A monitoring and early warning system and method for glacial lake outburst in high altitude areas

By combining satellite remote sensing and LiDAR data, the ice loss area and fault layer are analyzed, and the early warning frequency is dynamically adjusted, the problem of early warning of ice lake collapse is solved, the accuracy and timeliness of early warning are improved, and the efficiency of response to ice lake collapse incidents is ensured.

CN118865646BActive Publication Date: 2024-12-06CHINA GEOLOGICAL SURVEY MILITARY-CIVILIAN INTEGRATED GEOLOGICAL SURVEY CENT
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411338747.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-25
Publication Date
2024-12-06
Estimated Expiration
2044-09-25

AI Technical Summary

Technical Problem

The existing technology is difficult to early warning of ice lake collapse in advance, especially the hidden sub-ice erosion situation, and it is difficult to early warning in advance through conventional detection methods, resulting in the sudden nature of ice lake collapse in progress, which makes it impossible to effectively early warning.

Method used

Using satellite remote sensing technology and sensor equipment, through image processing and LiDAR data analysis, the ice loss area and fault layer are determined, the estimated volume is calculated, the early warning frequency is dynamically adjusted, and the periodic flow leakage strategy is implemented to improve the flexibility and accuracy of early warning. .

Benefits of technology

An effective warning of ice stability is achieved, potential ice collapse risks are identified in advance, and the risk of ice lake collapse is reduced, and the effectiveness and timeliness of early warnings are improved, and personnel safety is provided.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118865646B_ABST
    Figure CN118865646B_ABST
Patent Text Reader

Abstract

The present invention discloses a monitoring and early warning system and method for ice lake burst in high altitude areas, which relates to the field of ice lake monitoring and early warning technology. The early warning system comprises an ice layer determination module, a preliminary early warning module, a fault analysis module and an early warning adjustment module which are operated in sequence. The technical key points are as follows: through precise ice layer stability analysis and dynamic early warning frequency adjustment, the risk of ice layer burst can be judged more accurately, and the safety of personnel and equipment can be improved; at the same time, the scientific discharge strategy execution also provides clearer guidance for decision makers, and improves the efficiency of responding to ice layer burst events; the technical solution takes into account the volume of different ice lake areas, the estimated volume of broken ice layers and the fixed early warning frequency factors, so that the early warning system is more flexible and adaptable to different actual situations.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of glacial lake monitoring and early warning, and in particular to a glacial lake burst monitoring and early warning system and method for high-altitude areas. Background Art

[0002] GLBOB monitoring and early warning involves real-time monitoring of glacial lake water levels, temperatures, and topographic changes through a series of technical means and measures. Based on this data, the risk of GLBOB is assessed and early warning is issued, enabling timely response measures to avoid or reduce potential losses from GLBOB disasters. Key monitoring and early warning methods include:

[0003] Satellite remote sensing monitoring: Utilizing high-precision satellite remote sensing technology, dynamic monitoring of the glacial lake's surface area and water level changes is carried out to understand its basic conditions. This method offers advantages such as wide coverage and rapid data acquisition, making it a key means of monitoring glacial lake outbursts. On-site monitoring equipment: Water level gauges, thermometers, pressure sensors, and other monitoring equipment are installed around the glacial lake to monitor key parameters such as the lake's water level and temperature in real time. These devices can transmit the monitoring data to a central server for analysis and processing. Manual inspections: Professional personnel conduct on-site inspections to observe changes in the glacial lake's topography and the accumulation of moraine, providing supplementary information for satellite remote sensing and on-site monitoring equipment.

[0004] However, only mainstream or conventional means are used for monitoring and early warning of glacial lake outbursts. For some special cases, such as subglacial erosion, it is difficult to provide early warning through conventional detection methods because the formation and expansion process of subglacial tunnels or erosion spaces is relatively hidden and difficult to detect. Long-term erosion can also lead to sudden ice avalanches. This unexpected situation can also cause glacial lake outbursts, and due to its sudden nature, it is currently impossible to provide effective early warning through conventional methods. Summary of the Invention

[0005] (1) Technical problems solved

[0006] In response to the shortcomings of the existing technology, the present invention provides a glacial lake burst monitoring and early warning system and method for high-altitude areas. This technical solution takes into account the volume of different glacial lake areas, the estimated volume of the broken ice layer, and the fixed early warning frequency factors, making the early warning system more flexible and adaptable to different actual situations, and solving the problems raised in the background technology.

[0007] (2) Technical solution

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0009] A glacial lake outburst monitoring and early warning system for high-altitude areas, comprising:

[0010] The ice layer determination module selects a glacial lake area and uses satellite remote sensing technology and sensor equipment to collect image data at the junction of the glacial lake and the ice layer. The image data is then processed and analyzed to determine the estimated area S of the ice loss area and its changing trend.

[0011] The preliminary warning module compares the estimated area S of the current ice loss area with the set warning area M. If the estimated area S of the ice loss area exceeds the warning area M, the warning judgment mechanism is triggered, and the change trend of the estimated area S of the ice loss area is extracted to determine whether the current ice change rate has reached the preset warning value. If so, a first-level warning signal is issued and the fault analysis strategy is triggered; if not, no response action is taken;

[0012] The fault analysis module uses LiDAR data to generate an ice thickness map and identifies H discontinuous layers. When H = 1, the discontinuous layer is used as the target fracture layer. When H ≠ 1, the cross-sectional area of ​​different discontinuous layers is continuously monitored, and the layer with the largest cross-sectional area is selected as the target fracture layer. Using the target fracture layer as the segmentation contour, image processing software is used to obtain the cross-sectional area and lateral length of the fractured ice layer. The estimated volume of the fractured ice layer is calculated based on the error rate of the image processing software. The module then collects and builds a computational model based on the pre-processed estimated volume, the volume of the corresponding glacial lake area, and the fixed warning frequency to generate an estimated value for the dynamic warning frequency.

[0013] The early warning adjustment module adjusts the frequency of the current second-level early warning signal according to the dynamic early warning frequency estimate, and implements the regular discharge strategy according to the dynamic early warning frequency estimate, and performs regular gate opening operations at the discharge outlet of the corresponding glacial lake.

[0014] Furthermore, sensor equipment includes unmanned underwater vehicles and remotely operated underwater vehicles that are equipped with cameras;

[0015] The process of processing and analyzing image data is as follows:

[0016] The collected image data were processed using an image classification algorithm to distinguish between ice and non-ice areas. A geographic information system software was used to obtain an ice cross-sectional distribution map, and buoys fitted to the ice edge were used to determine the three endpoints of the ice loss area in the ice cross-sectional distribution map. An algorithm based on peak and trough detection was used to calculate the estimated area S of the ice loss area. The estimated area S was combined with a time series, and statistical analysis tools were used to analyze the changing trend of the estimated area S of the ice loss area.

[0017] Furthermore, when determining the three endpoints of the ice loss area, the buoy that is in contact with the edge of the ice layer and floating on the ice lake is point A, the position of the buoy perpendicular to the ice lake bed is point B, and the intersection of the ice lake bed and the ice layer is point C. The area enclosed by points A, B, and C is the ice loss area.

[0018] Furthermore, under the condition that the ice edge presents a wave curve, the steps of running the peak and trough detection algorithm are as follows:

[0019] Peak and trough detection: Traverse the contour line corresponding to the ice edge between points A and C, record the coordinates of each point, calculate the curvature of each point to identify peaks and troughs, and use the detected peaks and troughs as segmentation points;

[0020] Adjust the number of selection points between segmentation points: obtain the straight-line distance between adjacent segmentation points, build a rule engine, set a fixed threshold, use the fixed threshold as the interval between adjacent segmentation points, select a corresponding number of selection points, and evenly set selection points within each segment formed between adjacent segmentation points;

[0021] Calculation of the estimated ice loss area: A small area is formed between any two adjacent points between points A and C. The small area is a trapezoidal or triangular area. Use the corresponding area calculation formula: area of ​​each small area = 1 / 2 * (upper base + lower base) * height. Add up the areas of all small areas to obtain the estimated area S of the ice loss area.

[0022] Furthermore, the process of determining whether the current ice layer change rate has reached a preset warning value is as follows:

[0023] Collect time series data: record a series of time points t1, t2, ..., t n and the corresponding estimated ice loss areas S1, S2, ..., S n ; Use Python to perform linear regression analysis and build a linear regression model to obtain the value of β1;

[0024] The linear regression model is expressed as follows:

[0025] S t =β0+β1*t+∈ t ;

[0026] Where S t represents the estimated area of ​​ice loss at time point t, β0 represents the intercept term, β1 represents the slope term, ∈ t Represents the error term; the linear regression model is used to calculate the value of β1, which is the speed at which the estimated area S of the ice loss region changes over time, that is, the current ice change rate.

[0027] Furthermore, the process of generating the ice thickness map is as follows:

[0028] The target ice layer is scanned using a LiDAR system, and the reflected LiDAR data, which contains the three-dimensional coordinate information of the ice surface, is collected. Noise points are removed and filtered from the collected LiDAR data. The LiDAR data is then analyzed using TerraScan software to generate a digital elevation model of the ice surface. The known sea level data and the generated ice surface DEM are combined to calculate the ice thickness and create an ice thickness map.

[0029] Furthermore, after H discontinuous layers are identified, the process of obtaining the target fault layer is as follows:

[0030] On an ice thickness map, discontinuities appear as areas of sudden thickness changes. Ground-penetrating radar (GPR) is used to analyze the ice thickness map to identify H discontinuities. When H ≠ 1, edge detection technology is used to determine the boundary of each identified discontinuity on the ice thickness map. After determining the boundary of each discontinuity, a GIS system is used to calculate the cross-sectional area of ​​each discontinuity. All discontinuities of different cross-sectional areas are compared, and the one with the largest cross-sectional area is selected as the target fracture layer.

[0031] Furthermore, the process of calculating the estimated volume of the fractured ice layer is as follows:

[0032] Image processing software was used to analyze the ice thickness map. The target fracture layer was used as the edge, and the outline of the ice layer to be dropped was extracted along the target fracture layer. The area enclosed by the outline, i.e., the cross-sectional area of ​​the ice layer to be dropped, was calculated. On the ice thickness map, the measurement tools in the image processing software were used to measure the length of the target fracture layer along its horizontal direction, i.e., the transverse length. The formula used to calculate the estimated volume of the fractured ice layer was as follows:

[0033] V drop =(A section +δ A *A section )*(L horizontal +δ L *L horizontal );

[0034] Where V drop represents the estimated volume of the fractured ice layer, A section Indicates the cross-sectional area of ​​the ice layer to be dropped, L horizontal represents the lateral length of the target fault layer, δ A Indicates the error rate of the cross-sectional area obtained by image analysis, δ L Indicates the error rate of the measured horizontal length.

[0035] Furthermore, the process of generating dynamic warning frequency estimates is as follows:

[0036]

[0037] R=V drop / V la ;(2)

[0038] In formula (1), f adj represents the estimated value of dynamic warning frequency, f bas represents the fixed warning frequency, α represents the adjustment factor used to control the amplitude of the frequency increase, R represents the ratio of the estimated volume of the broken ice layer to the volume of the corresponding glacial lake area, and R0 represents the set risk threshold; in formula (2), V la Indicates the volume of the corresponding glacial lake area.

[0039] A glacial lake outburst monitoring and early warning method for high-altitude areas includes the following steps:

[0040] S1. Select a glacial lake area and use satellite remote sensing technology and sensor equipment to collect image data at the junction of the glacial lake and the ice layer. Then process and analyze the image data to determine the estimated area S of the ice loss area and its changing trend;

[0041] S2. Compare the estimated area S of the current ice loss region with the set warning area M. If the estimated area S of the ice loss region exceeds the warning area M, trigger the warning judgment mechanism, extract the change trend of the estimated area S of the ice loss region, and determine whether the current ice change rate reaches the preset warning value. If so, issue a first-level warning signal and trigger the fault analysis strategy; if not, do not take any response action;

[0042] S3. After generating an ice thickness map using LiDAR data, H discontinuous layers are identified. When H = 1, the discontinuous layer is used as the target fracture layer. When H ≠ 1, the cross-sectional areas of the different discontinuous layers are continuously monitored, and the layer with the largest cross-sectional area is selected as the target fracture layer. Using the target fracture layer as the segmentation contour, image processing software is used to obtain the cross-sectional area and lateral length of the fractured ice layer. Combined with the error rate of the image processing software, the estimated volume of the fractured ice layer is calculated. The pre-processed estimated volume, the volume of the corresponding glacial lake area, and the fixed warning frequency are collected and used to build a computational model to generate an estimated value for the dynamic warning frequency.

[0043] S4. Adjust the frequency of issuing the current second-level warning signal based on the dynamic warning frequency estimate, and implement a regular discharge strategy based on the dynamic warning frequency estimate, and perform regular gate opening operations at the discharge outlet of the corresponding glacial lake.

[0044] (3) Beneficial effects

[0045] The present invention provides a glacial lake outburst monitoring and early warning system and method for high-altitude areas, which has the following beneficial effects:

[0046] This solution uses an algorithm based on peak and trough detection to accurately calculate the estimated area of ​​ice loss, even when the ice edge exhibits a wavy curve. This method takes into account the complex shape of the ice edge, improving the accuracy of area calculation. Furthermore, by comparing the current estimated area of ​​ice loss with the set warning area and analyzing the rate of ice change, it provides an effective early warning of ice stability. This early warning mechanism can identify potential ice collapse risks in advance, providing a crucial safeguard for personnel safety.

[0047] This solution generates ice thickness maps using LiDAR data and combines them with ground-penetrating radar (GPR) analysis to accurately identify discontinuities and subsequently determine the target fracture layer. This fracture layer serves as a dividing line, facilitating the subsequent estimation of the volume of the fractured ice layer. Based on this estimated volume and the corresponding glacial lake volume, the warning frequency can be dynamically adjusted. A regular discharge strategy is implemented based on the dynamic warning frequency estimate, helping to reduce the risk of glacial lake outbursts while also improving the effectiveness and timeliness of warnings.

[0048] Through precise ice stability analysis and dynamic warning frequency adjustment, this solution can more accurately determine the risk of ice burst and improve the safety of personnel and equipment. At the same time, the implementation of scientific discharge strategies also provides clearer guidance for decision-makers and improves the efficiency of responding to ice burst events. This technical solution takes into account the volume of different glacial lake areas, the estimated volume of broken ice layers, and the fixed warning frequency factors, making the warning system more flexible and adaptable to different actual situations. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is a modular schematic diagram of a glacial lake outburst monitoring and early warning system for high-altitude areas according to the present invention;

[0050] Figure 2 This is the actual reference map for monitoring and early warning of glacial lakes in the present invention;

[0051] Figure 3 The present invention is a flowchart of the steps of a glacial lake outburst monitoring and early warning method for high-altitude areas. DETAILED DESCRIPTION

[0052] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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.

[0053] Glacial lake outbursts at high altitudes are usually caused by a combination of factors;

[0054] Some common or major causes of GLOF include global warming, glacial lake dam stability, extreme weather events, and glacial activity. Based on these major factors, the following approaches are used for early warning and monitoring of GLOF:

[0055] Strengthen remote sensing monitoring: Regularly monitor glacial lakes in high-altitude areas using satellite remote sensing technology to promptly identify changing trends in parameters such as glacial lake area and water level, providing data support for early warning. Establish a comprehensive monitoring network: Establish a comprehensive monitoring network around glacial lakes, including water level monitoring stations and meteorological observation stations, to monitor the hydrological and meteorological conditions of glacial lakes and glacial activity in real time. Conduct on-site inspections and risk assessments: Regularly organize expert teams to conduct on-site inspections and risk assessments of glacial lakes in high-altitude areas to understand the stability of glacial lake embankments, surrounding geological conditions, and potential risk factors. Strengthen international cooperation and exchanges: Glacial lake outbursts are a global natural disaster problem, requiring countries to strengthen cooperation and exchanges, share monitoring data, research results, and early warning experiences, and enhance the global capacity to respond to glacial lake outburst disasters.

[0056] However, in addition to the conventional or main causes of glacial lake outbursts mentioned above, there is also subglacial erosion. This situation is difficult to warn in advance through conventional detection methods because the formation and expansion of subglacial tunnels or erosion spaces are relatively hidden and difficult to detect. Long-term erosion can also lead to sudden ice avalanches. This unexpected situation can also cause glacial lake outbursts, and due to its sudden nature, it is currently impossible to provide effective warnings through conventional methods.

[0057] Therefore, in response to the special glacial lake outburst situation under the above-mentioned chain reaction, optical imaging and SAR data are combined to achieve preliminary early warning, and dynamic monitoring and analysis of the fault conditions of the ice layer in the glacial lake are carried out, thereby achieving the effect of dynamically adjusting the early warning frequency to cope with sudden glacial lake outbursts. To a certain extent, effective and accurate early warning processing is achieved, ensuring the safety of personnel.

[0058] Example 1:

[0059] See also Figure 1 and Figure 2This embodiment provides a glacial lake outburst monitoring and early warning system for high-altitude areas. The early warning system includes an ice layer determination module, a preliminary early warning module, a fault analysis module, and an early warning adjustment module that operate in sequence.

[0060] Ice Determination Module:

[0061] Select a glacial lake area and use satellite remote sensing technology and sensor equipment to collect image data at the junction of the glacial lake and the ice layer. Then process and analyze the image data to determine the estimated area S of the ice loss area and its changing trend.

[0062] Satellite remote sensing monitoring: Using high-resolution satellite remote sensing technology, high-definition images of the study area (i.e., the interface between the glacial lake and the ice sheet) are regularly acquired to capture subtle differences in ice distribution and changes.

[0063] Sensor equipment includes unmanned underwater vehicles (AUVs), autonomous underwater vehicles. AUVs can penetrate deep underwater, carrying sensors such as high-definition cameras and sonar, to conduct close-up observations and measurements of the bottom and edge of the ice, autonomously navigate complex underwater environments, and collect high-quality data. Remotely operated vehicles (ROVs), tethered remotely operated vehicles (ROVs), are connected to a control station on the surface via a cable, allowing operators to control the ROV's movements and the working status of its sensors in real time. ROVs carry more powerful lighting and video equipment, as well as a variety of scientific instruments, enabling more detailed observation and sampling of underwater ice.

[0064] The process of processing and analyzing image data is as follows:

[0065] Image classification and recognition: Using image classification algorithms, collected image data is automatically processed to accurately distinguish between ice and non-ice areas. Accurate area calculation: Using geographic information system (GIS) software, ice cross-sectional distribution maps are obtained. Buoys aligned with the ice edge are used to determine the three endpoints of the ice loss area within the ice cross-sectional distribution map. An algorithm based on peak and trough detection is then used to calculate the estimated area S of the ice loss area. Time series and trend analysis: The estimated area S is combined with the time series, and statistical analysis tools (such as linear regression and trend prediction) are used to analyze the changing trends of the estimated ice loss area S.

[0066] Among them, the image classification tools (such as supervised classification and unsupervised classification) in GIS software, that is, the corresponding image classification algorithms, are used to automatically process the collected images and accurately distinguish between ice and non-ice areas;

[0067] When determining the three endpoints of the ice loss area, the buoy that is in contact with the edge of the ice layer and floating on the ice lake is point A, the position of the buoy perpendicular to the ice lake bed is point B, and the intersection of the ice lake bed and the ice layer is point C. The area enclosed by points A, B, and C is the ice loss area. The distance between AB is obtained by the infrared rangefinder mounted on the buoy. For an underwater robot equipped with an infrared receiving mechanism at point B, the infrared rangefinder emits infrared light perpendicular to the ice lake bed. After the infrared receiving mechanism receives the light, the initial position of the underwater robot is determined to be point B. The underwater robot moves in a straight line toward the ice layer from point B until it is blocked by the ice layer and cannot move, which means it reaches point C. The movement distance of the underwater robot is the distance between BC.

[0068] Normally, the junction between the ice layer and the bottom of the glacial lake is smooth, so the position of the underwater robot when it moves to the ice layer can represent the distance between BC. The distances and positions of points A, B, and C are displayed on the ice layer cross-section distribution map, which facilitates the subsequent acquisition of the vertical distances between different segmented points or evenly selected points and AB.

[0069] Under the condition that the ice edge presents a wave curve, the steps of selecting the peak and trough detection algorithm are as follows:

[0070] Peak and trough detection:

[0071] Traverse the contour line corresponding to the edge of the ice layer between points A and C, record the coordinates of each point, and calculate the curvature or second-order derivative of each point to identify the peaks and troughs. The points with the maximum curvature correspond to the peaks, and the points with the minimum curvature correspond to the troughs. The detected peaks and troughs are used as segmentation points.

[0072] Adjust the number of selection points between segment points:

[0073] Obtain the straight-line distance between adjacent segmentation points, build a rule engine, set a fixed threshold, use the fixed threshold as the interval between adjacent segmentation points, select a corresponding number of selection points, and evenly set selection points within each segment formed between adjacent segmentation points to ensure the approximate accuracy of the curve; the curve determines the number of points required to approximate the curve in each segment based on the distance. The larger the distance, the more points are selected.

[0074] The fixed threshold value is set as needed, and the fixed threshold value is greater than 1 and is a positive integer;

[0075] Calculation of estimated ice loss area:

[0076] A small area is formed between any two adjacent points between points A and C. The small area is a trapezoidal or triangular area, and the corresponding area calculation formula is used; the area formula is: the area of ​​the corresponding trapezoidal or triangular area = 1 / 2 * (upper base + lower base) * height. The area of ​​all small areas is added up to obtain the estimated area S of the ice loss area.

[0077] Initial warning module:

[0078] The estimated area S of the current ice loss area is compared with the set warning area M. If the estimated area S of the ice loss area exceeds the warning area M, the early warning judgment mechanism is triggered, the change trend of the estimated area S of the ice loss area is extracted, and it is determined whether the current ice change rate has reached the preset warning value. If it has reached the value, a first-level early warning signal is issued to remind staff to stay away from the area above the ice loss area and trigger the fault analysis strategy. If it has not reached the value, no response action is taken.

[0079] The setting process of the warning area M is as follows:

[0080] Collect historical data: Count the area lost in past ice loss events and record the severity of the corresponding events; Determine thresholds: Through statistical analysis, identify the ice loss area threshold that leads to serious consequences (such as ice collapse and casualties); Set safety margins: Set a safety margin based on current ice conditions and environmental factors; Calculate the warning area M: Warning area M = historical threshold - safety margin;

[0081] For example, suppose data from past ice loss events in a certain region shows that ice stability begins to decline significantly when the ice loss area exceeds 1,000 square meters. At the same time, considering that the current ice thickness is relatively thin and the recent temperature is high, for safety reasons, a safety margin of 200 square meters is used, and the emergency area M is set to 800 square meters. If the current ice thickness is relatively thick and the recent temperature is low, the corresponding safety margin is 0, and the emergency area M is set to 1,000 square meters.

[0082] The determination of the above historical thresholds and safety margins relies on statistical analysis and expert judgment;

[0083] The process of determining whether the current ice layer change rate has reached the preset warning value is as follows:

[0084] When calculating the current ice layer change rate, the trend analysis method in time series analysis is used, and a linear regression model is adopted to calculate the change rate of the estimated ice loss area S over time, which is the current ice layer change rate;

[0085] Set a series of time points t1, t2, ..., t n and the corresponding estimated ice loss areas S1, S2, ..., Sn , the linear regression model is expressed as follows:

[0086] S t =β0+β1*t+∈ t ;

[0087] Where S t represents the estimated area of ​​ice loss at time point t, β0 represents the intercept term, β1 represents the slope term, ∈ t Represents the error term; through the linear regression model, the value of β1 is calculated, which represents the rate of change of the estimated area S of the ice loss area over time, that is, the current ice change rate;

[0088] The specific calculation steps are as follows:

[0089] After collecting time series data, use statistical software or programming tools (such as Python and R) to perform linear regression analysis to estimate the values ​​of β0 and β1. The value of β1 is the rate of change of the estimated area S of the current ice loss area over time.

[0090] The process of presetting the warning value is as follows:

[0091] Collect historical speed data: Count changes in ice loss rates over the past; Determine baseline speed: Select the average loss rate over a past time period T as the baseline speed; Set a warning growth rate: Based on current ice conditions and environmental factors, set a warning growth rate, indicating that when the loss rate reaches this value, serious consequences may occur; Calculate the warning value: Warning value = baseline speed - warning growth rate;

[0092] For example, suppose data from past ice loss events in a certain region shows that the average loss rate over the past time period T was 1,000 square meters per hour. Considering the current thin ice thickness and recent high temperatures, for safety reasons, the warning growth rate is set to 200 square meters per hour, and the warning value is set to 800 square meters per hour. If the current ice thickness is thicker and the recent temperature is lower, the corresponding warning growth rate is 0, and the warning value is set to 1,000 square meters per hour.

[0093] The determination of the above-mentioned baseline speed and warning growth rate relies on computational analysis and expert judgment.

[0094] Specifically, by combining satellite remote sensing technology with a variety of sensor equipment (such as AUVs and ROVs), high-precision, multi-dimensional monitoring of the interface between ice and glacial lakes was achieved. Using image classification algorithms and GIS software, it was possible to accurately distinguish between ice and non-ice areas, improving the accuracy of identifying ice loss areas.

[0095] By employing an algorithm based on peak and trough detection, the estimated area of ​​ice loss can be accurately calculated even when the ice edge exhibits a wavy curve. This method takes into account the complex shape of the ice edge, improving the accuracy of area calculation. Furthermore, by comparing the current estimated ice loss area with the set warning area and analyzing the rate of ice change, an effective early warning of ice stability is achieved. This early warning mechanism can identify potential ice collapse risks in advance, providing a crucial safeguard for personnel safety.

[0096] In summary, this technical solution has achieved significant progress and beneficial effects in ice layer monitoring and early warning, improving the accuracy of ice layer stability monitoring and the effectiveness of early warning, and providing important guarantees for personnel safety and scientific decision-making.

[0097] Fault analysis module:

[0098] A fault analysis strategy was implemented. After generating an ice thickness map using LiDAR data, H discontinuous layers were identified. When H = 1, this discontinuous layer was used as the target fracture layer. When H ≠ 1, the cross-sectional areas of the different discontinuous layers were continuously monitored, and the one with the largest cross-sectional area was selected as the target fracture layer. Using the target fracture layer as the segmentation contour, image processing software was used to obtain the cross-sectional area and lateral length of the fractured ice layer. The estimated volume of the fractured ice layer was calculated based on the error rate of the image processing software. A computational model was then built based on the pre-processed estimated volume, the volume of the corresponding glacial lake area, and the fixed warning frequency to generate an estimated dynamic warning frequency.

[0099] The process of generating the ice thickness map is as follows:

[0100] The target ice layer is scanned using a LiDAR system, collecting reflected LiDAR data containing the three-dimensional coordinates of the ice surface. The collected LiDAR data is preprocessed, including noise removal, filtering, and classification, to improve data quality. The preprocessed LiDAR data is then analyzed using TerraScan software to generate a digital elevation model (DEM) of the ice surface. This DEM is combined with known sea level or ground elevation data to calculate the ice thickness and create an ice thickness map.

[0101] After identifying H discontinuous layers, the process of obtaining the target fault layer is as follows:

[0102] On an ice thickness map, discontinuities appear as areas of sudden thickness changes. Ground penetrating radar (GPR) is used to analyze the thickness map to identify H discontinuities. For each identified discontinuity, edge detection techniques are used to determine its boundary on the ice thickness map when H ≠ 1. Once the boundaries of the discontinuities are determined, a GIS system is used to calculate the cross-sectional area of ​​each discontinuity. Specifically, the pixels or grid cells within the boundary are counted and converted into actual area units. All identified and quantified discontinuities are compared, and the one with the largest cross-sectional area is selected as the target fracture.

[0103] The process for calculating the estimated volume of the fractured ice layer is as follows:

[0104] Use image processing software, including ArcGIS or ENVI, to analyze the ice thickness map. Use the target fracture layer as the edge and extract the outline of the ice to be dropped along the target fracture layer. Ignore the cross-sectional area of ​​the target fracture layer and calculate the area enclosed by the outline, i.e., the cross-sectional area of ​​the ice to be dropped. On the ice thickness map, use the measurement tools in the image processing software to measure the length of the target fracture layer along its lateral direction, i.e., the transverse length. The formula used to calculate the estimated volume of the fractured ice layer is as follows:

[0105] V drop =(A section +δ A *A section )*(L horizontal +δ L *L horizontal );

[0106] Where V drop represents the estimated volume of the fractured ice layer, A section Indicates the cross-sectional area of ​​the ice layer to be dropped, L horizontal represents the lateral length of the target fault layer, δ A The error rate of the cross-sectional area obtained by image analysis, that is, the error rate of the image processing software itself, δ L The error rate of the measured lateral length is the error rate of the measuring tool itself. Taking the error rate into account, the estimated maximum volume of the fractured ice layer is obtained, which is the estimated volume of the fractured ice layer.

[0107] It should be noted that Figure 2 The L in the mark indicates the horizontal length, that is, the corresponding L horizontal ;

[0108] The estimated volume, the volume of the corresponding glacial lake area, and the fixed warning frequency are preprocessed as dimensionless data to unify the units of each parameter for ease of subsequent calculation and processing.

[0109] The process of generating dynamic warning frequency estimates is as follows:

[0110]

[0111] R=V drop / V la ;(2)

[0112] In formula (1), f adj represents the estimated value of dynamic warning frequency, f bas represents the fixed warning frequency, that is, the fixed basic frequency of regular warning (such as warning once every 10 days), α represents the adjustment factor, which is used to control the amplitude of the frequency increase, and the value of α is 0-1, R represents the ratio of the estimated volume of the broken ice layer to the volume of the corresponding glacial lake area, and R0 represents the set risk threshold. According to historical burst data, under normal circumstances, when the ice layer volume accounts for 10% of the glacial lake volume, it means that the possibility of glacial lake burst is very high, so the set value of the risk threshold is usually 0.1; In formula (2), V la represents the volume of the corresponding glacial lake area;

[0113] It should be noted that max(0, R-R0) ensures that when R ≤ R0, the frequency does not increase (i.e., maintains a fixed base frequency), while when R > R0, the frequency decreases based on the proportion exceeding the risk threshold. The frequency adjustment logic is: when R increases, that is, the ratio of ice volume to glacial lake volume increases, indicating an increase in risk, so the estimated dynamic warning frequency should be reduced, which means that the warning interval should be shortened to improve the intensity and timeliness of monitoring.

[0114] Assumptions:

[0115] V drop =10000m 3 (estimated volume of fractured ice);

[0116] V la =100000m 3 (corresponding to the volume of the glacial lake);

[0117] f bas = 10 days (fixed basic frequency of regular warning);

[0118] R0 = 0.1 (risk threshold, i.e., ice volume accounts for 10% of the glacial lake volume);

[0119] α = 1 (adjustment factor);

[0120] Calculated:

[0121] R=0.1, since R=R0, calculate f according to the formula adj Still the basic frequency; but if V dropIncreased to 20,000m 3 , then: R = 0.2, calculated: f adj =10 / 1.1≈9.09 days. In practical applications, the frequency can be rounded down to the nearest integer day, that is, f adj ≈9 days; in this way, we have correctly constructed a calculation model that dynamically adjusts the warning frequency based on the ratio of ice volume to glacial lake volume, and provided specific formulas and practical application examples. When the ice volume ratio increases, the warning frequency will also increase accordingly.

[0122] Early warning adjustment module:

[0123] Adjust the frequency of issuing the second-level warning signal according to the dynamic warning frequency estimate, and implement the regular discharge strategy according to the dynamic warning frequency estimate, and perform regular gate opening operations at the discharge outlet of the corresponding ice lake; slow flow pits can be arranged in advance downstream of the discharge outlet to avoid sudden changes in the downstream water level when the gate is opened regularly to discharge flood water.

[0124] Specifically, by generating ice thickness maps using LiDAR data and combining them with ground-penetrating radar (GPR) analysis, discontinuities can be accurately identified, leading to the determination of target fractures. Using these fractures as dividing lines, the volume of the fractured ice layer can be estimated. Based on this estimated volume and the corresponding glacial lake volume, the warning frequency can be dynamically adjusted. Implementing a regular discharge strategy based on the dynamic warning frequency estimate helps reduce the risk of glacial lake outbursts while also improving the effectiveness and timeliness of warnings.

[0125] Precise ice stability analysis and dynamic warning frequency adjustment enable more accurate assessment of ice burst risks, improving the safety of personnel and equipment. Scientifically implemented discharge strategies also provide clearer guidance for decision-makers, improving the efficiency of responding to ice burst events. This technical solution considers factors such as the volume of different glacial lake areas, the estimated volume of fractured ice, and a fixed warning frequency, making the warning system more flexible and adaptable to different situations.

[0126] In summary, this technical solution has achieved significant progress and beneficial results in ice stability monitoring, warning frequency adjustment, and discharge strategy implementation. It has improved the accuracy and timeliness of ice burst risk warnings, provided strong protection for the safety of personnel and equipment, and also provided important support for scientific decision-making.

[0127] Example 2:

[0128] See also Figure 3 Based on Example 1, this embodiment further provides a glacial lake outburst monitoring and early warning method for high-altitude areas, including the following specific steps:

[0129] S1. Select a glacial lake area and use satellite remote sensing technology and sensor equipment to collect image data at the junction of the glacial lake and the ice layer. Then process and analyze the image data to determine the estimated area S of the ice loss area and its changing trend;

[0130] S2. Compare the estimated area S of the current ice loss region with the set warning area M. If the estimated area S of the ice loss region exceeds the warning area M, trigger the warning judgment mechanism, extract the change trend of the estimated area S of the ice loss region, and determine whether the current ice change rate reaches the preset warning value. If so, issue a first-level warning signal and trigger the fault analysis strategy; if not, do not take any response action;

[0131] S3. After generating an ice thickness map using LiDAR data, H discontinuous layers are identified. When H = 1, the discontinuous layer is used as the target fracture layer. When H ≠ 1, the cross-sectional areas of the different discontinuous layers are continuously monitored, and the layer with the largest cross-sectional area is selected as the target fracture layer. Using the target fracture layer as the segmentation contour, image processing software is used to obtain the cross-sectional area and lateral length of the fractured ice layer. Combined with the error rate of the image processing software, the estimated volume of the fractured ice layer is calculated. The pre-processed estimated volume, the volume of the corresponding glacial lake area, and the fixed warning frequency are collected and used to build a computational model to generate an estimated value for the dynamic warning frequency.

[0132] S4. Adjust the frequency of issuing the current second-level warning signal based on the dynamic warning frequency estimate, and implement a regular discharge strategy based on the dynamic warning frequency estimate, and perform regular gate opening operations at the discharge outlet of the corresponding glacial lake.

[0133] In the application, the several formulas involved are all calculated by taking their numerical values ​​after removing the dimensions, and the formulas are established by collecting a large amount of data and performing software simulation to obtain a formula for the most recent real situation. Some coefficients or weights in the formulas are set by technical personnel in this field according to actual conditions, so they will not be elaborated here.

[0134] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art will appreciate that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution.

[0135] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, and may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment as needed.

[0136] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

Claims

1. A glacial lake outburst monitoring and early warning system for high altitude areas, characterized in that: include: The ice layer determination module selects an ice lake area, uses satellite remote sensing technology and sensor equipment to collect image data at the junction of the ice lake and the ice layer, and processes and analyzes the image data to determine the estimated area S of the ice layer loss area and its change trend; The preliminary warning module compares the estimated area S of the current ice loss area with the set warning area M. When the estimated area S of the ice loss area exceeds the warning area M, the warning judgment mechanism is triggered to extract the change trend of the estimated area S of the ice loss area and determine whether the current ice change speed reaches the preset warning value. If so, a first-level warning signal is issued and the fault analysis strategy is triggered. If it is not reached, no response action will be taken; The fault analysis module executes the fault analysis strategy and generates an ice thickness map using LiDAR data, and then identifies H discontinuous layers; when H=1, the discontinuous layer is used as the target fault layer; When H≠1, the cross-sectional areas of different discontinuous layers continue to be monitored, and the one with the largest cross-sectional area is extracted as the target fault layer; The target fracture layer is used as the segmentation contour, and the image processing software is used to obtain the cross-sectional area of ​​the fractured ice layer and the lateral length of the target fracture layer. Combined with the error rate of the image processing software, the estimated volume of the fractured ice layer is calculated. The calculation model is built based on the pre-processed estimated volume, the volume of the corresponding ice lake area, and the fixed warning frequency to generate a dynamic warning frequency estimate. The early warning adjustment module adjusts the frequency of the current second-level early warning signal according to the dynamic early warning frequency estimate, and implements the regular discharge strategy according to the dynamic early warning frequency estimate, and performs regular gate opening operations at the discharge outlet of the corresponding ice lake.

2. The glacial lake outburst monitoring and early warning system for high altitude areas according to claim 1 is characterized by: Sensor equipment includes unmanned submersibles and remotely operated vehicles that build cameras; The process of processing and analyzing image data is as follows: The collected image data are processed using image classification algorithms to distinguish ice and non-ice areas; the cross-sectional distribution map of the ice is obtained with the help of geographic information system software, and the three endpoints of the ice loss area are determined in the cross-sectional distribution map with the help of buoys that fit the edge of the ice; the estimated area S of the ice loss area is calculated using an algorithm based on peak and trough detection; The estimated area S is combined with the time series, and statistical analysis tools are used to analyze the changing trend of the estimated area S of the ice loss area.

3. The glacial lake outburst monitoring and early warning system for high altitude areas according to claim 2 is characterized by: When determining the three endpoints of the ice loss area, the buoy that is in contact with the edge of the ice and floating on the ice lake is point A, the position of the buoy perpendicular to the ice lake bed is point B, and the intersection of the ice lake bed and the ice layer is point C. The area enclosed by points A, B, and C is the ice loss area.

4. The glacial lake outburst monitoring and early warning system for high altitude areas according to claim 3 is characterized by: Under the condition that the edge of the ice presents a wave curve, the steps for running the peak and trough detection algorithm are as follows: Peak and trough detection: traverse the contour line corresponding to the edge of the ice layer between point A and point C, record the coordinates of each point, calculate the curvature of each point to identify the peaks and troughs, and use the detected peaks and troughs as segmentation points; Adjust the number of selection points between segmentation points: obtain the straight-line distance between adjacent segmentation points, build a rule engine, set a fixed threshold, use the fixed threshold as the interval between adjacent segmentation points, select a corresponding number of selection points, and evenly set the selection points in the segment formed between each adjacent segmentation point; Calculation of estimated area of ​​ice loss area: a small area is formed between any two adjacent points between point A and point C. The small area is a trapezoidal or triangular area. The corresponding area calculation formula is used. The area of ​​each small area = 1 / 2*(upper base + lower base)*height. The areas of all small areas are added up to obtain the estimated area S of the ice loss area.

5. The glacial lake outburst monitoring and early warning system for high altitude areas according to claim 1 is characterized by: The process of determining whether the current ice layer change rate has reached the preset warning value is as follows: Collect time series data: record a series of time points t1, t2, ..., t n and the corresponding estimated ice loss areas S1, S2, ..., S n ; Use Python to perform linear regression analysis and build a linear regression model to obtain the value of β1; The linear regression model is expressed as follows: S t =β0+β1*t+∈ t ; In the formula, S t represents the estimated area of ​​ice loss at time point t, β0 represents the intercept term, β1 represents the slope term, ∈ t represents the error term; The linear regression model is used to calculate the value of β1, which is the change rate of the estimated area S of the ice loss region over time, that is, the current ice change rate.

6. The glacial lake outburst monitoring and early warning system for high altitude areas according to claim 1 is characterized by: The process of generating an ice thickness map is as follows: The target ice layer is scanned using the LiDAR system to collect reflected LiDAR data containing the three-dimensional coordinate information of the ice surface. Noise points are removed and filtered from the collected LiDAR data. The LiDAR data is solved using TerraScan software to generate a digital elevation model of the ice surface. The known sea level data and the generated ice surface DEM are combined to calculate the thickness of the ice layer and generate an ice thickness map.

7. The glacial lake outburst monitoring and early warning system for high altitude areas according to claim 6 is characterized by: After identifying H discontinuous layers, the process of obtaining the target fault layer is as follows: On the ice thickness map, the discontinuous layer appears as an area with a sudden change in thickness. The ice thickness map is analyzed by ground penetrating radar to identify H discontinuous layers. When H≠1, for each identified discontinuous layer, its boundary on the ice thickness map is determined by edge detection technology. After determining the boundary of the discontinuous layer, the GIS system is used to calculate the cross-sectional area of ​​each discontinuous layer. All discontinuous layers with different cross-sectional areas are compared, and the one with the largest cross-sectional area is extracted as the target fracture layer.

8. The glacial lake outburst monitoring and early warning system for high altitude areas according to claim 7 is characterized by: The process for calculating the estimated volume of the fractured ice layer is as follows: The ice thickness map was analyzed using image processing software. The target fracture layer was taken as the edge, and the outline of the ice layer to be fallen was extracted along the target fracture layer. The area enclosed by the outline, i.e., the cross-sectional area of ​​the ice layer to be fallen, was calculated. On the ice thickness map, the length of the target fracture layer was measured along the lateral direction using the measurement tool in the image processing software, i.e., the lateral length. The formula used to calculate the estimated volume of the fractured ice layer is as follows: V drop =(A section +d A *A section )*(L horizontal +d L *L horizontal ); Where V drop represents the estimated volume of the fractured ice layer, A section represents the cross-sectional area of ​​the ice layer to be dropped, L horizontal represents the lateral length of the target fault layer, δ A represents the error rate of the cross-sectional area obtained by image analysis, δ L Indicates the error rate of the measured horizontal length.

9. The glacial lake outburst monitoring and early warning system for high altitude areas according to claim 8 is characterized by: The process of generating dynamic warning frequency estimates is as follows: In formula (1), f adj represents the estimated value of dynamic warning frequency, f bas represents the fixed warning frequency, α represents the adjustment factor used to control the amplitude of the frequency increase, R represents the ratio of the estimated volume of the broken ice layer to the volume of the corresponding ice lake area, and R0 represents the set risk threshold; in formula (2), V la Indicates the volume of the corresponding glacial lake area.

10. A method for monitoring and warning of glacial lake outburst in high altitude areas, using any system described in claims 1 to 9, characterized in that: The steps include: S1. Select a glacial lake area, use satellite remote sensing technology and sensor equipment to collect image data of the junction between the glacial lake and the ice layer, and process and analyze the image data to determine the estimated area S of the ice loss area and its change trend; S2. Compare the estimated area S of the current ice loss area with the set warning area M. If the estimated area S of the ice loss area exceeds the warning area M, trigger the early warning judgment mechanism, extract the change trend of the estimated area S of the ice loss area, and determine whether the current ice change speed reaches the preset warning value. If so, issue a first-level early warning signal and trigger the fault analysis strategy; If it is not reached, no response action will be taken; S3. After generating an ice thickness map using LiDAR data, H discontinuous layers are identified; when H=1, the discontinuous layer is used as the target fracture layer; When H≠1, the cross-sectional areas of different discontinuous layers continue to be monitored, and the one with the largest cross-sectional area is extracted as the target fault layer; The target fracture layer is used as the segmentation contour, and the image processing software is used to obtain the cross-sectional area of ​​the fractured ice layer and the lateral length of the target fracture layer. Combined with the error rate of the image processing software, the estimated volume of the fractured ice layer is calculated. The calculation model is built based on the pre-processed estimated volume, the volume of the corresponding ice lake area, and the fixed warning frequency to generate a dynamic warning frequency estimate. S4. Adjust the frequency of issuing the second-level warning signal according to the dynamic warning frequency estimate, and implement the regular discharge strategy according to the dynamic warning frequency estimate, and perform regular gate opening operations at the discharge outlet of the corresponding ice lake.