyclone is acquired, and the complex disaster chain space-time evolution prediction model is input to predict the space-time evolution trend of the complex disaster chain, including:

[0031] Real-time multi-source observation data of the tropical cyclone is acquired and integrated to obtain a fusion data set;

[0032] The fusion data set is subjected to structured processing to determine the space-time distribution pattern of disaster chain related features, and abnormal features are determined according to the space-time distribution pattern, and the weight distribution of key influencing factors in the abnormal features is acquired by using a random forest algorithm;

[0033] The parameters of the complex disaster chain space-time evolution prediction model are updated based on the weight distribution of the key influencing factors, the fusion data set is input into the updated complex disaster chain space-time evolution prediction model, and the space-time evolution trend of the complex disaster chain is acquired.

[0034] Optionally, the space-time evolution trend of the complex disaster chain is subjected to quantitative analysis to determine the risk levels of different regions, including:

[0035] The influence range is determined based on the space-time evolution trend of the complex disaster chain;

[0036] The population density of each region is acquired, and the overlap degree of the influence range and each region is analyzed;

[0037] The risk levels of each region are determined based on the population density and the overlap degree.

[0038] Optionally, the method further includes predicting the disaster loss of each region according to a preset risk loss evaluation system, wherein the risk loss evaluation system is constructed by the correlation between historical disaster loss data and risk levels.

[0039] The present application has the following advantages:

[0040] The present application acquires multi-source observation data and performs space-time registration and standardization processing, analyzes the evolution characteristics of disaster elements by using a nonlinear change feature extraction method, constructs a multi-dimensional disaster element quantization model and fuses it into a comprehensive disaster intensity index, models and predicts the disaster intensity index by using a space-time evolution analysis algorithm, dynamically adjusts the model parameters in combination with real-time monitoring data, realizes the prediction of the evolution trend of the complex disaster chain, and further quantitatively analyzes the prediction result by using a risk evaluation algorithm to determine the risk level distribution of different regions. The present application can effectively evaluate the risk of the complex disaster chain caused by the tropical cyclone, and provide a scientific basis for disaster prevention and mitigation decision-making. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings described below only illustrate some of the embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative labor.

[0042] Figure 1 A flow chart of a regional composite disaster chain risk assessment method under the influence of a tropical cyclone. DETAILED DESCRIPTION

[0043] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0044] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0045] The present embodiment provides a regional composite disaster chain risk assessment method under the influence of a tropical cyclone, as shown in Figure 1 , comprising:

[0046] Obtain historical multi-source observation data of a tropical cyclone and preprocess, and construct a multi-dimensional disaster element data set;

[0047] Based on the multi-dimensional disaster element data set, analyze the nonlinear variation characteristic parameters of each disaster element, and based on the nonlinear variation characteristic parameters, construct a multi-dimensional disaster element quantization model and a comprehensive disaster intensity index;

[0048] Model the historical variation trajectory of the comprehensive disaster intensity index, extract abnormal patterns and spatio-temporal characteristic parameters, and construct a composite disaster chain spatio-temporal evolution prediction model;

[0049] Obtain real-time multi-source observation data of a tropical cyclone, input the composite disaster chain spatio-temporal evolution prediction model to predict the spatio-temporal evolution trend of the composite disaster chain, and quantitatively analyze the spatio-temporal evolution trend of the composite disaster chain to determine the risk level of different regions, and complete the regional composite disaster chain risk assessment under the influence of a tropical cyclone.

[0050] Specifically, the embodiment obtains multi-source observation data, performs spatio-temporal registration and standardization processing, analyzes the evolution characteristics of disaster elements by using a nonlinear change feature extraction method, constructs a multi-dimensional disaster element quantization model, and fuses it into a comprehensive disaster intensity index; the disaster intensity index is modeled and predicted by using a spatio-temporal evolution analysis algorithm, the model parameters are dynamically adjusted combined with real-time monitoring data, the evolution trend of the compound disaster chain is predicted, and the risk grade distribution of different regions is determined by quantitatively analyzing and predicting the results by using a risk assessment algorithm. The embodiment can effectively evaluate the risk of the compound disaster chain caused by a tropical cyclone, and provide a scientific basis for disaster prevention and mitigation decision-making.

[0051] Further, the historical multi-source observation data of the tropical cyclone are obtained and preprocessed, and a multi-dimensional disaster element data set is constructed, including:

[0052] The historical multi-source observation data of the tropical cyclone are obtained from satellite remote sensing, ground meteorological stations and ocean buoys, including cyclone intensity data, rainfall distribution data and wind field structure data;

[0053] After the spatio-temporal registration, interpolation and correction processing of the historical multi-source observation data, the feature parameters of the cyclone intensity, rainfall distribution and wind field structure are extracted respectively, and a feature parameter set is obtained;

[0054] Based on the feature parameter set, the cyclone intensity data and the wind field structure data are comprehensively analyzed by using a random forest algorithm, the core influence area and the intensity grade distribution of the cyclone are determined;

[0055] Through the core influence area and the intensity grade distribution of the cyclone, combined with the rainfall distribution data, spatial superposition analysis is performed, a risk distribution layer is obtained, and disaster element data are fused, and the multi-dimensional disaster element data set is constructed.

[0056] Specifically, the embodiment first obtains the cloud system structure and intensity distribution data of the cyclone through satellite remote sensing images, such as using the infrared channel data of Fengyun-4 satellite, extracts the area with a cloud top temperature lower than -60°C as the cyclone core area, calculates the reflectivity by combining the visible light channel, and estimates the cyclone intensity to be 12 (wind speed about 33 m / s). At the same time, the ground meteorological station records the pressure and wind speed data updated every hour, such as a station recording a pressure of 980 hundred pascals and a wind speed of 25 m / s, and generates a regional pressure field by an interpolation algorithm (such as inverse distance weighted method) with an error controlled within ±2 hundred pascals. The ocean buoy monitoring information provides sea surface temperature and wave height data, such as a buoy recording a sea temperature of 28.5°C and a wave height of 3.2 meters, and predicts the cyclone path deviation trend by time series analysis, and optimizes the prediction accuracy to ±5 kilometers by combining the Kalman filter algorithm. Subsequently, for spatio-temporal registration, the data of different resolutions and coordinate systems are unified, such as satellite images with a resolution of 1 km, ground station data with point-like distribution, and ocean buoy data with a time interval of 30 minutes, and the spatial resolution is unified to 500 m grid by Kriging interpolation method, the time interval is unified to every hour, the projection conversion is carried out by using WGS84 coordinate system, and the error is controlled within ±0.5 meters, so as to ensure the consistency of the data. Finally, a standardized multi-dimensional disaster factor data set is generated, and the cyclone intensity (wind speed 33 m / s), rainfall distribution (hourly rainfall 50 mm, based on satellite inversion and ground data fusion) and wind field structure (based on Doppler radar wind speed vector field, maximum wind speed radius about 50 km) are integrated into a NetCDF format file.

[0057] Further, based on the multi-dimensional disaster factor data set, the non-linear variation characteristic parameters of each disaster factor are analyzed, including:

[0058] The multi-dimensional disaster factor data set is analyzed by using a non-linear variation analysis method to obtain preliminary non-linear variation characteristic parameters and dynamic change description of each disaster factor;

[0059] Based on the dynamic change description, the change rate value of the cyclone intensity is calculated, and the mutation event is judged by the change rate value to obtain the mutation event identification of the cyclone intensity;

[0060] Based on the mutation event identification, the spatial characteristics of the rainfall distribution at the time of the mutation event are analyzed in combination with the spatial pattern of the rainfall distribution to obtain the non-linear variation parameters of the rainfall distribution;

[0061] Based on the non-linear variation parameters, the change characteristics of the wind field structure under the influence of the mutation event are extracted in combination with the evolution track of the wind field structure to obtain the dynamic evolution mode of the wind field structure;

[0062] The correlation influence degree of each disaster element is determined by the mutation event, the nonlinear change parameter, and the dynamic evolution mode, and the preliminary nonlinear change characteristic parameter of each disaster element is integrated based on the correlation influence degree to obtain a final nonlinear change characteristic parameter.

[0063] Specifically, the cyclone intensity time series data is first processed by using a wavelet transform method. It is assumed that the data set contains cyclone intensity values recorded every hour, the time span is 72 hours, the intensity value range is between 20 and 80 meters per second, the signal is decomposed to level 3 by using discrete wavelet transform (DWT), the high-frequency component is extracted by using a Daubechies wavelet basis function, the wavelet coefficients at each time point are calculated, it is found that the intensity change rate reaches 5.2 meters per second per hour at the 24th hour, which exceeds the preset threshold of 3.0 meters per second per hour, the system automatically marks it as an intensity mutation event, and records the mutation time and the correlation coefficient 0.85 as characteristic parameters. Then, the rainfall distribution spatial pattern is analyzed, the rainfall grid data (resolution is 0.1 degree x 0.1 degree, coverage is 1000 kilometers x 1000 kilometers) is processed by using principal component analysis (PCA) combined with a nonlinear kernel function (such as a radial basis function kernel), the first two principal components are extracted, the contribution rates are 65% and 20% respectively, the spatial gradient change rate of the principal components is calculated, it is found that the central area rainfall intensity gradient is 12.5 millimeters per hour per kilometer, which represents the nonlinear concentration characteristics of the rainfall distribution, and the parameter table is generated to store the gradient value and the distribution pattern coefficient 0.72. Subsequently, the wind field structure evolution trajectory is analyzed, the Lyapunov exponent calculation method in chaos theory is adopted, the wind field vector data (updated every 6 hours, wind speed range is 5 to 50 meters per second) is input, the phase space is constructed by using the time delay embedding method, the delay time is set to 3 hours, the embedding dimension is 5, the maximum Lyapunov exponent is 0.15, which indicates that the wind field evolution has strong nonlinear chaotic characteristics, and the system automatically saves the exponent value and the trajectory fractal dimension 1.8 as characteristic parameters. Finally, the above cyclone intensity mutation event, rainfall distribution characteristic parameters and wind field chaos index are integrated into a disaster element database to form a nonlinear change characteristic parameter set.

[0064] Further, a multi-dimensional disaster element quantization model and a comprehensive disaster intensity index are constructed based on the nonlinear change characteristic parameters, including:

[0065] The nonlinear change characteristic parameters are standardized to obtain a disaster element matrix, that is, the multi-dimensional disaster element quantization model;

[0066] The disaster element matrix is subjected to weighted fusion calculation to obtain a comprehensive disaster intensity value and to judge a comprehensive disaster intensity level distribution. If the comprehensive disaster intensity level distribution exceeds a preset level threshold, the weight coefficient is adjusted and optimized, the comprehensive disaster intensity value is recalculated until the preset threshold range is met, and a comprehensive disaster intensity index is obtained.

[0067] Specifically, for the quantification of the cyclone intensity index, based on historical meteorological data, the cyclone center pressure value is 960 hundred pascals, which is converted into an intensity index using a standardization formula. The calculation formula is: index = (standard pressure value - actual pressure value) / standard deviation, where the standard pressure value is 1013 hundred pascals and the standard deviation is 20 hundred pascals. The calculated index is (1013-960) / 20 = 2.65, indicating that the cyclone intensity is high. Next, for the rainfall intensity distribution matrix, assuming that the rainfall data in a certain area is a 5x5 grid matrix, the center point rainfall is 100 millimeters, and the surrounding gradually decreases to 20 millimeters. A spatial interpolation algorithm (such as inverse distance weighting) is used to generate a smooth distribution matrix, and the weight of each grid point is calculated. The center point weight is 1, and the surrounding weight decreases inversely with the square of the distance. The maximum value of the standardized rainfall intensity matrix is 0.85. Then, for the wind field velocity vector field, assuming that the wind speed at a certain point is 25 meters per second and the wind direction is northeast 45 degrees, the vector is decomposed into x-axis component 17.68 meters per second and y-axis component 17.68 meters per second. Combined with the data of multiple observation points in the region, a continuous wind field distribution map is generated using the Kriging interpolation method, and the average wind speed intensity is calculated as 0.78. Finally, the above indexes are integrated using a weighted fusion algorithm. The cyclone intensity index weight is set to 0.4, the rainfall intensity weight is set to 0.3, and the wind field speed weight is set to 0.3. The comprehensive disaster intensity index = 2.65x0.4 + 0.85x0.3 + 0.78x0.3 = 1.06 + 0.255 + 0.234 = 1.549, and the evaluation value is 1.549, indicating that the disaster intensity is at a medium-high level.

[0068] Further, the historical change trajectory of the comprehensive disaster intensity index is modeled, and abnormal patterns and spatiotemporal feature parameters are extracted to construct a complex disaster chain spatiotemporal evolution prediction model, including:

[0069] Long-term recorded data of the comprehensive disaster intensity index is obtained, and time series analysis method is used to extract the feature of the trajectory change of the comprehensive disaster intensity, to determine the key time nodes and spatial distribution rules;

[0070] Based on the key time nodes and spatial distribution rules, the deviation value of the current time comprehensive disaster intensity index from the historical same period data is calculated to obtain the deviation distribution result;

[0071] If the deviation distribution result exceeds the preset statistical threshold, it is determined that the comprehensive disaster intensity index at the current time is an abnormal pattern, and the spatio-temporal characteristic parameters of the abnormal pattern are obtained;

[0072] Combined with the spatio-temporal characteristic parameters and long-term record data, the spatio-temporal evolution prediction model of the compound disaster chain is constructed.

[0073] Specifically, the historical data of the disaster intensity index is first processed by a spatio-temporal evolution analysis algorithm, and the disaster intensity index data of the past 10 years is collected, which is recorded monthly every year, a total of 120 data points, and the data range is between 0 and 100. For example, the disaster intensity index in January 2020 is 45.5, and the disaster intensity index in January 2021 is 48.2, and so on. An ARIMA model (autoregressive integrated moving average model) in time series analysis is used for modeling, and the parameters are set to (p=1, d=1, q=1). The predicted value and residual error of each time point are obtained by fitting the historical data, and the mean and standard deviation of the historical same period are calculated. For example, the historical mean in January is 46.8, and the standard deviation is 2.5. Then, the statistical threshold is set to the mean plus or minus 2 times the standard deviation, i.e. the threshold range is 41.8 to 51.8. If the disaster intensity index in January 2023 is 55.3, it is out of the threshold range, and it is determined to be an abnormal evolution pattern. Further, by analyzing the time distribution and spatial distribution of the abnormal points, combining with the geographic information system (GIS) technology, the disaster intensity index is associated with the geographic position, and it is assumed that the abnormal points are concentrated in a certain area (such as coordinates range east longitude 120-122 degrees, north latitude 30-32 degrees), and the spatial interpolation algorithm (such as Kriging interpolation) is used to generate the disaster intensity heat map, and it is found that the intensity abnormal area coincides with the high incidence area of the historical disaster chain (such as flood-landslide). Finally, based on the above analysis, a spatio-temporal evolution prediction model of the disaster chain is constructed, a Bayesian network algorithm is used, historical disaster chain data and current abnormal index are input, and model parameters are output, for example, the disaster chain trigger probability is 0.75, the propagation speed parameter is 0.3 units / day, and it is predicted that the disaster chain may expand to the east longitude 123 degrees and north latitude 33 degrees area in the next month.

[0074] Further, real-time multi-source observation data of tropical cyclones is obtained, and the spatio-temporal evolution prediction model of the compound disaster chain is input to predict the spatio-temporal evolution trend of the compound disaster chain, including:

[0075] Real-time multi-source observation data of tropical cyclones is obtained and integrated to obtain a fusion data set;

[0076] The fusion data set is structurally processed to determine the spatio-temporal distribution pattern of the disaster chain related characteristics, and the abnormal characteristics are determined according to the spatio-temporal distribution pattern, and the weight distribution of the key influencing factors in the abnormal characteristics is obtained by using a random forest algorithm;

[0077] updating parameters of the complex disaster chain spatiotemporal evolution prediction model based on the weight distribution of the key influencing factors, inputting the fused data set into the updated complex disaster chain spatiotemporal evolution prediction model, and obtaining a complex disaster chain spatiotemporal evolution trend.

[0078] Specifically, the embodiment first obtains cyclone path information from satellite remote sensing data, assumes that the current cyclone center position is 120.5 degrees east longitude and 25.3 degrees north latitude, the moving speed is 15 kilometers per hour, and the direction is northwest by west, performs smoothing processing on the cyclone path by using a Kalman filtering algorithm, combines historical trajectory data, and calculates a predicted position in the next 6 hours as 119.8 degrees east longitude and 26.1 degrees north latitude, with an error range controlled within ±0.2 degrees. Then, rainfall cloud image data is fused, rainfall intensity distribution is obtained by using radar monitoring, it is assumed that the rainfall in the current core area is 50 millimeters per hour, cloud image features are extracted by using a convolutional neural network model, it is predicted that the rainfall in the next 3 hours will increase to 60 millimeters per hour, and the coverage range will expand to a radius of 100 kilometers. At the same time, wind field distribution information is collected, it is assumed that the current maximum wind speed is 30 meters per second, and the wind field radius is 200 kilometers, based on a numerical weather prediction model WRF, the boundary conditions are updated in combination with real-time wind speed data, it is predicted that the wind speed in the next 12 hours will increase to 35 meters per second, and the wind field radius will expand to 250 kilometers. Subsequently, the above multi-source data is input into a spatiotemporal evolution prediction model, a long short-term memory network (LSTM) algorithm is used, parameters of cyclone path, rainfall, and wind field are combined, model weights are dynamically adjusted, an evolution trend of a complex disaster chain is predicted, and it is concluded that a storm surge may be caused by the cyclone in the next 24 hours, affecting the coastal area, the storm surge height is expected to be 1.5 meters, and the influence range is about 50 kilometers of coastline. Finally, by using a Bayesian updating method, model parameters are corrected every hour according to real-time monitoring data, for example, the cyclone moving speed error is adjusted from ±0.2 degrees to ±0.15 degrees, the prediction accuracy is improved, and the reliability of the evolution trend prediction result of the disaster chain is ensured.

[0079] Further, the complex disaster chain spatiotemporal evolution trend is quantitatively analyzed to determine risk levels of different regions, including:

[0080] determining an influence range based on the complex disaster chain spatiotemporal evolution trend;

[0081] obtaining population densities of the regions and analyzing overlapping degrees of the influence range and the regions;

[0082] judging risk levels of the regions based on the population densities and the overlapping degrees.

[0083] Further, the method further includes predicting disaster losses of the regions according to a preset risk loss evaluation system, wherein the risk loss evaluation system is constructed by correlation between historical disaster loss data and risk levels.

[0084] Specifically, the embodiment first obtains disaster influence range data through a disaster chain evolution prediction model, assumes that the occurrence probability of a flood and landslide compound disaster in a certain area is 0.75 and the influence range coverage radius is 10 kilometers, and the center point coordinates are (X: 120.5, Y: 30.2) by using a Bayesian network model based on historical data. Then, a spatial overlay analysis algorithm is used to compare the predicted influence range with geographic information data of a population-dense area, assumes that the center point coordinates of the population-dense area are (X: 120.6, Y: 30.1), and the overlap area proportion of the two areas is 0.65 by calculation, and the preset danger threshold is 0.5. Since the overlap degree exceeds the threshold, it is automatically determined that the area is a potential high-risk area. Subsequently, a risk assessment algorithm is used to further quantify the risk level, combines the disaster occurrence probability 0.75, the population density 5000 per square kilometer, and the overlap area proportion 0.65, and calculates the risk index as 0.75 x 0.65 x 5000 = 2437.5. Assuming that the risk level classification standard is: the index is greater than 2000 for high risk, 1000 to 2000 for medium risk, and less than 1000 for low risk, the area is classified as high risk. Finally, a risk level distribution map is generated based on geographic information system (GIS) technology, the risk indexes of all areas are spatially visualized, assumes that there are 10 sub-areas, of which 3 risk indexes exceed 2000 and are marked as red high risk, 5 are between 1000 and 2000 and are marked as yellow medium risk, and 2 are below 1000 and are marked as green low risk, and the distribution map is output and stored as a vector format file.

[0085] The above-described embodiments are only descriptions of the preferred modes of the present application and do not limit the scope of the present application. Without departing from the design spirit of the present application, various modifications and improvements to the technical solutions of the present application made by those of ordinary skill in the art shall fall within the protection scope of the present application as defined by the claims.

Claims

1. A method for assessing the risk of a regional composite disaster chain under the influence of a tropical cyclone, characterized in that: include: Obtain historical multi-source observation data on tropical cyclones and pre-process them to construct a multi-dimensional disaster factor dataset; Analyzing nonlinear change characteristic parameters of each disaster factor based on the multidimensional disaster factor data set, and constructing a multidimensional disaster factor quantification model and a comprehensive disaster intensity index based on the nonlinear change characteristic parameters; Modeling the historical change trajectory of the comprehensive disaster intensity index, extracting abnormal patterns and spatiotemporal characteristic parameters, and constructing a spatiotemporal evolution prediction model for the composite disaster chain; Real-time multi-source observation data of tropical cyclones are obtained, and the data are input into the spatiotemporal evolution prediction model of the composite disaster chain to predict the spatiotemporal evolution trend of the composite disaster chain. The spatiotemporal evolution trend of the composite disaster chain is quantitatively analyzed to determine the risk levels of different regions and complete the regional composite disaster chain risk assessment under the influence of tropical cyclones.

2. The method for assessing the risk of a regional composite disaster chain under the influence of a tropical cyclone according to claim 1, characterized in that: Historical multi-source observation data of tropical cyclones are obtained and preprocessed to construct a multi-dimensional disaster factor dataset, including: Acquire historical multi-source observational data on tropical cyclones from satellite remote sensing, ground-based meteorological stations, and ocean buoys, including cyclone intensity data, rainfall distribution data, and wind field structure data; After performing spatiotemporal registration, interpolation, and correction processing on the historical multi-source observation data, characteristic parameters of the cyclone intensity, rainfall distribution, and wind field structure are extracted respectively to obtain a characteristic parameter set; Based on the characteristic parameter set, a random forest algorithm is used to comprehensively analyze the cyclone intensity data and wind field structure data to determine the core impact area and intensity level distribution of the cyclone; The multidimensional disaster element data set is constructed by performing spatial overlay analysis based on the core impact area and intensity level distribution of the cyclone and combining it with rainfall distribution data to obtain the risk distribution layer and fuse the disaster element data.

3. The method for assessing the risk of a regional composite disaster chain under the influence of a tropical cyclone according to claim 1, wherein: Analyzing the nonlinear change characteristic parameters of each disaster factor based on the multidimensional disaster factor dataset includes: Using a nonlinear change analysis method to analyze the multidimensional disaster factor dataset, and obtaining preliminary nonlinear change characteristic parameters and dynamic change descriptions of each disaster factor; Based on the dynamic change description, a change rate value of the cyclone intensity is calculated, and a mutation event is determined by the change rate value to obtain a mutation event identifier of the cyclone intensity; Based on the mutation event identifier, combined with the spatial pattern of rainfall distribution, the spatial characteristics of rainfall distribution when the mutation event occurs are analyzed to obtain nonlinear change parameters of rainfall distribution; Based on the nonlinear change parameters and combined with the evolution trajectory of the wind farm structure, the change characteristics of the wind farm structure under the influence of the sudden change event are extracted to obtain the dynamic evolution pattern of the wind farm structure; The correlation influence degree of each disaster factor is determined by the mutation event identifier, nonlinear change parameter, and dynamic evolution mode, and the preliminary nonlinear change characteristic parameters of each disaster factor are integrated based on the correlation influence degree to obtain the final nonlinear change characteristic parameters.

4. The method for risk assessment of a regional composite disaster chain under the influence of a tropical cyclone according to claim 1, characterized in that: A multi-dimensional disaster factor quantitative model and a comprehensive disaster intensity index are constructed based on the nonlinear change characteristic parameters, including: Standardizing the nonlinear change characteristic parameters to obtain a disaster factor matrix, i.e., the multidimensional disaster factor quantification model; A weighted fusion calculation is performed on the disaster factor matrix to obtain a comprehensive disaster intensity value and determine the comprehensive disaster intensity level distribution. If the comprehensive disaster intensity level distribution exceeds the preset level threshold, the weight coefficient is adjusted and optimized, and the comprehensive disaster intensity value is recalculated until it meets the preset threshold range to obtain a comprehensive disaster intensity index.

5. The method for risk assessment of regional composite disaster chains under the influence of tropical cyclones according to claim 1, characterized in that: The historical change trajectory of the comprehensive disaster intensity index is modeled, and abnormal patterns and spatiotemporal characteristic parameters are extracted to construct a spatiotemporal evolution prediction model for the composite disaster chain, including: Obtain long-term recorded data of comprehensive disaster intensity indicators, use time series analysis methods to extract features of the trajectory changes of comprehensive disaster intensity, and determine key time nodes and spatial distribution patterns; Based on key time nodes and spatial distribution patterns, calculate the deviation between the current comprehensive disaster intensity index and historical data for the same period, and obtain the deviation distribution results; If the deviation distribution result exceeds the preset statistical threshold, the comprehensive disaster intensity index at the current moment is determined to be an abnormal mode, and the spatiotemporal characteristic parameters of the abnormal mode are obtained; Combining the spatiotemporal characteristic parameters and long-term recorded data, a spatiotemporal evolution prediction model of the composite disaster chain is constructed.

6. The method for assessing the risk of a regional composite disaster chain under the influence of a tropical cyclone according to claim 1, characterized in that: Acquire real-time multi-source observation data of tropical cyclones and input them into the spatiotemporal evolution prediction model of the composite disaster chain to predict the spatiotemporal evolution trend of the composite disaster chain, including: Acquire and integrate real-time multi-source observation data of tropical cyclones to obtain a fused dataset; Performing structured processing on the fused dataset to determine the spatiotemporal distribution pattern of disaster chain-related features, judging abnormal features based on the spatiotemporal distribution pattern, and obtaining the weight distribution of key influencing factors in the abnormal features using a random forest algorithm; Based on the weight distribution of the key influencing factors, the parameters of the composite disaster chain spatiotemporal evolution prediction model are updated, and the fusion data set is input into the updated composite disaster chain spatiotemporal evolution prediction model to obtain the spatiotemporal evolution trend of the composite disaster chain.

7. The method for risk assessment of a regional composite disaster chain under the influence of a tropical cyclone according to claim 1, characterized in that: Quantitatively analyze the spatiotemporal evolution trend of the complex disaster chain to determine the risk levels of different regions, including: Determine the impact scope based on the spatiotemporal evolution trend of the composite disaster chain; Obtain the population density of each region and analyze the degree of overlap between the impact area and each region; Based on the population density and degree of overlap, the risk level of each area is determined.

8. The method for assessing the risk of a regional composite disaster chain under the influence of a tropical cyclone according to any one of claims 1 to 7, characterized in that: The method further includes: predicting disaster losses in each region according to a preset risk loss assessment system, wherein the risk loss assessment system is constructed by correlation between historical disaster loss data and risk levels.

Images