Geological disaster space-time distribution probability prediction method and system under different rainfall conditions
By combining the exponential decay model and the contribution rate of the discriminant coefficient, the problems of insufficient rainfall accumulation effect and type discrimination in traditional geological disaster prediction methods are solved, and high-precision spatiotemporal distribution probability prediction and gridded early warning are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- YUNNAN POWER GRID CO LTD ELECTRIC POWER RES INST
- Filing Date
- 2025-12-18
- Publication Date
- 2026-05-12
AI Technical Summary
Traditional geological disaster prediction methods fail to fully reflect the cumulative effect and attenuation process of previous rainfall, lack refined discrimination of rainfall process types, cannot distinguish the different inducing mechanisms of short-term heavy rainfall and long-duration rainfall, and the prediction results are mostly qualitative or semi-quantitative, making it difficult to achieve gridded spatiotemporal probability prediction.
The total effective precipitation and segmented cumulative precipitation were calculated using an exponential decay model. By combining the discrimination coefficient and the actual contribution rate, a disaster-rainfall sample set was constructed to refine the identification of the dominant rainfall type that induces disasters and to conduct spatiotemporal distribution probability analysis.
It has improved the accuracy and practicality of geological disaster prediction, realized gridded, high spatial resolution risk prediction, reduced the false alarm rate, and improved the pertinence and accuracy of early warning.
Smart Images
Figure CN122024408A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of geological disaster prediction technology, and in particular to a method and system for predicting the spatiotemporal distribution probability of geological disasters under different rainfall conditions. Background Technology
[0002] Geological disasters are a significant type of natural disaster threatening people's lives and property. Rainfall is a major factor in triggering geological disasters such as landslides, collapses, and debris flows. Traditional methods for predicting rainfall-related geological disasters mainly employ the critical rainfall method or statistical regression models. These methods have significant limitations: first, they only consider total precipitation or daily rainfall, failing to fully reflect the cumulative effect and attenuation process of previous rainfall; second, they lack refined differentiation of rainfall event types, making it impossible to distinguish the different triggering mechanisms of short-term heavy rainfall and long-duration rainfall; and third, the prediction results are mostly qualitative or semi-quantitative, making it difficult to achieve gridded spatiotemporal probability prediction.
[0003] While some existing studies have attempted to introduce the concept of effective precipitation, they often employ simple linear summation or fixed weighting coefficients, failing to accurately reflect the physical processes of soil moisture infiltration and evapotranspiration. Furthermore, they do not adequately consider the differences in rainfall response characteristics among different disaster types (such as landslides, debris flows, and collapses), resulting in limited prediction accuracy. In addition, existing methods often neglect the spatial mismatch between disaster locations and meteorological stations when constructing disaster-rainfall sample sets, directly using data from the nearest station, leading to data correlation errors.
[0004] Therefore, there is an urgent need for a method to predict the spatiotemporal distribution probability of geological disasters that can comprehensively consider the cumulative effect of previous rainfall, the identification of rainfall process types, and the response characteristics of different disaster types, so as to improve the prediction accuracy and practicality. Summary of the Invention
[0005] Based on this, it is necessary to propose a method for predicting the spatiotemporal distribution probability of geological disasters under different rainfall conditions to address the above problems.
[0006] A method for predicting the spatiotemporal distribution probability of geological disasters under different rainfall conditions, the method comprising the following steps: Acquire historical geological disaster data and meteorological precipitation data, and construct a disaster-rainfall sample set; Based on the exponential decay model, the total effective precipitation and segmented cumulative precipitation for each sample are calculated. Based on the comparison between the total effective precipitation and the segmented cumulative precipitation of each sample in the disaster-rainfall sample set and a preset threshold, precipitation-type disaster samples are selected. Based on the discrimination coefficient and actual contribution rate, the precipitation-type disaster samples are discriminated to determine the dominant rainfall type that induces the disaster; Based on the dominant rainfall type, a spatiotemporal distribution probability analysis of geological disasters is conducted to generate geological disaster occurrence prediction results.
[0007] In the above scheme, the construction of the disaster-rainfall sample set specifically includes: For each disaster record in the historical geological disaster database, obtain its occurrence time and geographical location information; Based on the aforementioned geographical location information, a proximity matching algorithm is used in the meteorological precipitation database to associate disaster sites with the meteorological stations that are spatially closest to them. Precipitation data from associated meteorological stations at the time of the disaster and in the preceding period are extracted and combined with disaster records to form a disaster-rainfall sample.
[0008] In the above scheme, calculating the total effective precipitation for each sample specifically includes: The total effective precipitation for each sample is calculated using the following formula:
[0009] in, When i=1, that is, the daily precipitation on the day of the disaster and the (i-1)th day before. Rainfall attenuation coefficient, The total effective precipitation for each sample is N, and the total number of days for the calculation is N.
[0010] In the above scheme, the segmented cumulative precipitation includes short-term cumulative precipitation, medium-term cumulative precipitation, and long-term cumulative precipitation; The aforementioned short-term cumulative precipitation is:
[0011] The aforementioned medium-term cumulative precipitation is:
[0012] The long-term cumulative precipitation is:
[0013] in, When i=1, it refers to the daily precipitation on the day the disaster occurs and the (i-1)th day before.
[0014] In the above scheme, the step of selecting precipitation-type disaster samples based on the comparison between the total effective precipitation and segmented cumulative precipitation of each sample in the disaster-rainfall sample set and a preset threshold specifically includes: Set a first threshold and a second threshold for each sample; If the total effective precipitation of the sample is greater than or equal to the first threshold, or the cumulative precipitation of any segment is greater than or equal to the second threshold, then the sample is determined to be a precipitation-type disaster sample. Otherwise, the sample will be discarded as an irrelevant noise sample.
[0015] In the above scheme, the step of judging the precipitation-type disaster samples based on the discriminant coefficient and actual contribution rate to determine the dominant rainfall type that induces the disaster specifically includes: Determine the discriminant coefficient based on the type of disaster. :
[0016] Among them, m is dynamically adjusted according to different disaster types. At the time, it was a landslide. At that time, mudslides, Time represents collapse, and k represents the rainfall attenuation coefficient; Calculate the actual contribution rate of short-term precipitation in the current sample. :
[0017] in, For the j-th precipitation type disaster sample, the short-term cumulative precipitation, The actual contribution rate of short- and medium-term precipitation This represents the total effective precipitation for the j-th precipitation type disaster sample. Compare the actual contribution rate of short-term precipitation in the current sample. With discriminant coefficient : If the actual contribution rate of short-term precipitation in the current sample Discrimination coefficient If so, the sample is determined to be dominated by short-term precipitation. If the actual contribution rate of short-term precipitation in the current sample ≤Discrimination coefficient If so, the sample is determined to be dominated by long-duration precipitation.
[0018] In the above scheme, the step of performing spatiotemporal distribution probability analysis of geological disasters based on the dominant rainfall type to generate geological disaster occurrence prediction results includes: Based on the dominant rainfall type, the predicted area is dynamically divided into a short-term precipitation-dominant area, a long-duration precipitation-dominant area, and a short-duration heavy precipitation-induced area. For each partition, the historical spatiotemporal distribution patterns of disasters corresponding to the dominant rainfall type are matched; Spatiotemporal distribution patterns include the probability function of the time lag characteristics of disaster occurrence and the spatial density distribution characteristics; Based on the matched spatiotemporal distribution pattern, the calculation of different spatial grids within each partition in the future within a specified time window is performed. The probability of geological disasters occurring within the mouth; The calculation results are compared with the warning level thresholds to generate a data set that includes the probability of disaster occurrence and the dominant rainfall category. A gridded geological disaster risk prediction map based on type, spatiotemporal location, and early warning level.
[0019] This application also proposes a probability prediction system for the spatiotemporal distribution of geological disasters under different rainfall conditions. The system includes: a sample set construction unit, a sample data calculation unit, a sample screening unit, a sample type determination unit, and a spatiotemporal analysis and prediction unit. The sample set construction unit is used to acquire historical geological disaster data and meteorological precipitation data, and to construct a disaster-rainfall sample set; The sample data calculation unit is used to calculate the total effective precipitation and segmented cumulative precipitation for each sample based on the exponential decay model. The sample screening unit is used to screen precipitation-type disaster samples based on the comparison between the total effective precipitation and the segmented cumulative precipitation of each sample in the disaster-rainfall sample set and a preset threshold. The sample type determination unit is used to determine the dominant rainfall type that induces the disaster based on the discrimination coefficient and the actual contribution rate. The spatiotemporal analysis and prediction unit is used to perform spatiotemporal distribution probability analysis of geological disasters based on the dominant rainfall type, and generate geological disaster occurrence prediction results.
[0020] This application also proposes a readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the following steps: Acquire historical geological disaster data and meteorological precipitation data, and construct a disaster-rainfall sample set; Based on the exponential decay model, the total effective precipitation and segmented cumulative precipitation for each sample are calculated. Based on the comparison between the total effective precipitation and the segmented cumulative precipitation of each sample in the disaster-precipitation sample set and the preset threshold, precipitation-type disaster samples are selected. Based on the discrimination coefficient and actual contribution rate, precipitation-type disaster samples are discriminated to determine the dominant rainfall type that induces the disaster; Based on the dominant rainfall type, a spatiotemporal distribution probability analysis of geological disasters is conducted to generate geological disaster occurrence prediction results.
[0021] This application also proposes a computer device, including a memory and a processor, wherein the memory stores a computer program, and the computer program is executed by the processor in the following steps: Acquire historical geological disaster data and meteorological precipitation data, and construct a disaster-rainfall sample set; Based on the exponential decay model, the total effective precipitation and segmented cumulative precipitation for each sample are calculated. Based on the comparison between the total effective precipitation and the segmented cumulative precipitation of each sample in the disaster-precipitation sample set and the preset threshold, precipitation-type disaster samples are selected. Based on the discrimination coefficient and actual contribution rate, precipitation-type disaster samples are discriminated to determine the dominant rainfall type that induces the disaster; Based on the dominant rainfall type, a spatiotemporal distribution probability analysis of geological disasters is conducted to generate geological disaster occurrence prediction results.
[0022] The embodiments of this invention offer the following advantages: By introducing an exponential decay model to calculate total effective precipitation and segmented cumulative precipitation, this invention can accurately simulate the infiltration and dissipation process of water in soil and rock, thereby scientifically quantifying the cumulative contribution of previous rainfall to current slope stability. This effectively solves the problem of missed reporting of long-duration disasters caused by neglecting the lag effect of rainfall in traditional methods, and improves the ability of early warning models to characterize complex rainfall processes. Simultaneously, by constructing a discrimination algorithm based on the discriminant coefficient and actual contribution rate, it can automatically and accurately classify disaster-inducing rainfall into short-term precipitation-dominated, long-duration precipitation-dominated, and short-duration heavy rainfall-induced types. This refined discrimination overcomes the shortcomings of the existing one-size-fits-all threshold method, enabling the early warning system to identify the risk characteristics of different rainfall patterns, significantly improving the targeting and accuracy of early warnings, and effectively reducing the false alarm rate. Finally, based on the identified dominant rainfall type, targeted spatiotemporal distribution probability analysis is performed. By associating different rainfall types with the spatiotemporal patterns of historical disasters, gridded, high spatial resolution risk prediction can be achieved. Attached Figure Description
[0023] 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 some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0024] in: Figure 1 This is a schematic diagram of a method for predicting the spatiotemporal distribution probability of geological disasters under different rainfall conditions in one embodiment; Figure 2 This is a time frequency distribution map of geological disasters in one embodiment. Detailed Implementation
[0025] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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.
[0026] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention; however, it will be apparent to those skilled in the art that the invention may be practiced without one or more of these details; in other instances, certain technical features well-known in the art have not been described in order to avoid confusion with the invention. It should be understood that the invention can be practiced in different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided to make the disclosure thorough and complete and to fully convey the scope of the invention to those skilled in the art.
[0027] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. When used herein, the singular forms “a,” “an,” and “the” are also intended to include the plural forms, unless the context clearly indicates otherwise. The terms “comprising” and / or “including,” when used in this specification, identify the presence of said features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups. When used herein, the term “and / or” includes any and all combinations of the associated listed items.
[0028] To fully understand the present invention, a detailed structure will be presented in the following description in order to illustrate the technical solution proposed by the present invention; optional embodiments of the present invention are described in detail below, however, in addition to these detailed descriptions, the present invention may have other embodiments.
[0029] like Figure 1 As shown, in one embodiment, a method for predicting the spatiotemporal distribution probability of geological hazards under different rainfall conditions is provided. This method includes steps S101 to S105, which are detailed below: S101. Obtain historical geological disaster data and meteorological precipitation data, and construct a disaster-rainfall sample set; Specifically, historical geological disaster data includes the time of occurrence of the disaster accurate to the hour, latitude and longitude coordinates, and disaster types such as landslides, debris flows, and collapses; meteorological precipitation data includes hourly and daily precipitation data from ground observation stations and regional automatic stations.
[0030] This step systematically integrates multi-source heterogeneous data and utilizes spatial matching algorithms to construct a high-precision disaster-rainfall sample set, laying a high-quality, highly correlated data foundation for subsequent accurate predictions. It solves the problem of spatial mismatch between disaster locations and meteorological stations, ensuring that each disaster sample can be accurately associated with its actual rainfall triggering conditions. This fundamentally avoids analytical biases caused by data mismatch and is the essential guarantee for the effectiveness of the entire prediction method.
[0031] In some embodiments, constructing a disaster-rainfall sample set specifically includes: For each disaster record in the historical geological disaster database, obtain its occurrence time and geographical location information; Based on geographic location information, a proximity matching algorithm is used in the meteorological precipitation database to associate disaster sites with the meteorological stations that are spatially closest. Precipitation data from associated meteorological stations at the time of the disaster and in the preceding period are extracted and combined with disaster records to form a disaster-rainfall sample.
[0032] S102. Based on the exponential decay model, calculate the total effective precipitation and segmented cumulative precipitation for each sample. By introducing an exponential decay model to calculate effective precipitation, this approach addresses the core deficiency in existing technologies that fail to adequately consider the dynamic decay of effective precipitation. Instead of simply summing up rainfall, it simulates the infiltration and dissipation of water in the soil, precisely quantifying the cumulative contribution of the lag effect of previous rainfall to current slope stability. This enables the model to effectively identify disaster risks caused by long-duration, slow infiltration, significantly reducing the underreporting rate of such disasters.
[0033] In some embodiments, calculating the total effective precipitation for each sample specifically includes: The total effective precipitation for each sample is calculated using the following formula:
[0034] in, When i=1, that is, the daily precipitation on the day of the disaster and the (i-1)th day before. Rainfall attenuation coefficient, The total effective precipitation for each sample is N, and the total number of days for the calculation is N.
[0035] Specifically, the rainfall attenuation coefficient k is set to 0.78, and the total number of days for calculation is determined according to the type of disaster: for landslide disasters, N is set to 11; for debris flow disasters, N is set to 9; and for collapse disasters, N is set to 8.
[0036] In some embodiments, segmented cumulative precipitation includes short-term cumulative precipitation, medium-term cumulative precipitation, and long-term cumulative precipitation; Short-term cumulative precipitation is:
[0037] The medium-term cumulative precipitation is:
[0038] Long-term cumulative precipitation is:
[0039] in, When i=1, it refers to the daily precipitation on the day the disaster occurs and the (i-1)th day before.
[0040] In some embodiments, the daily precipitation data for the day of the disaster and the preceding 10 days are read to calculate the effective precipitation. The calculation formula is as follows:
[0041] in, The actual precipitation on day i is represented by k, which is the rainfall attenuation coefficient. In this embodiment, k is set to a fixed value of 0.78 based on the measured soil permeability characteristics. Meanwhile, parallel computing for short periods (1-3 days, denoted as...) ), mid-term (4-6 days, denoted as ) and long-term (7~11 days, denoted as Segmented cumulative precipitation; After the calculation is completed, if the sample satisfies If the cumulative precipitation value in any segment exceeds 10 mm, it is judged as a precipitation-type disaster sample; otherwise, it is discarded as irrelevant noise.
[0042] S103. Based on the comparison between the total effective precipitation and the segmented cumulative precipitation of each sample in the disaster-rainfall sample set and the preset threshold, precipitation-type disaster samples are selected. This step, through initial screening using preset thresholds, achieves a dual improvement in computational efficiency and model accuracy. It effectively eliminates disaster samples caused by non-rainfall factors such as earthquakes and human engineering activities, preventing this noisy data from interfering with subsequent analysis of rainfall-induced patterns. This mechanism ensures that the probabilistic model focuses on learning the true physical relationship between rainfall and disasters, improving model purity and predictive accuracy while reducing unnecessary computational burden.
[0043] In some embodiments, precipitation-type disaster samples are selected based on a comparison between the total effective precipitation and the segmented cumulative precipitation of each sample in the disaster-rainfall sample set and a preset threshold. Specifically, this includes: Set a first threshold and a second threshold for each sample; If the total effective precipitation of the sample is greater than or equal to the first threshold, or the cumulative precipitation of any segment is greater than or equal to the second threshold, then the sample is determined to be a precipitation-type disaster sample. Otherwise, the sample will be discarded as an irrelevant noise sample.
[0044] S104. Based on the discrimination coefficient and actual contribution rate, the precipitation-type disaster samples are discriminated to determine the dominant rainfall type that induces the disaster; This step enables refined and automated identification of rainfall-inducing patterns, completely changing the traditional one-size-fits-all approach to early warning. Through a unique discrimination coefficient and contribution rate algorithm, it can clearly distinguish between different inducing mechanisms, such as short-term precipitation-dominated, long-duration precipitation-dominated, and even short-duration heavy precipitation-dominated. This refined identification capability makes early warning criteria more physically meaningful, significantly improving the targeting of warnings and thus significantly reducing the false alarm rate caused by confusion in rainfall patterns.
[0045] In some embodiments, precipitation-type disaster samples are discriminated based on the discrimination coefficient and the actual contribution rate to determine the dominant rainfall type that induces the disaster, specifically including: Determine the discriminant coefficient based on the type of disaster. :
[0046] Among them, m is dynamically adjusted according to different disaster types. At the time, it was a landslide. At that time, mudslides, Time represents collapse, and k represents the rainfall attenuation coefficient; Calculate the actual contribution rate of short-term precipitation in the current sample. :
[0047] in, For the j-th precipitation type disaster sample, the short-term cumulative precipitation, The actual contribution rate of short- and medium-term precipitation This represents the total effective precipitation for the j-th precipitation type disaster sample. Compare the actual contribution rate of short-term precipitation in the current sample. With discriminant coefficient : If the actual contribution rate of short-term precipitation in the current sample Discrimination coefficient If so, the sample is determined to be dominated by short-term precipitation. If the actual contribution rate of short-term precipitation in the current sample ≤Discrimination coefficient If so, the sample is determined to be dominated by long-duration precipitation.
[0048] In some embodiments, a step for identifying short-duration heavy precipitation is also included: Calculate the ratio of the precipitation in the 12 hours prior to the sample to the precipitation in the previous 3 days; If the ratio of the precipitation in the previous 12 hours to the precipitation in the previous 3 days is greater than the first heavy precipitation threshold, the dominant rainfall type of the sample will be corrected to short-term heavy precipitation induced type.
[0049] Preferred, The calculated results are usually in the range of 0.56 to 0.61, and the first heavy precipitation threshold is 0.8.
[0050] S105. Based on the dominant rainfall type, conduct a spatiotemporal distribution probability analysis of geological disasters and generate geological disaster occurrence prediction results.
[0051] This step, based on refined rainfall type identification results, performs spatiotemporal distribution probability analysis. Its technical advantage lies in achieving a leap from vague temporal warnings to precise spatiotemporal predictions. It can generate gridded probabilistic prediction results containing specific latitude, longitude, and time windows, improving spatial accuracy from the traditional administrative county level to the risk grid level. Simultaneously, combined with a feedback optimization mechanism, this application can dynamically respond to real-time rainfall changes and correct prediction results in real time, ensuring the timeliness and practical value of early warning information and providing accurate and reliable decision-making basis for emergency response.
[0052] In some embodiments, a spatiotemporal distribution probability analysis of geological hazards is performed based on the dominant rainfall type to generate geological hazard occurrence prediction results, including: Based on the dominant rainfall type, the forecast area is dynamically divided into short-term precipitation-dominant areas, long-duration precipitation-dominant areas, and short-duration heavy precipitation-induced areas. For each partition, the historical spatiotemporal distribution patterns of disasters corresponding to the dominant rainfall type are matched; Spatiotemporal distribution patterns include the probability function of the time lag characteristics of disaster occurrence and the spatial density distribution characteristics; Based on the matched spatiotemporal distribution pattern, the calculation of different spatial grids within each partition in the future within a specified time window is performed. The probability of geological disasters occurring within the mouth; The calculation results are compared with the warning level thresholds to generate a data set that includes the probability of disaster occurrence and the dominant rainfall category. A gridded geological disaster risk prediction map based on type, spatiotemporal location, and early warning level.
[0053] This application also proposes a probability prediction system for the spatiotemporal distribution of geological disasters under different rainfall conditions. The system includes: a sample set construction unit, a sample data calculation unit, a sample screening unit, a sample type determination unit, and a spatiotemporal analysis and prediction unit. The sample set construction unit is used to acquire historical geological disaster data and meteorological precipitation data, and to construct a disaster-rainfall sample set; The sample data calculation unit is used to calculate the total effective precipitation and segmented cumulative precipitation for each sample based on the exponential decay model. The sample screening unit is used to screen precipitation-type disaster samples based on the comparison between the total effective precipitation and the segmented cumulative precipitation of each sample in the disaster-rainfall sample set and a preset threshold. The sample type determination unit is used to determine the dominant rainfall type that induces the disaster based on the discrimination coefficient and the actual contribution rate. The spatiotemporal analysis and prediction unit is used to perform spatiotemporal distribution probability analysis of geological disasters based on the dominant rainfall type, and generate geological disaster occurrence prediction results.
[0054] The aforementioned spatiotemporal analysis and prediction unit also incorporates an automated feedback optimization loop: the system monitors the probability error of the prediction results in real time, and once the error exceeds 5%, it automatically triggers parameter adjustment commands to re-optimize the spatiotemporal feature weights until the error converges to an acceptable range. Finally, the system transforms the calculated probability of occurrence, spatiotemporal distribution characteristics, and warning levels for each type of disaster into visual charts and gridded maps, which are then displayed to users through the warning platform interface, thereby achieving accurate prediction of the spatiotemporal distribution of geological disasters.
[0055] In some embodiments, the geological disaster spatiotemporal distribution probability prediction system under different rainfall conditions first performs a preliminary screening mechanism for precipitation-type disasters, and the system sets an effective precipitation threshold. Subsequent deep probability calculations are only initiated when the calculated value exceeds this threshold, effectively eliminating interfering samples. Subsequently, the model uses a contribution rate algorithm for refined filtering of rainfall types. The system has built-in tiered thresholds for different rainfall intensities, including those for identifying heavy rainfall events (…). ) and rainstorm events ( The standard is as follows. Specifically, for extremely sudden and intense short-duration precipitation, the model employs a dual-validation strategy: while calculating the intensity ratio, it additionally scans for hourly precipitation exceeding [a certain threshold] in the preceding 24 hours. The frequency of occurrence ensures keen awareness of extreme weather events.
[0056] First, filter for precipitation-type disasters (threshold). Then calculate the contribution rate. Filtering rainfall types, thresholds include heavy precipitation events ( ) and rainstorm events ( For short-duration heavy precipitation, an additional check is performed on the frequency of hourly precipitation exceeding 20 mm / h in the preceding 24 hours. Probability calculations provide spatiotemporal location (latitude and longitude, time window), categorizing it into type zones: short-term dominant zone (…). ), long-duration time zone ( ), short-term strong area ( ).
[0057] Based on the above calculations, the system accurately outputs spatiotemporal location information including latitude and longitude coordinates and time windows, and dynamically divides the prediction area into three dominant zones: those that meet the contribution rate... Greater than the discriminant coefficient The dominant area of short-term precipitation ( ),satisfy Less than or equal to The dominant area of long-duration precipitation ( ), and areas of short-duration heavy precipitation that meet the characteristics of high intensity ratio ( Or high-frequency heavy rainfall).
[0058] In addition, the system can automatically adjust the effective precipitation calculation period according to the physical characteristics of different disasters: for landslide disasters, the total effective precipitation calculation is based on 11 days (m=11); for debris flow, it is based on 9 days (m=9); and for landslide, it is based on 8 days (m=8).
[0059] In addition, based on a sample of 1868 precipitation-related disasters, the data showed: 1305 landslides (70%), 328 debris flows (18%), and 194 collapses (10%). Short-term precipitation accounted for 78%–83%, with 88%–96% of events occurring within the first 3 days, and short-duration heavy rainfall accounting for 14%–23%. The temporal results are as follows: Figure 2 As shown, the most frequent short-duration heavy rainfall occurs 1-4 hours before a landslide, with the most concentrated rainfall occurring 1 hour before the debris flow. Error analysis shows a prediction accuracy >95%.
[0060] In summary, this invention provides a probabilistic discrimination model based on different rainfall conditions, including a formula for calculating effective precipitation and a discrimination coefficient. and contribution rate The application enables the prediction of the spatiotemporal distribution probability of landslides, debris flows, and collapses.
[0061] This application also proposes a readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the following steps: Acquire historical geological disaster data and meteorological precipitation data, and construct a disaster-rainfall sample set; Based on the exponential decay model, the total effective precipitation and segmented cumulative precipitation for each sample are calculated. Based on the comparison between the total effective precipitation and the segmented cumulative precipitation of each sample in the disaster-precipitation sample set and the preset threshold, precipitation-type disaster samples are selected. Based on the discrimination coefficient and actual contribution rate, precipitation-type disaster samples are discriminated to determine the dominant rainfall type that induces the disaster; Based on the dominant rainfall type, a spatiotemporal distribution probability analysis of geological disasters is conducted to generate geological disaster occurrence prediction results.
[0062] This application also proposes a computer device, including a memory and a processor, wherein the memory stores a computer program, and the computer program is executed by the processor in the following steps: Acquire historical geological disaster data and meteorological precipitation data, and construct a disaster-rainfall sample set; Based on the exponential decay model, the total effective precipitation and segmented cumulative precipitation for each sample are calculated. Based on the comparison between the total effective precipitation and the segmented cumulative precipitation of each sample in the disaster-precipitation sample set and the preset threshold, precipitation-type disaster samples are selected. Based on the discrimination coefficient and actual contribution rate, precipitation-type disaster samples are discriminated to determine the dominant rainfall type that induces the disaster; Based on the dominant rainfall type, a spatiotemporal distribution probability analysis of geological disasters is conducted to generate geological disaster occurrence prediction results.
[0063] Those skilled in the art will understand that implementing all or part of the processes in the above embodiments can be accomplished by instructing related hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0064] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0065] The embodiments described above are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application's patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. The embodiments disclosed above are merely preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, equivalent variations made according to the claims of this invention are still within the scope of this invention.
Claims
1. A method for predicting the spatiotemporal distribution probability of geological disasters under different rainfall conditions, characterized in that, The method includes: Acquire historical geological disaster data and meteorological precipitation data, and construct a disaster-rainfall sample set; Based on the exponential decay model, the total effective precipitation and segmented cumulative precipitation for each sample are calculated. Based on the comparison between the total effective precipitation and the segmented cumulative precipitation of each sample in the disaster-rainfall sample set and a preset threshold, precipitation-type disaster samples are selected. Based on the discrimination coefficient and actual contribution rate, the precipitation-type disaster samples are discriminated to determine the dominant rainfall type that induces the disaster; Based on the dominant rainfall type, a spatiotemporal distribution probability analysis of geological disasters is conducted to generate geological disaster occurrence prediction results.
2. The method for predicting the spatiotemporal distribution probability of geological disasters under different rainfall conditions according to claim 1, characterized in that, The construction of the disaster-rainfall sample set specifically includes: For each disaster record in the historical geological disaster database, obtain its occurrence time and geographical location information; Based on the aforementioned geographical location information, a proximity matching algorithm is used in the meteorological precipitation database to associate disaster sites with the meteorological stations that are spatially closest to them. Precipitation data from associated meteorological stations at the time of the disaster and in the preceding period are extracted and combined with disaster records to form a disaster-rainfall sample.
3. The method for predicting the spatiotemporal distribution probability of geological disasters under different rainfall conditions according to claim 2, characterized in that, The calculation of the total effective precipitation for each sample specifically includes: The total effective precipitation for each sample is calculated using the following formula: in, When i=1, that is, the daily precipitation on the day of the disaster and the (i-1)th day before. Rainfall attenuation coefficient, The total effective precipitation for each sample is N, and the total number of days for the calculation is N.
4. The method for predicting the spatiotemporal distribution probability of geological disasters under different rainfall conditions according to claim 3, characterized in that, The segmented cumulative precipitation includes short-term cumulative precipitation, medium-term cumulative precipitation, and long-term cumulative precipitation; The aforementioned short-term cumulative precipitation is: The aforementioned medium-term cumulative precipitation is: The long-term cumulative precipitation is: in, When i=1, it refers to the daily precipitation on the day the disaster occurs and the (i-1)th day before.
5. The method for predicting the spatiotemporal distribution probability of geological disasters under different rainfall conditions according to claim 4, characterized in that, The step of selecting precipitation-type disaster samples based on the comparison between the total effective precipitation and segmented cumulative precipitation of each sample in the disaster-rainfall sample set and a preset threshold specifically includes: Set a first threshold and a second threshold for each sample; If the total effective precipitation of the sample is greater than or equal to the first threshold, or the cumulative precipitation of any segment is greater than or equal to the second threshold, then the sample is determined to be a precipitation-type disaster sample. Otherwise, the sample will be discarded as an irrelevant noise sample.
6. The method for predicting the spatiotemporal distribution probability of geological disasters under different rainfall conditions according to claim 5, characterized in that, The process of judging the precipitation-type disaster samples based on the discriminant coefficient and actual contribution rate to determine the dominant rainfall type that induces the disaster specifically includes: Determine the discriminant coefficient based on the type of disaster. : Among them, m is dynamically adjusted according to different disaster types. At the time, it was a landslide. At that time, mudslides, Time represents collapse, and k represents the rainfall attenuation coefficient; Calculate the actual contribution rate of short-term precipitation in the current sample. : in, For the j-th precipitation type disaster sample, the short-term cumulative precipitation, The actual contribution rate of short- and medium-term precipitation This represents the total effective precipitation for the j-th precipitation type disaster sample. Compare the actual contribution rate of short-term precipitation in the current sample. With discriminant coefficient : If the actual contribution rate of short-term precipitation in the current sample Discrimination coefficient If so, the sample is determined to be dominated by short-term precipitation. If the actual contribution rate of short-term precipitation in the current sample ≤Discrimination coefficient If so, the sample is determined to be dominated by long-duration precipitation.
7. The method for predicting the spatiotemporal distribution probability of geological disasters under different rainfall conditions according to claim 6, characterized in that, The step of performing a spatiotemporal distribution probability analysis of geological disasters based on the dominant rainfall type to generate geological disaster occurrence prediction results includes: Based on the dominant rainfall type, the predicted area is dynamically divided into a short-term precipitation-dominant area, a long-duration precipitation-dominant area, and a short-duration heavy precipitation-induced area. For each partition, the historical spatiotemporal distribution patterns of disasters corresponding to the dominant rainfall type are matched; Spatiotemporal distribution patterns include the probability function of the time lag characteristics of disaster occurrence and the spatial density distribution characteristics; Based on the matched spatiotemporal distribution pattern, the calculation of different spatial grids within each partition in the future within a specified time window is performed. The probability of geological disasters occurring within the mouth; The calculation results are compared with the warning level thresholds to generate a data set that includes the probability of disaster occurrence and the dominant rainfall category. A gridded geological disaster risk prediction map based on type, spatiotemporal location, and early warning level.
8. A system for predicting the spatiotemporal distribution probability of geological disasters under different rainfall conditions, characterized in that, The system includes: a sample set construction unit, a sample data calculation unit, a sample screening unit, a sample type determination unit, and a spatiotemporal analysis and prediction unit; The sample set construction unit is used to acquire historical geological disaster data and meteorological precipitation data, and to construct a disaster-rainfall sample set; The sample data calculation unit is used to calculate the total effective precipitation and segmented cumulative precipitation for each sample based on the exponential decay model. The sample screening unit is used to screen precipitation-type disaster samples based on the comparison between the total effective precipitation and the segmented cumulative precipitation of each sample in the disaster-rainfall sample set and a preset threshold. The sample type determination unit is used to determine the precipitation-type disaster sample based on the discrimination coefficient and the actual contribution rate, and to determine the dominant rainfall type that induces the disaster. The spatiotemporal analysis and prediction unit is used to perform spatiotemporal distribution probability analysis of geological disasters based on the dominant rainfall type, and generate geological disaster occurrence prediction results.
9. A readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, the processor performs the steps of the method as described in any one of claims 1 to 7.
10. A computer device, comprising a memory and a processor, characterized in that, The memory stores a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 7.