Landslide geological disaster risk assessment method and equipment under combined action of earthquake and rainfall
By constructing a dual-branch risk assessment model, combining multi-source time-series data and physical mechanisms, dynamic assessment of landslide geological hazard risk under the combined effects of earthquakes and rainfall was achieved. This solves the problems of static and isolated analysis in existing technologies, improves the accuracy and time-series dynamism of risk assessment, and provides a scientific basis for disaster prevention and mitigation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NAT INST OF NATURAL HAZARDS MINISTRY OF EMERGENCY MANAGEMENT OF CHINA
- Filing Date
- 2026-01-16
- Publication Date
- 2026-04-28
AI Technical Summary
Existing technologies for assessing the risk of landslide geological hazards under the combined effects of earthquakes and rainfall suffer from static and isolated analysis, insufficient model assimilation capabilities, and a lack of temporal dynamism. They cannot effectively quantify and integrate the dynamic interactions of multi-source time-series observation data and embed physical mechanisms, resulting in low risk assessment accuracy and an inability to characterize the temporal evolution of risks.
Multi-source time-series data are used to quantify earthquake damage and dynamic hydrological response factors. A dual-branch risk assessment model is constructed, which combines physical mechanisms and data-driven models. A coupled inference module is used to realize time-series dynamic risk assessment and output a time-series dynamic geological disaster risk probability map.
It improves the accuracy and interpretability of risk assessment, reveals the evolution of risk over time, provides a scientific basis for medium- and long-term disaster prevention and mitigation after earthquakes, and enhances the robustness of the model in extreme and unknown scenarios.
Smart Images
Figure CN121938152A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of slope geological hazard risk assessment technology, and more specifically to a method and equipment for assessing landslide geological hazard risk under the combined effects of earthquakes and rainfall. Background Technology
[0002] Earthquakes and rainfall are the two main factors that trigger geological disasters (such as landslides, collapses, and debris flows). Current risk assessment methods have the following limitations: 1. Static and Isolated Analysis. Traditional methods often involve simple algebraic superposition of earthquake-induced risk maps and rainfall-induced risk maps, such as weighted superposition or the maximum value method. This approach ignores the physical mechanisms of the dynamic coupling between earthquakes and rainfall. For example, earthquakes cause damage to soil and rock masses and generate numerous cracks, which significantly alter the hydrogeological conditions of a watershed. This fundamentally changes the infiltration, transport, and pore water pressure response patterns of subsequent rainfall. This preconditioning effect has not been effectively quantified in existing methods.
[0003] 2. Insufficient model assimilation capability. Existing physical models, such as the Newmark slider model and the infinite slope stability model, have clear mechanisms, but their parameters are difficult to obtain and they are difficult to assimilate massive amounts of time-series remote sensing observation data. While pure data-driven machine learning models (such as random forests and CNNs) are good at learning patterns from data, their interpretability is poor and their reliability is questionable when extrapolating to coupled scenarios that have not been experienced before (such as strong earthquakes + extreme rainfall).
[0004] 3. Lack of temporal dynamism. Risk evolves dynamically over time. After an earthquake, the soil and rock mass undergoes a time-dependent process of "damage-healing," and its sensitivity to rainfall changes accordingly. Most current methods provide a static "snapshot" of risk, failing to depict the temporal evolution of risk over months to years after the earthquake.
[0005] Therefore, how to construct a productized assessment framework that can deeply integrate multi-source time-series observation data, embed physical mechanisms, and output time-series dynamic risks is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention
[0006] In view of this, the present invention provides a method and equipment for risk assessment of landslide geological hazards under the combined effects of earthquakes and rainfall, which can quantify the dynamic interaction between earthquake damage and rainfall and achieve time-series dynamic risk assessment. This invention not only improves the accuracy of risk assessment but also reveals the evolutionary pattern of risk over time, providing a scientific basis for medium- and long-term disaster prevention and mitigation after earthquakes.
[0007] To achieve the above objectives, the present invention adopts the following technical solution: A method for assessing the risk of landslide geological hazards under the combined effects of earthquakes and rainfall includes the following steps: S1. Acquire and preprocess multi-source time-series data for the evaluation area; S2. Based on the multi-source time-series data, quantify the earthquake damage factor and dynamic hydrological response factor, and combine the two to generate a coupling triggering factor; S3. Construct a dual-branch risk assessment model, and use the multi-source time series data and the coupling triggering factor to train and calibrate the dual-branch risk assessment model. S4. Using the trained dual-branch risk assessment model and combining dynamic input data from different time periods, generate and output a time-series dynamic geological disaster risk probability map of the assessment area.
[0008] Furthermore, the multi-source time-series data includes: earthquake-related data, rainfall time-series data, geological environment baseline data, disaster logging data, and surface change monitoring data.
[0009] Furthermore, the earthquake damage factor is calculated by integrating the peak ground acceleration field and the post-earthquake creep rate obtained based on time-series InSAR data, and is used to quantify the degree of earthquake damage to the soil and rock mass. The dynamic hydrological response factor uses the earthquake damage factor as a correction parameter to correct the topographic humidity index, generating an earthquake-corrected infiltration potential index, which is then calculated in conjunction with effective rainfall or soil moisture.
[0010] Furthermore, the formula for calculating the earthquake-corrected infiltration potential index is as follows:
[0011] Where α is the infiltration enhancement coefficient, Earthquake damage factor, This indicates the terrain humidity index.
[0012] Furthermore, the dual-branch risk assessment model includes a physical mechanism branch and a data-driven branch; The physical mechanism branch dynamically corrects the pore water pressure ratio of the infinite slope model based on the coupling triggering factor to calculate the dynamic safety factor; the data-driven branch uses multi-source time-series data and the coupling triggering factor as input to generate a disaster occurrence probability map based on a neural network model.
[0013] Furthermore, in the physical mechanism branch, the expression for the dynamic pore water pressure ratio is:
[0014] In the formula, The background pore water pressure ratio of the region; The pore water pressure response coefficient controls the degree of amplification of pore water pressure. This is the coupling triggering factor.
[0015] Furthermore, the expression for calculating the dynamic safety factor is:
[0016] In the formula, The effective cohesion of the soil and rock mass; The effective internal friction angle of the rock and soil mass; This is the saturated unit weight of the soil and rock mass; d β represents the soil layer thickness; β represents the slope gradient. The dynamic pore water pressure ratio.
[0017] Furthermore, the dual-branch risk assessment model also includes a coupled reasoning module, which is used to fuse the dynamic safety coefficient map output by the physical mechanism branch and the disaster occurrence probability map output by the data-driven branch to generate a comprehensive geological disaster risk probability map and uncertainty assessment.
[0018] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a method for assessing the risk of landslide geological hazards under the combined effects of earthquakes and rainfall.
[0019] As can be seen from the above technical solution, compared with the prior art, the present invention provides a method and equipment for assessing the risk of landslide geological hazards under the combined effects of earthquake and rainfall, which has the following beneficial effects: 1. By introducing time-series InSAR data and coupling factors, the timeliness of earthquake damage and its impact on hydrological processes were quantified, achieving a qualitative change from static superposition to dynamic coupling.
[0020] 2. The dual-branch model combines the interpretability and extrapolation capability of the physical model with the high accuracy and strong fitting capability of the data-driven model. By coupling the inference module, it achieves complementary advantages and improves the robustness of the model in extreme and unknown scenarios.
[0021] 3. The output time-series risk map can clearly indicate which areas, at what time periods, and with what intensity of rainfall are at the highest risk after an earthquake, providing accurate decision support for post-disaster emergency response, temporary resettlement site planning, and long-term monitoring of major engineering facilities. Attached Figure Description
[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.
[0023] Figure 1 This is a schematic diagram of the method flow provided by the present invention. Detailed Implementation
[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0025] See Figure 1 This invention discloses a method for assessing the risk of landslide geological hazards under the combined effects of earthquakes and rainfall, comprising the following steps: S1. Acquire and preprocess multi-source time-series data for the evaluation area; S2. Based on multi-source time series data, quantify earthquake damage factors and dynamic hydrological response factors, and combine the two to generate coupling triggering factors; S3. Construct a dual-branch risk assessment model and use multi-source time series data and coupling triggering factors to train and calibrate the dual-branch risk assessment model. S4. Using the trained dual-branch risk assessment model and combining dynamic input data from different time periods, generate and output a time-series dynamic geological disaster risk probability map of the assessment area.
[0026] Specifically, the core of the invention lies in proposing a dual-branch coupled model of physical mechanism and data-driven approach. This model includes a physical mechanism branch and a data-driven branch, ultimately achieving fusion decision-making through a coupled inference module. The overall technical solution flow is shown below: Step 1: Preparation and preprocessing of multi-source time series data. Collect the following multimodal data of the study area and unify the spatiotemporal resolution and coordinate system.
[0027] Earthquake-related data: earthquake magnitude, focal mechanism, peak ground acceleration (PGA) field, coseismic deformation field (from InSAR), and aftershock sequence.
[0028] Rainfall time series data: long-term series rainfall data (GPM, TRMM satellite rainfall products), calculating cumulative rainfall and rainfall intensity at different time scales.
[0029] Geological environment baseline data: digital elevation model, lithology, geological structure, land use, and vegetation cover index.
[0030] Disaster cataloging data: a directory of historical geological disaster sites, especially newly occurring disaster sites after earthquakes.
[0031] Surface change monitoring data: time-series InSAR data, to obtain cumulative surface deformation and time-series deformation rate before, after, and during earthquakes.
[0032] The second step is to quantify the coupling factor between earthquake damage effects and rainfall infiltration potential, abandoning simple PGA or rainfall superposition, and constructing a new factor for dynamic coupling.
[0033] Earthquake damage factor ( Deq The post-earthquake creep rate (V_post) is a key parameter for quantifying the extent of earthquake damage to the Earth's surface. Its recommended value range is [0, 1]. The closer it is to 1, the more severe the earthquake damage and the more fully developed the fractures. In addition to considering the static PGA, this study introduces the post-earthquake creep rate (V_post) based on time-series InSAR as a proxy indicator for the long-term damage effect of the earthquake. Deq = f(PGA, V_post); A region with a high PGA value and a sustained high creep rate indicates that the soil and rock mass in that region is severely damaged and continuously unstable.
[0034] Dynamic hydrological response factor (R_hydro): Considering the enhanced infiltration capacity caused by post-earthquake fracture development. A simplified coupled hydrological model is constructed, which incorporates earthquake damage factors. Deq As a parameter, the traditional topographic moisture index (TWI) is modified to generate a "seismically corrected infiltration potential index" (ETWI). ETWI is obtained by correcting the TWI for seismic damage; its core lies in utilizing... Deq This "amplifies" the topography's water catchment and infiltration potential. The formula is as follows:
[0035] α is the infiltration enhancement coefficient, an adjustable parameter greater than 0 (e.g., α = 2.0). This parameter is crucial for model calibration, as it controls the maximum extent to which seismic damage affects infiltration capacity.
[0036] Earthquake damage factor ( D eq The parameter that quantifies the degree of earthquake damage to the Earth's surface is recommended to have a value range of [0,1]. The closer it is to 1, the more severe the earthquake damage. D eq In addition to considering static PGA, post-earthquake creep rate via temporal InSAR is also taken into account. A high PGA region superimposed with a sustained high creep rate indicates severe damage and continuous instability of the soil and rock mass in that region.
[0037] TWI stands for Terrain Humidity Index, a terrain analysis index based on digital elevation models. It reflects the ability of terrain to control hydrological processes and can be calculated using ArcGIS software based on DEM data.
[0038] R_hydro is determined by ETWI and real-time effective rainfall (or soil moisture).
[0039] Coupling trigger factor (C_tr): will D eq By combining with R_hydro through a nonlinear function, a coupling factor is generated that comprehensively represents the interaction between "fragile geological bodies (earthquake damage)" and "triggering conditions (rainfall)".
[0040] Step 3: Construct a two-branch risk assessment model: Branch A - Physical Mechanism Branch: Based on the infinite slope model, but with its key parameters dynamically adjusted. The coupling factor C_tr obtained in step two is used as a dynamic correction term for the pore water pressure ratio (r_u), and the dynamic safety factor of each grid cell is calculated. FS dynamic This branch ensures that the model has a clear physical mechanism.
[0041] Formula for calculating dynamic pore water pressure ratio:
[0042] Formula for calculating dynamic safety factor: The regional background pore water pressure ratio is a baseline value (e.g., 0.3-0.5), which can be obtained through investigation or calibration. : Pore water pressure response coefficient (a constant greater than 0), which controls the degree of amplification of pore water pressure; Effective cohesion of soil and rock mass; Effective internal friction angle of soil and rock mass; : Saturated unit weight of soil and rock mass; d β: Soil layer thickness; β: Slope gradient; For dynamic pore water pressure ratio, As a coupling triggering factor, it is to use D eq By combining R_hydro with a nonlinear function, a coupling factor is generated that comprehensively characterizes earthquake damage and rainfall-induced geological hazards.
[0043] Branch B - Data-Driven Branch: This branch uses a hybrid Transformer+UNet model as its core. The model input consists of multi-source time-series data prepared in Steps 1 and 2, including earthquake magnitude, focal mechanism, peak ground acceleration (PGA) field, coseismic deformation field, aftershock sequence, long-term rainfall data, cumulative rainfall, rainfall intensity, digital elevation model, lithology, geological structure, land use, vegetation cover index, time-series InSAR data, cumulative surface deformation, time-series deformation rate, and earthquake-corrected infiltration potential index. This generates a probability map.
[0044] Coupled Inference Module: It is not a simple weighted average, but a learnable "arbitrator".
[0045] Input: FS_dynamic graph output from branch A (physical mechanism), probability graph output from branch B (data-driven), and uncertainty estimate.
[0046] Mechanism: A lightweight feedforward neural network (Multilayer Perceptron, MLP) is employed. Its unique feature is that it does not directly predict the safety factor, but instead learns how to assign optimal fusion weights to the physical mechanism branch and the data-driven branch. When there are a large number of disaster samples, the results of the data-driven branch are trusted more; in extreme cases such as strongly coupled scenarios not present in the training data, the physical mechanism branch is relied upon for extrapolation; when the results of the two branches differ significantly, a high uncertainty flag is output, indicating the need for manual judgment.
[0047] Output: Final comprehensive geological hazard risk probability map, with an uncertainty assessment.
[0048] Step 4: Time-series dynamic risk assessment and verification: Dynamic assessment: By using a sliding time window, rainfall data and post-earthquake InSAR creep rate data from different time periods are input into a trained bi-branch model to obtain a series of risk probability maps on time slices, thereby enabling visualization and quantitative analysis of the spatiotemporal evolution of risk.
[0049] On the other hand, embodiments of the present invention also disclose an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a method for assessing the risk of landslide geological hazards under the combined effects of earthquakes and rainfall.
[0050] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.
[0051] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for assessing the risk of landslide geological hazards under the combined effects of earthquakes and rainfall, characterized in that, Includes the following steps: S1. Acquire and preprocess multi-source time-series data for the evaluation area; S2. Based on the multi-source time-series data, quantify the earthquake damage factor and dynamic hydrological response factor, and combine the two to generate a coupling triggering factor; S3. Construct a dual-branch risk assessment model, and use the multi-source time series data and the coupling triggering factor to train and calibrate the dual-branch risk assessment model. S4. Using the trained dual-branch risk assessment model and combining dynamic input data from different time periods, generate and output a time-series dynamic geological disaster risk probability map of the assessment area.
2. The method for assessing the risk of landslide geological hazards under the combined effects of earthquake and rainfall as described in claim 1, characterized in that, The multi-source time-series data includes: earthquake-related data, rainfall time-series data, geological environment baseline data, disaster logging data, and surface change monitoring data.
3. The method for assessing the risk of landslide geological hazards under the combined effects of earthquake and rainfall as described in claim 1, characterized in that, The earthquake damage factor is calculated by integrating the peak ground acceleration field and the post-earthquake creep rate obtained based on time-series InSAR data, and is used to quantify the degree of damage to the rock and soil mass caused by the earthquake. The dynamic hydrological response factor uses the earthquake damage factor as a correction parameter to correct the topographic humidity index, generating an earthquake-corrected infiltration potential index, which is then calculated in conjunction with effective rainfall or soil moisture.
4. The landslide geological hazard risk assessment method under the combined effects of earthquake and rainfall as described in claim 3, characterized in that, The formula for calculating the earthquake-corrected infiltration potential index is as follows: Where α is the infiltration enhancement coefficient, Earthquake damage factor, This indicates the terrain humidity index.
5. The method for assessing the risk of landslide geological hazards under the combined effects of earthquake and rainfall as described in claim 1, characterized in that, The dual-branch risk assessment model includes a physical mechanism branch and a data-driven branch; The physical mechanism branch dynamically corrects the pore water pressure ratio of the infinite slope model based on the coupling triggering factor to calculate the dynamic safety factor; the data-driven branch uses multi-source time-series data and the coupling triggering factor as input to generate a disaster occurrence probability map based on a neural network model.
6. The method for assessing the risk of landslide geological hazards under the combined effects of earthquake and rainfall as described in claim 5, characterized in that, In the physical mechanism branch, the expression for the dynamic pore water pressure ratio is: In the formula, The background pore water pressure ratio of the region; The pore water pressure response coefficient controls the degree of amplification of pore water pressure. This is the coupling triggering factor.
7. The method for assessing the risk of landslide geological hazards under the combined effects of earthquake and rainfall as described in claim 6, characterized in that, The expression for calculating the dynamic safety factor is: In the formula, It is the effective cohesion of the soil and rock mass; The effective internal friction angle of the rock and soil mass; The saturated unit weight of the soil and rock mass; d β represents the soil layer thickness; β represents the slope gradient. The dynamic pore water pressure ratio.
8. The method for assessing the risk of landslide geological hazards under the combined effects of earthquake and rainfall as described in claim 5, characterized in that, The dual-branch risk assessment model also includes a coupled reasoning module, which is used to fuse the dynamic safety coefficient map output by the physical mechanism branch with the disaster occurrence probability map output by the data-driven branch to generate a comprehensive geological disaster risk probability map and uncertainty assessment.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements a landslide geological hazard risk assessment method under the combined effects of earthquake and rainfall as described in any one of claims 1 to 8.