Multi-disaster chain type marine disaster risk assessment method based on regional disaster-bearing body distribution

Abstract

The invention discloses a multi-disaster chain type marine disaster risk assessment method based on regional disaster-bearing body distribution, relates to the technical field of marine disaster prevention and control, and aims to clarify the range of a research region, extract all disaster-bearing bodies in the research region, analyze the interaction relationship of the disaster-bearing bodies and establish an association probability matrix. By integrating a disaster chain theory, a probability statistical model and GIS spatial analysis, systematic identification and quantification of multi-disaster chain type marine disaster risks are realized, a logical relationship among a native event, a secondary event and a derivative event is defined, a disaster chain type structure chart is constructed, the total probability of a disaster-bearing body is calculated through a probability iterative algorithm, and the risk of the multi-disaster chain type marine disaster risk is identified and quantified. According to the method, direct and indirect conduction risks are comprehensively reflected, individual risk values are calculated in combination with disaster-bearing body vulnerability, and disaster chain overall risk values are generated through superposition, so that the limitation of isolated analysis of a single disaster in a traditional method is avoided, and complex interaction and cumulative effects of a disaster chain can be more comprehensively captured.

Description

Technical Field

[0001] The present invention relates to the technical field of marine disaster prevention and control, and in particular to a multi-hazard chain marine disaster risk assessment method based on regional disaster-bearing body distribution. Background Art

[0002] Regional disaster-prone bodies refer to various objects or systems that may be affected by natural disasters in a specific area, including ecosystems, infrastructure, socio-economic activities, population density, etc., which may be damaged or affected when a disaster occurs. In offshore areas, they are susceptible to marine disasters, including typhoons, storm surges, tsunamis, red tides, sea level rise, coastal erosion, etc. Due to the chain transmission effect between marine disasters, a single disaster may trigger secondary or derivative disasters. Therefore, risk assessment is often conducted through multi-hazard chain marine disaster risk assessment to target the interrelationships and impacts of multiple marine disasters when they occur.

[0003] In the existing technology, multi-hazard chain assessment often uses a simple conditional probability model, which easily ignores the nonlinear interactions between hazards and makes it difficult to quantify the mutation effects of critical points in the disaster chain. Therefore, how to analyze the comprehensive risk of the disaster chain based on the risks of all hazard-prone objects in the disaster chain and determine the regional risk based on the distribution of hazard-prone objects in the region is the problem to be solved by the present invention. To this end, a multi-hazard chain marine disaster risk assessment method based on the distribution of regional hazard-prone objects is proposed. Summary of the Invention

[0004] The present invention aims to provide a multi-hazard chain marine disaster risk assessment method based on the distribution of regional disaster-prone bodies to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: The multi-hazard chain marine disaster risk assessment method based on the regional hazard distribution includes the following steps: S1. Define the scope of the study area, extract all hazard-prone bodies within the study area, analyze their interactions, and establish an association probability matrix; S2. Identify the chain effects between disaster events, clarify the primary events, secondary events, derivative events and the disaster chain structure diagram; S3. Analyze the probability of direct effects of marine disaster-causing factors on each hazard-bearing body, that is, the probability of occurrence of primary events; S4. Based on the interaction between hazard-bearing bodies, calculate the marginal probability and correlation probability of secondary or derivative events induced between hazard-bearing bodies; S5. Use the probability iteration algorithm to calculate the point probability and edge probability based on the probability of the original event, and calculate the total probability of each hazard-bearing body layer by layer; S6. Combine the total probability of the hazard-bearing body with its losses to calculate the risk of the hazard-bearing body, and obtain the overall risk value of the disaster chain through superposition; S7. Calculate and determine the overall regional risk based on the distribution of hazard-prone objects within the study area.

[0006] A further improvement of the technical solution of the present invention is that in S1, the process of establishing the association probability matrix is: Use maps to clearly define the geographical boundaries of the study area, use GIS to extract all hazard-prone objects in the study area, classify and record their attributes, and form a hazard-prone object list; Based on the disaster chain theory, the physical, functional or spatial connections between hazard-prone objects are identified, the interaction logic between hazard-prone objects is determined through historical data, and a directed network diagram is constructed to visualize the relationship path; The hierarchical analysis method is used to calculate the association probability, ensure the consistency of proportion, analyze the interaction intensity between the disaster-bearing bodies, and establish the association probability matrix.

[0007] A further improvement of the technical solution of the present invention is that: S2 specifically includes: Based on historical disaster data and disaster-causing mechanisms, we analyze the chain structure of disaster events, identify primary events and the disaster-affected bodies they directly affect, and analyze the secondary and derivative events triggered by changes in the state of the disaster-affected bodies. We then construct a disaster chain logic diagram through an event tree to clarify the causal transmission paths and hierarchical structure between events. Primary events, secondary events, derivative events and their associated disaster-prone bodies are abstracted as directed network nodes, and the chain effects between events are converted into directed edges. Graph theory methods are used to calculate the betweenness centrality and identify key hub events. At the same time, disaster transmission paths with high probability or high loss are identified as vulnerable paths to form a disaster chain structure diagram.

[0008] A further improvement of the technical solution of the present invention is that: S3 specifically includes: Based on historical disaster data, the intensity, frequency, and spatial distribution characteristics of disaster-causing factors are analyzed. The probability of occurrence of disaster-causing factors of different intensities is calculated using a probability statistical method based on extreme value analysis. This forms a hazard level map. Combined with GIS spatial overlay, the probability distribution of disaster-causing factors in the study area is determined. Based on the type of hazard-prone objects and their physical properties, a sensitivity index system is constructed. The response probability of hazard-prone objects is analyzed through vulnerability curves, and the degree of response of hazard-prone objects to hazard factors of different intensities is quantified. Based on the multiplicative relationship between hazard (probability of occurrence of disaster-causing factors) and sensitivity (probability of response of disaster-affected bodies), the direct action probability is calculated, and a spatial explicit method based on raster operations is used to generate a primary event occurrence probability matrix pixel by pixel.

[0009] A further improvement of the technical solution of the present invention is that: S4 specifically includes: Collect data on the attributes, spatial locations, and historical disaster events of hazard-prone objects, clarify the interaction relationships between hazard-prone objects, and based on the interaction mechanism of hazard-prone objects and combined with the collected data, determine the factors affecting the marginal probability, including the intensity of the hazard-causing factor and the sensitivity of the hazard-prone object. Use a probabilistic statistical model to quantify the impact of each factor on the marginal probability, and calculate the marginal probability of secondary or derivative events induced by hazard-prone objects. Based on the marginal probability calculation results, the correlation between disaster-prone bodies is analyzed, and the probability of secondary or derivative events induced by disaster-prone bodies is analyzed using probabilistic reasoning methods.

[0010] A further improvement of the technical solution of the present invention is that in S5, the process of calculating the total probability of each disaster-prone body by stacking layer by layer is as follows: Based on historical disaster data and disaster chain theory, the initial occurrence probability of the primary event is determined, and the marginal probability of the primary event on the directly related disaster-bearing body is quantified through regression analysis. The initial probability transfer matrix is ​​constructed to clarify the direct path of disaster transmission and its basic occurrence probability. Using a forward iterative algorithm, starting from the primary event, the point probability of the secondary disaster-prone body is calculated layer by layer. The point probability of each node is equal to the sum of the product of the point probabilities of all its predecessor nodes and the corresponding edge probabilities. Efficient propagation is achieved through matrix operations until all disaster-prone bodies are covered, forming a complete disaster transmission probability field. The iterative results are checked for steady-state conditions. The calculation is terminated when the point probability change between two consecutive iterations is less than the preset probability threshold to ensure the convergence of the probability distribution. The total probability of each disaster-prone body, i.e., the point probability, is output to comprehensively reflect the superimposed risk of the original event and the disaster chain transmission.

[0011] A further improvement of the technical solution of the present invention is that in S6, the process of obtaining the overall risk value of the disaster chain by superposition is: Based on the total probability calculation results of each hazard-prone body and the losses of each hazard-prone body, the individual risk value is calculated. The individual risk value is the product of the total probability of each hazard-prone body and its losses. The losses are related to the hazard-prone body's own properties and are calculated using the vulnerability of the hazard-prone body. The individual risk values ​​of all hazard-prone objects in the disaster chain are integrated and superimposed to output the overall risk value of the disaster chain, which represents the comprehensive destructive potential of the event or process.

[0012] A further improvement of the technical solution of the present invention is that the calculation formula of the individual risk value is: ; Where, is the individual risk value of the hazard-prone body z, representing the absolute risk level it faces. is the final risk probability of the disaster-bearing body z, is the loss of the hazard-bearing body z, which is determined by its vulnerability. The value of approaches 0, and the disaster-bearing body has no risk. The larger the value of , the higher the risk; The calculation formula for the overall risk value of the disaster chain is: ; Where, It is the overall risk value of the disaster chain, reflecting the comprehensive destructive potential of the entire event process, and Q is the number of all disaster-prone objects in the disaster chain.

[0013] A further improvement of the technical solution of the present invention is that in S7, the process of calculating and determining the overall regional risk is: Based on the geographic information system of the study area, the individual risk value of each hazard-prone body is mapped to its spatial location to generate a risk value distribution raster layer. Spatial interpolation is used to ensure that the risk value covers all geographical units, forming a continuous risk intensity surface that reflects the local risk level at different locations in the region. Using spatial statistical methods, the overall risk value of the disaster chain is aggregated according to the geographical boundaries of the study area. The overall risk value of the disaster chain is compared with the preset risk threshold to divide the risk level. A regional risk classification map is generated to intuitively display the spatial heterogeneity of the risk. Among them, the risk level is divided into high risk level, medium risk level and low risk level. Combined with the regional risk classification map, mark the risk levels in the study area and highlight areas with high risk levels to determine the risk prevention and control priority list.

[0014] Due to the adoption of the above technical solution, the present invention has the following technical advancements compared to the prior art: The present invention provides a multi-hazard chain marine disaster risk assessment method based on the distribution of regional hazard-prone bodies. By integrating disaster chain theory, probability statistical models and GIS spatial analysis, it can achieve systematic identification and quantification of multi-hazard chain marine disaster risks, clarify the logical relationship between primary events, secondary events and derivative events, construct a disaster chain structure diagram, calculate the total probability of hazard-prone bodies through a probabilistic iterative algorithm, comprehensively reflect direct and indirect transmission risks, calculate individual risk values ​​based on the vulnerability of hazard-prone bodies, and superimpose them to generate an overall risk value for the disaster chain. This avoids the limitations of traditional methods that analyze single disasters in isolation, and can more comprehensively capture the complex interactions and cumulative effects of disaster chains.

[0015] The present invention provides a multi-hazard chain marine disaster risk assessment method based on the distribution of regional hazard-prone bodies. It uses GIS technology to map the risk value of the hazard-prone body to the actual geographical location, and generates a continuous risk intensity surface through spatial interpolation and rasterization processing, ensuring the spatial accuracy of the risk assessment results. It can intuitively display the risk distribution characteristics of different locations in the region. At the same time, it aggregates risk values ​​and divides risk levels through spatial statistical methods to generate a regional risk classification map, providing an accurate geographical reference for regional disaster prevention planning, helping to identify high-risk hotspots and optimize resource allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0017] Figure 1 Schematic diagram of the workflow of the multi-hazard chain marine disaster risk assessment method based on the regional hazard-bearing body distribution of the present invention; Figure 2 The figure is a flow chart of the method for multi-hazard chain marine disaster risk assessment based on regional disaster-bearing body distribution of the present invention. DETAILED DESCRIPTION

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0019] Example 1, as Figure 1 、 Figure 2 As shown, the present invention provides a multi-hazard chain marine disaster risk assessment method based on regional hazard-bearing body distribution, comprising the following steps: S1. Define the scope of the study area, extract all hazard-prone objects within the study area, analyze their interaction relationships, and establish an association probability matrix. Use maps to define the geographic boundaries of the study area. Use GIS to extract all hazard-prone objects in the study area and classify and record their attributes to form an inventory of hazard-prone objects. Based on the hazard chain theory, identify the physical, functional, or spatial associations between hazard-prone objects. Determine the interaction logic between hazard-prone objects through historical data. Construct a directed network diagram to visualize the relationship path. Use the analytic hierarchy process to calculate the association probability and ensure proportional consistency. Analyze the interaction strength between hazard-prone objects and establish an association probability matrix. The specific work content includes: using geographic information system (GIS) technology to load a map of the study area, defining geographic boundaries through polygon drawing, generating a vector layer of the study area, determining the assessment scope through spatial data, and using the spatial analysis function of GIS to extract all hazard-prone objects within the study area, including infrastructure, residential areas, ecological protection areas, etc., and classify and record their attributes. The hazard-prone object attribute information is integrated to form a structured hazard-prone object list. The hazard-prone object attribute information includes ID, hazard-prone object name, coordinates, scale, and functional category. The spatial data sources include OpenStreetMap, public databases, and field survey data. Based on the disaster chain theory, the interaction mechanisms between the various hazard-prone objects in the study area were analyzed, including physical dependence, functional association, or spatial proximity. Historical disaster event data was used to analyze the damage sequence and association paths of the hazard-prone objects, determine the logical relationship between the interactions between the hazard-prone objects, and use graphical tools to construct a directed network diagram to visualize the disaster chain path. The hazard-prone objects were used as network nodes, and the interactions between the hazard-prone objects as directed edges, intuitively displaying the relationship paths between the hazard-prone objects. The analytic hierarchy process (AHP) was used to calculate the association probability between each pair of hazard-prone objects, quantify the interaction strength between the hazard-prone objects, and then organize and arrange the association probabilities to form an association probability matrix to reflect the relative interaction strength between the hazard-prone objects. S2. Identify the chain effects between disaster events, clarify the primary events, secondary events, derivative events, and the disaster chain structure diagram. Based on historical disaster data and disaster-causing mechanisms, analyze the disaster event chain structure, identify the primary events and the disaster-prone bodies they directly affect, and analyze the secondary events and derivative events triggered by the changes in the state of the disaster-prone bodies. Then, construct a disaster chain logical relationship diagram through the event tree, clarify the causal transmission path and hierarchical structure between each event, abstract the primary events, secondary events, derivative events, and their associated disaster-prone bodies into directed network nodes, convert the chain effects between events into directed edges, and use graph theory methods to calculate the betweenness centrality to identify key hub events. At the same time, identify disaster transmission paths including high probability or high loss as vulnerable paths to form a disaster chain structure diagram; The specific work content is as follows: Based on historical disaster data and disaster-causing mechanisms, analyze the chain structure of disaster events, clarify the primary events and the disaster-affected bodies they directly affect, identify the triggered secondary events and derivative events by analyzing the state changes (damage, functional failure, etc.) of the disaster-affected bodies under the influence of disasters, and use the event tree analysis method to decompose the disaster evolution process layer by layer, construct a disaster chain logical relationship diagram, and present the causal transmission path and hierarchical structure. The first layer is the primary event, the second layer is the secondary event, and the third layer is the derivative event. Among them, the branch nodes of the event tree represent different stages of disaster development, and the edges represent the triggering probability or conditions between events, systematically describing the complete life cycle of the disaster chain. The primary event is the initial disaster event directly triggered by the hazard factor and is the starting point of the disaster chain. It is directly triggered by the external hazard factor acting on the disaster-affected body and does not depend on the occurrence of other disaster events. The secondary event is a disaster event triggered by the primary event and indirectly caused by the change in the state of the hazard-affected body. The primary event is required as a triggering condition, and the hazard-affected body is transformed into a new hazard factor in this process. Derivative events are social, economic, or environmental systemic impacts further triggered by secondary events, and are more indirect. Compared with the first two types of events, derivative events have longer causal chains and wider impact ranges. Events and associated hazard-bearing bodies in the disaster chain are abstracted as nodes of a directed network, with primary events, secondary events, and derivative events as the starting, intermediate, and terminal nodes, respectively. The chain effects between events are converted into directed edges, and the edge weights are quantified through historical statistics. Graph theory methods are used to calculate betweenness centrality and identify key hub events (nodes that frequently trigger secondary disasters). By identifying disaster transmission paths with high probability or high loss, i.e., vulnerable paths, a disaster chain structure diagram is formed, revealing the weak links and core transmission mechanisms of the disaster chain. The calculation formula for betweenness centrality is: ; Where, is the betweenness centrality of node v, which measures its hubness in the network, is the total number of shortest paths from node s to node t, is through node v The number of shortest paths, s and t are any two nodes in the network except v, The higher the value of is, the stronger the hub of node v in the disaster chain is. The lower the value of , the node v is an edge event and has a smaller impact on disaster transmission; The calculation formula for high-probability or high-loss disaster transmission paths is: ; ; Where, is the joint probability of the path, used for high-probability path analysis; is the cumulative loss of the path, used for high-loss path analysis; is the edge probability from node i to i+1 (event triggering probability), is the edge loss weight from node i to i+1 (vulnerability of the disaster-bearing body), n is the number of nodes in the path, The closer it is to 1, the higher the possibility of conduction. The larger the value of , the more serious the potential loss; S3. Analyze the probability of direct effects of hazard factors on each hazard-prone body in marine disasters, that is, the probability of occurrence of primary events. Analyze the intensity, frequency, and spatial distribution characteristics of hazard factors based on historical disaster data. Use probability statistics based on extreme value analysis to calculate the probability of occurrence of hazard factors of different intensities, forming a hazard level map. Combined with GIS spatial overlay, determine the probability distribution of hazard factors within the study area. Construct a sensitivity index system based on the type and physical properties of the hazard-prone body. Analyze the response probability of the hazard-prone body using vulnerability curves to quantify the degree of response of the hazard-prone body to hazard factors of different intensities. Calculate the direct effect probability based on the product relationship between hazard (probability of occurrence of hazard factors) and sensitivity (probability of response of the hazard-prone body). Use a spatially explicit method based on raster operations to generate a primary event probability matrix pixel by pixel. The specific work content is: Based on historical disaster data, sort out the intensity, frequency and spatial distribution characteristics of disaster-causing factors, calculate the occurrence probability of disaster-causing factors of different intensities by statistically analyzing the extreme value parameters of different disaster types and combining probability statistics methods, and classify the hazard level (high, medium and low) based on historical disaster frequency data to form a hazard level map covering the study area. Through GIS spatial overlay analysis, integrate multiple disaster-causing factor layers to determine the comprehensive occurrence probability of disaster-causing factors in different regions; Based on the types of hazard-prone objects (such as buildings, infrastructure, population, and ecosystems) and their physical properties, a multidimensional sensitivity index system was constructed, covering structural resilience, functional redundancy, and exposure. The weights of each sensitivity index were determined using the analytic hierarchy process (AHP). Based on the vulnerability curve theory, a nonlinear relationship model was established between the response probability of the hazard-prone object and the intensity of the hazard factor. The response probability of the hazard-prone object under the influence of hazard factors of different intensities, namely the probability of damage or functional failure, was quantified, forming a sensitivity database covering all types of hazard-prone objects. Based on the product relationship between hazard (probability of occurrence of hazard factors) and sensitivity (probability of response of hazard-prone objects), the study area was divided into regular pixel units using raster operation technology. The direct effect probability of different hazard factor and hazard-prone object combinations was calculated pixel by pixel. Specifically, the hazard factor hazard layer and the hazard-prone object sensitivity layer were superimposed on each pixel, and a probability product operation was performed using GIS spatial analysis tools to generate a primary event probability matrix. The calculation formula for the occurrence probability of disaster factors of different intensities is: ; Where, is the probability of occurrence of disaster factors of different intensities, that is, the probability of occurrence of disaster factors of given intensity The probability of occurrence of the following; Is the intensity in historical records ≥ The number of disaster events, is the total number of observation years, is the intensity threshold of the hazard factor, A value of 0 means it never happens, and a value close to 1 means it happens every year. The higher, The smaller it is, the higher the frequency is. The bigger; The calculation formula for the response probability of the disaster-bearing body is: ; Where, is the response probability of the disaster-bearing body, that is, the probability of damage or failure of the disaster-bearing body under the intensity of the disaster factor I; I is the actual intensity of the disaster factor, is the damage intensity threshold of the disaster-bearing body, k, is the shape parameter of the fragility curve, which is fitted by historical data. A value of 0 indicates no damage at all, and a value close to 1 indicates certain damage. , tends to 0, , Rapidly approaches 1 (non-linear growth); The calculation formula for direct effect probability is: ; Where, It is the comprehensive probability of the disaster-causing factor directly acting on the disaster-bearing body, that is, the probability of the primary event occurring. A value of 0 indicates no risk, a value close to 1 indicates that damage is inevitable, and a high +High , Significantly increased, low or low , Approaching 0; S4. Based on the interaction between hazard-bearing bodies, calculate the marginal probability and association probability of secondary or derivative events induced between hazard-bearing bodies. Collect data on the attributes, spatial locations, and historical disaster events of hazard-bearing bodies to clarify the interaction relationship between hazard-bearing bodies. Based on the interaction mechanism of hazard-bearing bodies and combined with the collected data, determine the factors affecting the marginal probability, including the intensity of the hazard-causing factor and the sensitivity of the hazard-bearing body. Use a probabilistic statistical model to quantify the impact of each factor on the marginal probability, and calculate the marginal probability of secondary or derivative events induced between hazard-bearing bodies. Based on the results of the marginal probability calculation, analyze the association between hazard-bearing bodies and use probabilistic reasoning methods to analyze and derive the association probability of secondary or derivative events induced between hazard-bearing bodies. The specific work content includes: collecting attribute data of hazard-prone objects, including functional categories and physical characteristics, as well as spatial locations and historical disaster event records, to form a structured database. Using spatial analysis techniques, the distribution characteristics of hazard-prone objects and their potential role in the disaster chain are clarified. Based on the disaster chain theory, the interaction mechanism between hazard-prone objects is analyzed, including physical dependence, functional association and spatial proximity. A hazard-prone object relationship network is constructed. Using historical disaster event data, the relationship between changes in the state of hazard-prone objects and the triggering of secondary or derivative events is identified. Based on the collected data and the interaction mechanism between hazard-affected bodies, the factors affecting the marginal probability are determined, including the intensity of the hazard-causing factor and the sensitivity of the hazard-affected body. A probability statistical model based on regression analysis is used to quantify the contribution weight of each factor to the marginal probability. A marginal probability calculation framework is established, and the model parameters are fitted with historical disaster event data to obtain the marginal probability of secondary or derivative events induced between hazard-affected bodies, reflecting the potential possibility of secondary or derivative events induced between hazard-affected bodies. Based on the marginal probability calculation results, the correlation between hazard-affected bodies is further analyzed, and the probabilistic reasoning method (Markov chain) is used to derive the association probability of secondary or derivative events induced between hazard-affected bodies. The association probability represents the relative strength of the interaction between hazard-affected bodies and reflects the stability and directionality of event transmission in the disaster chain. The calculation formula for the marginal probability of secondary or derivative events between hazard-bearing bodies is: ; Where, is the probability that the hazard-bearing body c induces secondary / derivative events in the hazard-bearing body z, Is the Sigmoid function, which maps the linear combination to a probability value, is the intensity of the hazard factor on the hazard-bearing body c, standardized to [0, 1], is the sensitivity of the hazard-bearing body z, standardized to [0, 1], is the regression coefficient, which is fitted by historical data. is a random error term that follows a normal distribution. A value closer to 0 indicates that it is extremely difficult to trigger, and a value closer to 1 indicates that it is certain to trigger. The calculation formula for the association probability of secondary or derivative events induced by disaster-bearing bodies is: ; ; ; Where, is the long-term correlation probability between the disaster-bearing bodies c and z, reflecting the steady-state transmission intensity of the disaster chain. is the marginal probability of the f-th disaster event, are the transition weights of the Markov chain, is the actual number of observations of the hazard-bearing body c triggering the hazard-bearing body z in the f-th event, is the total number of times that the disaster-prone body c triggers any other disaster-prone body in the f-th disaster event (i.e., the sum of the observation frequencies of all outgoing edges), m is the index of all disaster-prone body nodes directly connected to c in the network, is the time decay weight, is the forgetting factor, T is the total number of observed events, The value of 0 means no association. A value of 1 indicates completely stable conduction; S5. Use the probability iteration algorithm to calculate the point probability and edge probability based on the probability of the original event, and calculate the total probability of each hazard-bearing body layer by layer; S6. Calculate the risk of the hazard-prone body by combining its total probability and loss, and obtain the overall risk value of the disaster chain by superposition. The loss is a quantity related to the hazard-prone body's own properties and is calculated using the vulnerability of the hazard-prone body. S7. Calculate and determine the overall regional risk based on the distribution of hazard-prone objects within the study area.

[0020] Example 2, as Figure 1 、 Figure 2 As shown, based on Example 1, the present invention provides a technical solution: Preferably, in S5, the process of calculating the total probability of each disaster-prone body layer by layer is: Based on historical disaster data and disaster chain theory, the initial occurrence probability of the primary event is determined, and the edge probability of the primary event on the directly related disaster-prone body is quantified through regression analysis. The initial probability transfer matrix is ​​constructed to clarify the direct path of disaster transmission and its basic occurrence probability. A forward iterative algorithm is used, starting from the primary event, to calculate the point probability of the secondary disaster-prone body layer by layer. The point probability of each node in each layer is equal to the sum of the products of the point probabilities of all its predecessor nodes and the corresponding edge probabilities. Efficient propagation is achieved through matrix operations until all disaster-prone bodies are covered, forming a complete disaster transmission probability field. The iterative results are checked for steady-state. The calculation is terminated when the change in the point probability of two consecutive iterations is less than the preset probability threshold to ensure the convergence of the probability distribution. The total probability of each disaster-prone body, i.e., the point probability, is output, which comprehensively reflects the superimposed risk of the primary event and disaster chain transmission. The specific work content is: analyze historical disaster data and disaster chain theory, determine the initial occurrence probability of primary events by statistically analyzing the frequency of historical events, and at the same time, analyze the interaction mechanism between primary events and directly related disaster-prone bodies, identify the triggering conditions of physical dependence, functional association or spatial proximity, and construct a regression model to quantify the edge probability. The model input includes the intensity of the disaster-causing factor and the sensitivity of the disaster-prone body. The regression coefficient is fitted by historical disaster data to determine the direct impact weight of the primary event on each disaster-prone body, form an initial probability transfer matrix, and clarify the direct path of disaster transmission and its basic occurrence possibility; adopt a forward iterative algorithm, take the primary event as the starting point, and calculate the point probability of the secondary disaster-prone body layer by layer. The point probability of each node is obtained by the weighted sum of the product of the point probability of all its predecessor nodes and the corresponding edge probability. It is propagated through matrix operations until all disaster-prone bodies are covered, dynamically simulate the transmission path of the disaster chain, and form a complete disaster transmission probability field; The iterative results are checked for steady-state conditions by comparing the point probability changes between two consecutive iterations. The calculation is terminated when the change is less than the preset probability threshold to ensure convergence of the probability distribution. This verifies the stability of the disaster transmission process and avoids probability oscillations caused by cyclic dependence or path redundancy. The final output is the total probability of each hazard-bearing body, i.e., the point probability, which comprehensively reflects the superimposed risk of the direct effect of the primary event and the indirect transmission of the disaster chain. The calculation formula for the total probability of each disaster-prone body is: ; Where, is the final risk probability of the disaster-prone body z, which takes into account the impact of all direct and indirect transmission paths. 0 means absolute safety, and close to 1 means inevitable damage. is the set of all disaster-bearing entities or primary events in the network that can reach z through directed edges. is the point probability of the precursor disaster-bearing body c. If c is a primary event, then is the initial probability of occurrence; is the edge probability of the disaster-bearing body c to z, Approaching 1 means there is at least one high-probability path. Approaching 0, indicating that all or Very low, the disaster-bearing body z is at the end of the disaster chain and the path is long; In S6, the process of obtaining the overall risk value of the disaster chain through superposition is: Based on the total probability calculation results of each hazard-prone body and the losses of each hazard-prone body, the individual risk value is calculated. The individual risk value is the product of the total probability of each hazard-prone body and its losses. The losses are a quantity related to the hazard-prone body's own properties. Using the vulnerability calculation of the hazard-prone body, the individual risk values ​​of all hazard-prone bodies in the disaster chain are integrated and superimposed to output the overall risk value of the disaster chain, which represents the comprehensive destructive potential of the event or process. The specific work content is as follows: Based on the disaster risk assessment system, with the total probability of each hazard-prone body as the core, combined with the loss situation of the hazard-prone body itself, the two are multiplied to obtain the individual risk value of each hazard-prone body. Among them, loss is closely related to the inherent attribute of the hazard-prone body and is determined by analyzing its vulnerability. Vulnerability reflects the vulnerability of the hazard-prone body when facing disasters. Different attributes lead to different disaster resistance capabilities. The loss results are obtained through the vulnerability curve to quantify the degree of damage to the hazard-prone body under the action of the disaster; after obtaining the individual risk value of each hazard-prone body, the individual risk values ​​of all hazard-prone bodies in the disaster chain are integrated. The individual risk values ​​of each hazard-prone body are integrated in an overlay manner, and the risks faced by each hazard-prone body are comprehensively considered to output the overall risk value of the disaster chain, which represents the comprehensive destructive potential of the entire disaster chain event or process; The formula for calculating individual risk value is: ; Where, is the individual risk value of the hazard-prone body z, representing the absolute risk level it faces. is the final risk probability of the disaster-bearing body z, is the loss of the hazard-bearing body z, which is determined by its vulnerability. The value of approaches 0, and the disaster-bearing body has no risk. The larger the value of , the higher the risk; The calculation formula for the overall risk value of the disaster chain is: ; Where, is the overall risk value of the disaster chain, reflecting the comprehensive destructive potential of the entire event process, Q is the number of all disaster-bearing bodies in the disaster chain, The larger the value of , the more destructive the disaster chain is; In S7, the process of calculating and determining the overall regional risk is as follows: Based on the geographic information system of the study area, the individual risk value of each hazard-prone body is mapped to its spatial location to generate a risk value distribution raster layer. Spatial interpolation is used to ensure that the risk value covers all geographical units, forming a continuous risk intensity surface that reflects the local risk levels at different locations within the region. Spatial statistical methods are used to aggregate the overall risk value of the disaster chain according to the geographical boundaries of the study area. The overall risk value of the disaster chain is compared with the preset risk threshold to divide the risk level, and a regional risk classification map is generated to intuitively display the spatial heterogeneity of the risk. Among them, the risk level is divided into high risk level, medium risk level and low risk level. Combined with the regional risk classification map, each risk level in the study area is marked, and the areas with high risk level are highlighted to determine the risk prevention and control priority list; Multiple risk levels correspond to multiple risk thresholds one by one, and the corresponding relationship is as follows: The risk thresholds for the high risk level are: ; The risk thresholds for the medium risk level are: ; The risk thresholds for the low risk level are: ; in, is the overall risk value of the disaster chain, is the lower threshold of high risk level and the upper threshold of medium risk level, The lower threshold for medium risk level and the upper threshold for low risk level; The specific work content is as follows: Based on the geographic information system of the study area, the individual risk value of each hazard-prone body is associated with its spatial coordinates. The discrete individual risk values ​​are mapped to their actual geographic locations through spatial data matching technology. The risk values ​​of geographical units that are not directly covered are estimated using spatial interpolation methods to ensure that all spatial locations have risk value attributes. Through rasterization processing, the continuous geographic space is divided into regular grid cells. Each cell is assigned an interpolated risk value to generate a risk value distribution raster layer. Using spatial statistical methods and taking the geographical boundaries of the study area as the aggregation unit, the individual risk values ​​of all disaster-prone bodies in the disaster chain are summarized and calculated to obtain the overall risk value of the disaster chain. By comparing with the preset risk thresholds, the overall risk value is divided into three risk levels: high, medium and low. The thematic mapping function of GIS is used to assign differentiated colors or symbols to different areas according to the risk level to generate a regional risk classification map; combined with the regional risk classification map, each risk level in the study area is spatially marked, with high-risk areas highlighted, and their geographical location and scope are clarified through the annotation function. Based on the spatial distribution characteristics of risk levels, a risk prevention and control priority list is constructed, and high-risk areas are prioritized in the focus of disaster prevention, emergency preparedness and resilience improvement. At the same time, combined with the potential threats of medium and low-risk areas, differentiated prevention and control strategies are formulated to form a risk management model with graded response and precise policy implementation.

[0021] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A multi-hazard chain marine disaster risk assessment method based on regional hazard-bearing body distribution is characterized by: The following steps are involved: S1. Define the scope of the study area, extract all hazard-prone bodies within the study area, analyze their interactions, and establish an association probability matrix; S2. Identify the chain effects between disaster events, clarify the primary events, secondary events, derivative events and the disaster chain structure diagram; S3. Analyze the probability of direct effects of marine disaster-causing factors on each hazard-bearing body, that is, the probability of occurrence of primary events; S4. Based on the interaction between hazard-bearing bodies, calculate the marginal probability and correlation probability of secondary or derivative events induced between hazard-bearing bodies; S5. Use the probability iteration algorithm to calculate the point probability and edge probability based on the probability of the original event, and calculate the total probability of each hazard-bearing body layer by layer; S6. Combine the total probability of the hazard-bearing body with its losses to calculate the risk of the hazard-bearing body, and obtain the overall risk value of the disaster chain through superposition; S7. Calculate and determine the overall regional risk based on the distribution of hazard-prone objects within the study area.

2. The multi-hazard chain marine disaster risk assessment method based on regional hazard-bearing body distribution according to claim 1 is characterized by: In S1, the process of establishing the association probability matrix is: Use maps to clearly define the geographical boundaries of the study area, use GIS to extract all hazard-prone objects in the study area, classify and record their attributes, and form a hazard-prone object list; Based on the disaster chain theory, the physical, functional or spatial connections between hazard-prone objects are identified, the interaction logic between hazard-prone objects is determined through historical data, and a directed network diagram is constructed to visualize the relationship path; The hierarchical analysis method is used to calculate the association probability, analyze the interaction intensity between disaster-sustaining bodies, and establish the association probability matrix.

3. The multi-hazard chain marine disaster risk assessment method based on regional hazard-bearing body distribution according to claim 1 is characterized by: The S2 specifically includes: Based on historical disaster data and disaster-causing mechanisms, we analyze the chain structure of disaster events, identify primary events and the disaster-affected bodies they directly affect, and analyze the secondary and derivative events triggered by changes in the state of the disaster-affected bodies. We then construct a disaster chain logic diagram through an event tree to clarify the causal transmission paths and hierarchical structure between events. Primary events, secondary events, derivative events and their associated disaster-prone bodies are abstracted as directed network nodes, and the chain effects between events are converted into directed edges. Graph theory methods are used to calculate the betweenness centrality and identify key hub events. At the same time, disaster transmission paths with high probability or high loss are identified as vulnerable paths to form a disaster chain structure diagram.

4. The multi-hazard chain marine disaster risk assessment method based on regional hazard-bearing body distribution according to claim 1 is characterized by: The S3 specifically includes: Based on historical disaster data, the intensity, frequency, and spatial distribution characteristics of disaster-causing factors are analyzed. The probability of occurrence of disaster-causing factors of different intensities is calculated using a probability statistical method based on extreme value analysis. This forms a hazard level map. Combined with GIS spatial overlay, the probability distribution of disaster-causing factors in the study area is determined. Based on the type of hazard-prone objects and their physical properties, a sensitivity index system is constructed. The response probability of hazard-prone objects is analyzed through vulnerability curves, and the degree of response of hazard-prone objects to hazard factors of different intensities is quantified. Based on the product relationship between hazard and sensitivity, the direct effect probability is calculated, and a spatial explicit method based on raster operations is used to generate the probability matrix of primary event occurrence pixel by pixel.

5. The multi-hazard chain marine disaster risk assessment method based on regional hazard-bearing body distribution according to claim 4 is characterized by: The S4 specifically includes: Collect data on the attributes, spatial locations, and historical disaster events of hazard-prone objects, clarify the interaction relationships between hazard-prone objects, and based on the interaction mechanism of hazard-prone objects and combined with the collected data, determine the factors affecting the marginal probability, including the intensity of the hazard-causing factor and the sensitivity of the hazard-prone object. Use a probabilistic statistical model to quantify the impact of each factor on the marginal probability, and calculate the marginal probability of secondary or derivative events induced by hazard-prone objects. Based on the marginal probability calculation results, the correlation between disaster-prone bodies is analyzed, and the probability of secondary or derivative events induced by disaster-prone bodies is analyzed using probabilistic reasoning methods.

6. The multi-hazard chain marine disaster risk assessment method based on regional hazard-bearing body distribution according to claim 1 is characterized by: In S5, the process of calculating the total probability of each disaster-prone body layer by layer is as follows: Based on historical disaster data and disaster chain theory, the initial occurrence probability of the primary event is determined, and the marginal probability of the primary event on the directly related hazard-bearing body is quantified through regression analysis to construct the initial probability transfer matrix; Using a forward iterative algorithm, starting from the primary event, the point probability of the secondary disaster-prone body is calculated layer by layer. The point probability of each layer node is equal to the sum of the product of the point probability of all its predecessor nodes and the corresponding edge probability, until all disaster-prone bodies are covered, forming a complete disaster transmission probability field. The iterative results are checked for steady-state conditions. When the point probability change between two consecutive iterations is less than the preset probability threshold, the calculation is terminated and the total probability of each disaster-prone body, i.e., the point probability, is output to comprehensively reflect the superimposed risk of the original event and the disaster chain transmission.

7. The multi-hazard chain marine disaster risk assessment method based on regional hazard-bearing body distribution according to claim 6 is characterized by: In S6, the process of obtaining the overall risk value of the disaster chain by superposition is: Based on the total probability calculation results of each hazard-prone body and the losses of each hazard-prone body, the individual risk value is calculated. The individual risk value is the product of the total probability of each hazard-prone body and its losses. The losses are related to the hazard-prone body's own properties and are calculated using the vulnerability of the hazard-prone body. The individual risk values ​​of all hazard-prone objects in the disaster chain are integrated and superimposed to output the overall risk value of the disaster chain, which represents the comprehensive destructive potential of the event or process.

8. The multi-hazard chain marine disaster risk assessment method based on regional hazard-bearing body distribution according to claim 7 is characterized by: The calculation formula of the individual risk value is: ； Where, is the individual risk value of the hazard-bearing body z, is the final risk probability of the disaster-bearing body z, is the loss of the hazard-bearing object z, determined by its vulnerability; The calculation formula for the overall risk value of the disaster chain is: ； Where, is the overall risk value of the disaster chain, and Q is the number of all disaster-prone objects in the disaster chain.

9. The multi-hazard chain marine disaster risk assessment method based on regional hazard-bearing body distribution according to claim 8 is characterized by: In S7, the process of calculating and determining the overall regional risk is as follows: Based on the geographic information system of the study area, the individual risk value of each hazard-prone body is mapped to its spatial location, generating a risk value distribution raster layer that covers all geographic units and forms a continuous risk intensity surface, reflecting the local risk level at different locations within the region. Using spatial statistical methods, the overall risk value of the disaster chain is aggregated according to the geographical boundaries of the study area. The overall risk value of the disaster chain is compared with the preset risk threshold to divide the risk level. A regional risk classification map is generated to intuitively display the spatial heterogeneity of the risk. Among them, the risk level is divided into high risk level, medium risk level and low risk level. Combined with the regional risk classification map, mark the risk levels in the study area and highlight areas with high risk levels to determine the risk prevention and control priority list.

Images