PDF(1798 KB)
基于多源遥感与时序异常特征的冰湖溃决灾害预警研究——以吉隆口岸为例
崔长露, 陈喆, 向大享, 刘淑凤, 姜莹
长江科学院院报 ›› 2026, Vol. 43 ›› Issue (10) : 87-95.
PDF(1798 KB)
PDF(1798 KB)
基于多源遥感与时序异常特征的冰湖溃决灾害预警研究——以吉隆口岸为例
Early Warning of Glacial Lake Outburst Floods Based on Multi-source Remote Sensing and Temporal Sudden Anomaly Characteristics: A Case Study of the Gyirong Port
冰湖溃决洪水是高寒山区极具破坏性的自然灾害。以2025年7月8日西藏吉隆口岸冰湖溃决事件为对象,利用Sentinel-2和近10项水文气象指标,采用Spearman秩相关、Z-score异常检测、时滞交叉相关等方法,识别冰湖溃决的影响因素与灾前异常信号。结果发现:① 冰湖面积与2 m气温(秩相关系数rs=0.55)、总蒸发量(rs=0.58)表现出较强的正相关,与雪当量(rs=-0.35)呈负相关;② 降雨与土壤水分对冰湖面积存在1~2期延迟水量响应,滞后2期相关系数均为0.75(显著性概率P=0.05),气温与总蒸发量表现为即时能量响应;③ 溃决前夕径流与面积呈负相关(rs=-0.57),2 m气温、总蒸发量、土壤水分在溃决前最近一期有效观测(距溃决事件3 d)Z-score绝对值均>1,可作为冰湖溃决短临预警的参考指标。研究实现了数据稀缺区域冰湖溃决的影响因素分析,可为高寒边境区域冰湖灾害应急响应与早期预警提供技术参考。
[Objective] Glacial lake outburst floods (GLOFs) pose severe threats to infrastructure and human lives in high-altitude Himalayan border regions. Existing GLOF studies mostly focus on long-term evolution and static risk assessment of large, steadily-observed glacial lakes, relying heavily on continuous ground-station measurements. There remain prominent research gaps in identifying pre-disaster anomaly signals for small, episodic glacial lakes under data-scarce conditions, especially when cloud contamination limits the availability of valid remote-sensing observations. [Methods] This study focuses on the glacial lake outburst event that occurred at the Gyirong Port in Tibet on July 8, 2025. The Sentinel-2 MSI 10-m optical imagery was adopted to extract multi-temporal glacial-lake boundaries and area sequences using the Normalized Difference Water Index (NDWI, threshold=0.2) combined with manual visual correction. ERA5-Land re-analysis products supplied regional-scale hydrometeorological variables including 2-m air temperature, precipitation, total evaporation, surface soil water, runoff, snow-water equivalent and snowmelt. Spearman rank correlation was applied to quantify monotonic relationships between lake area and hydrometeorological factors for small, non-normally-distributed samples. Z-score standardization was implemented to quantify relative anomalies of each indicator before outburst. Time-lagged cross-correlation with lag-0, lag-1 and lag-2 observation windows was further performed to distinguish immediate energy-driven responses versus delayed water-driven responses. [Results] The glacial lake experienced four distinct evolutionary phases: slow growth, rapid expansion, abrupt area reduction upon outburst, and post-outburst steady decay. Spearman correlation outputs showed that glacial-lake area had moderate positive correlations with 2-m air temperature (rs=0.55) and total evaporation (rs=0.58), and negative correlations with snow-water equivalent (rs=-0.35). Precipitation and surface soil water exhibited obvious delayed hydrological effects; their correlation coefficients with lake area both reached 0.75 at lag-2 (P=0.05), demonstrating a 1-2-phase delayed water-supply response. By contrast, 2-m temperature and total evaporation generated immediate responses at lag-0 without time delay. Approximately three days before the outburst, the Z-scores of 2-m air temperature, total evaporation, and surface soil water reached 1.45, 1.40, and -1.04, respectively, with all absolute values exceeding 1. [Conclusions] The results indicate that the studied glacial lake at Gyirong Port was controlled by two complementary processes: an immediate energy-driven response and a delayed water-supply response. Elevated air temperature and enhanced evaporation represented rapid energy forcing that promoted snow and ice melting and increased water availability, whereas precipitation and soil-water responses were delayed by approximately 1-2 observation periods, corresponding to a lagged water-transfer process from the watershed to the lake. The concurrence of sustained positive energy anomalies, anomalous soil-water conditions, rapid lake expansion, and reduced runoff suggests that enhanced meltwater accumulation and temporary hydrological retention jointly increased lake storage and may have reduced ice-dam stability before the outburst. The anomalous signals detected approximately three days before failure demonstrate that Z-score analysis can provide useful event-scale indicators for short-term GLOF risk identification. This study achieved a process-based attribution analysis of glacial lake outburst floods in areas with small sample sizes and sparse data, providing technical support for emergency response and early warning of glacial lake disasters in high-altitude border regions.
冰湖溃决 / 多源遥感 / Spearman秩相关 / 时序异常 / 吉隆口岸
Glacial lake outburst flood / multi-source remote sensing / Spearman rank correlation / temporal sudden anomaly / Gyirong Port
| [1] |
|
| [2] |
|
| [3] |
邹强, 周斌, 杨涛, 等. 喜马拉雅高海拔山区冰湖溃决时空分异特征[J]. 地球科学, 2024, 49(11):4047-4062.
(
|
| [4] |
|
| [5] |
Glacial lake outburst floods (GLOFs) represent a major hazard and can result in significant loss of life. Globally, since 1990, the number and size of glacial lakes has grown rapidly along with downstream population, while socio-economic vulnerability has decreased. Nevertheless, contemporary exposure and vulnerability to GLOFs at the global scale has never been quantified. Here we show that 15 million people globally are exposed to impacts from potential GLOFs. Populations in High Mountains Asia (HMA) are the most exposed and on average live closest to glacial lakes with ~1 million people living within 10 km of a glacial lake. More than half of the globally exposed population are found in just four countries: India, Pakistan, Peru, and China. While HMA has the highest potential for GLOF impacts, we highlight the Andes as a region of concern, with similar potential for GLOF impacts to HMA but comparatively few published research studies.© 2023. The Author(s).
|
| [6] |
Glacial lake dynamics in high-mountain regions serve as a sensitive proxy for cryospheric responses to climate warming. This study utilizes multi-temporal Sentinel-2 imagery and digital elevation model (DEM) data to quantify glacial lake evolution in the Donglin Tsangpo Watershed, a strategically important section of the China–Nepal Economic Corridor, from 2016 to 2024. The results show a significant expansion in both the number (from 43 to 56) and total area (from 3.97 km2 to 4.94 km2, +24.43%) of glacial lakes, primarily driven by the rapid emergence of very small lakes (0.02–0.05 km2) and a clear upward shift in elevation distribution, with new lakes forming above 5300 m and extending to elevations exceeding 5500 m. Analysis of Moderate Resolution Imaging Spectroradiometer (MODIS) land surface temperature (LST) reveals that this expansion coincided with pronounced positive thermal anomalies, particularly the 2020 extreme warm event (daytime +3.88 °C, nighttime +1.61 °C). Mechanistic analysis using the ERA5-Land reanalysis dataset further demonstrates that persistent positive downward longwave radiation (LW) anomalies (peaking at +10.71 W/m2 in 2021) effectively compensated for reduced shortwave input, inhibiting nocturnal refreezing and extending the effective ablation period. Furthermore, a rising liquid-to-solid precipitation ratio and extreme melt-day anomalies (up to +39.36 days) provided intensified hydrothermal inputs, driving the pronounced expansion of glacier-contact lakes despite non-linear interannual responses. This study also estimates individual lake volumes, identifying a transition toward rapid lake development that elevates potential downstream hazard exposure. These findings provide a high-resolution dataset and a robust physical framework for transboundary environmental monitoring and risk assessment in this climate-sensitive region.
|
| [7] |
李时, 李志, 马轶, 等. 基于风险源与动力机制的冰湖溃决判识新方法:以炯拉措冰湖为例[J]. 科学技术与工程, 2025, 25(33):14129-14140.
(
|
| [8] |
常家豪, 姚晓军, 王玉, 等. 新藏公路沿线流域冰湖变化与冰湖溃决洪水危险性评估[J]. 湖泊科学, 2025, 37(4): 1457-1469.
(
|
| [9] |
张波, 张瑞, 刘国祥, 等. 基于SAR影像的贡巴冰川末端冰湖年际变化监测及溃决规律分析[J]. 武汉大学学报(信息科学版), 2019, 44(7):1054-1064.
(
|
| [10] |
|
| [11] |
安兴荣, 何坤, 梁毅豪, 等. 西藏喜马拉雅中段波曲流域冰川冰湖多源遥感监测与灾害风险评估[J/OL]. 成都理工大学学报(自然科学版),[2026-02-13]. Doi:10.12474/cdlgzrks.2026011001.
(
|
| [12] |
张太刚, 王伟财, 汪鑫, 等. 青藏高原腹地冰坝湖的演化与溃决[J]. 中国科学:地球科学, 2026, 56(2):572-583.
(
|
| [13] |
贾洋, 崔鹏. 西藏冰湖溃决灾害事件极端气候特征[J]. 气候变化研究进展, 2020, 16(4):395-404.
(
|
| [14] |
Glacial lake outburst flood (GLOF) is one of the major natural disasters in the Qinghai-Tibetan Plateau (QTP). On 25 June 2020, the outburst of the Jiwenco Glacial Lake (JGL) in the upper reaches of Nidu river in Jiari County of the QTP reached the downstream Niwu Township on 26 June, causing damage to many bridges, roads, houses, and other infrastructure, and disrupting telecommunications for several days. Based on radar and optical image data, the evolution of the JGL before and after the outburst was analyzed. The results showed that the area and storage capacity of the JGL were 0.58 square kilometers and 0.071 cubic kilometers, respectively, before the outburst (29 May), and only 0.26 square kilometers and 0.017 cubic kilometers remained after the outburst (27 July). The outburst reservoir capacity was as high as 5.4 million cubic meters. The main cause of the JGL outburst was the heavy precipitation process before outburst and the ice/snow/landslides entering the lake was the direct inducement. The outburst flood/debris flow disaster also led to many sections of the river and buildings in Niwu Township at high risk. Therefore, it is urgent to pay more attention to glacial lake outburst floods and other low-probability disasters, and early real-time engineering measures should be taken to minimize their potential impacts.
|
| [15] |
徐航, 陈宁生, 杨志全. 西藏冰湖溃决频次与径流峰值流量相关性分析[J]. 科学技术与工程, 2023, 23(9):3641-3647.
(
|
| [16] |
刘晶晶, 马春, 苏鹏程. 以气温和降雨量为指标的冰湖溃决预警方法[J]. 南水北调与水利科技, 2018, 16(6): 1-8.
(
|
| [17] |
The southeast region of Tibet experiences frequent glacial lake outburst disasters, and disaster warning systems are thus crucial for disaster prevention and mitigation in the area. In this study, based on remote sensing images and historical data, 20 glacial lakes in southeast Tibet were selected as samples for risk analysis. A probability model of glacial lake outburst floods (GLOFs) in southeast Tibet was established using logistic regression for seven selected prediction indexes. By calculating the sensitivity and specificity of the model, the probability of identifying GLOFs was found to be 60%, with an identification degree of 86%. The under the ROC (receiver operating characteristic) curve index was prominently larger than 0.5, indicating the applicability of logistic regression for predicting GLOFs in southeast Tibet. The probability equation of the model shows that the area of the glacial lake, the distance of the glacial lake from the glacier, the slope of the glacier, the slope of the glacier tongue, and the dam backwater slope have a great influence on the probability of GLOFs. The results can provide a reference for the local governments to prevent disasters and reduce the damage of GLOFs.
|
| [18] |
|
| [19] |
\n Weather and climate conditions drive the evolution of tropical glaciers which play an important role as water reservoirs for Peruvian inhabitants in the arid coast and semi-arid Andean region. The scarcity of long-term high-quality observations over Peruvian glaciers has motivated the extensive use of reanalysis data to describe the climatic evolution of these glaciers. However, the representativeness and uncertainties of these reanalysis products over these glaciers are still poorly constrained. This study evaluates the ability of the ERA5-Land reanalysis (ERA5L) to reproduce hourly and monthly 2 m air temperature and relative humidity (T2m and Rh2m, respectively) over several Peruvian glaciers. We compared the ERA5L with data from four on-glacier automatic weather stations (AWS), whose hourly time series were completed with nearby stations, for the period January 2017 to December 2019. Results indicates a better performance of the reanalysis for T2m (\n r\n >0.80) than for Rh2m (∼0.4<\n r\n <∼0.6) in all four glaciers. Concerning the observations, both parameters show a daily cycle influenced by the presence of the glacier. This influence is more prominent during the dry months when the so-called glacier damping and cooling effects are stronger. On a monthly time scale, the ERA5L validation for both parameters are better in wet outer tropical sites (RMSE between ±0.2°C for T2m and between 3%–7% for Rh2m) rather than in dry outer tropical sites (RMSE between ±0.2°C for T2m and between 3%–7% for Rh2m). Among all sites considered in the study, the Rh2m bias is the highest in the Cavalca glacier (correlation of 0.81; RMSE 13%, MAE 11% and bias 8.3%) and the lowest in Artesonraju glacier (correlation of 0.96; RMSE 3%; MAE 2.3% and bias — 0.8%). Based on certain considerations outlined in this paper, it is appropriate to use ERA5L to characterize T2m and Rh2m conditions on Peruvian glaciers, particularly in the wet outer tropics.\n
|
| [20] |
Spatio-temporal reconstruction of winter glacier mass balance is important for assessing long-term impacts of climate change. However, high-altitude regions significantly lack reliable observations, which is limiting the calibration of glaciological and hydrological models. Reanalysis products provide estimates of snow precipitation also for remote high-mountain regions, but this data come with inherent uncertainty, and assessing their biases is difficult given the low quantity and quality of available (long-term) in situ observations.\nIn this study, we aim at improving knowledge on the spatio-temporal variations in winter glacier mass balance by exploring the combination of data from reanalyses and direct snow accumulation observations on glaciers using machine learning. We use the winter mass balance data of 95 glaciers distributed over the European Alps, western Canada, Central Asia and Scandinavia and compare them with the total precipitation from the ERA5 and the MERRA-2 reanalysis products during the snow accumulation seasons from 1981 until 2019. We develop and apply a machine learning model to adjust the precipitation from the reanalysis products along the elevation profile of the glaciers and consequently to reconstruct the winter mass balance in both space (for glaciers without observational data) and time (filling observational data gaps). The employed machine learning model is a gradient boosting regressor (GBR), which combines several meteorological variables from the reanalyses (e.g. air temperature, relative humidity) with topographical parameters. These GBR-derived estimates are evaluated against the winter mass balance data using (i) independent glaciers (site-independent GBR) and (ii) independent accumulation seasons (season-independent GBR). Both approaches resulted in reduced biases and increased correlation between the precipitation of the original reanalyses and the winter mass balance data of the glaciers. Generally, the GBR models have also shown a good representation of the spatial (vertical elevation intervals) and temporal (years) variability of the winter mass balance on individual glaciers.
|
| [21] |
Regional-scale surface energy balance (SEB) models of glacier melt require forcing by coarse-gridded data from reanalysis or global climate models that need to be downscaled to glacier scale. As on-glacier meteorological observations are rare, it generally remains unknown how exact the reanalysis and downscaled data are for local-scale SEB modeling. We address this question by evaluating the performance of reanalysis from the European Centre for Medium-Range Weather Forecasts (ERA5 and ERA5-Land reanalysis), with and without downscaling, at four glaciers in western Canada with available on-glacier meteorological measurements collected over different summer seasons. We dynamically downscale ERA5 with the Weather Research and Forecasting (WRF) model at 3.3 and 1.1 km grid spacing. We find that our SEB model, forced separately with the observations and the two reanalyses, yields less than 10 % difference in simulated total melt energy and shows strong correlations (0.86) in simulated time series of daily melt energy at each site. The good performance of the reanalysis-derived melt energy is partly due to cancellation of biases between overestimated incoming shortwave radiation and substantially underestimated wind speed and subsequently turbulent heat fluxes. Downscaling with WRF improves the simulation of wind speed, while other meteorological variables show similar performance to ERA5 without downscaling. The choice of WRF physics parameterization schemes is shown to have a relatively large impact on the simulations of SEB components but a smaller impact on the modeled total melt energy. The results increase our confidence in dynamical downscaling with WRF for long-term glacier melt modeling in this region.
|
| [22] |
Himalayan glacier changes in the context of global climate change have attracted worldwide attention due to their profound cryo-hydrological ramifications. However, an integrated understanding of the debris-free and debris-covered glacier evolution and its interaction with glacial lake is still lacking. Using one case study in the Gyirong River Basin located in the central Himalayas, this paper applied archival Landsat imagery and an automated mapping method to understand how glaciers and glacial lakes interactively evolved between 1988 and 2015. Our analyses identified 467 glaciers in 1988, containing 435 debris-free and 32 debris-covered glaciers, with a total area of 614.09 ± 36.69 km2. These glaciers decreased by 16.45% in area from 1988 to 2015, with an accelerated retreat rate after 1994. Debris-free glaciers retreated faster than debris-covered glaciers. As a result of glacial downwasting, supraglacial debris coverage expanded upward by 17.79 km2 (24.44%). Concurrent with glacial retreat, glacial lakes increased in both number (+41) and area (+54.11%). Glacier-connected lakes likely accelerated the glacial retreat via thermal energy transmission and contributed to over 15% of the area loss in their connected glaciers. On the other hand, significant glacial retreats led to disconnections from their proglacial lakes, which appeared to stabilize the lake areas. Continuous expansions in the lakes connected with debris-covered glaciers, therefore, need additional attention due to their potential outbursts. In comparison with precipitation variation, temperature increase was the primary driver of such glacier and glacial lake changes. In addition, debris coverage, size, altitude, and connectivity with glacial lakes also affected the degree of glacial changes and resulted in the spatial heterogeneity of glacial wastage across the Gyirong River Basin.
|
| [23] |
There is currently no glacial lake inventory data set for the entire high-mountain Asia (HMA) area. The definition and classification of glacial lakes remain controversial, presenting certain obstacles to extensive utilization of glacial lake inventory data. This study integrated glacier inventory data and 668 Landsat TM, ETM+, and OLI images and adopted manual visual interpretation to extract glacial lake boundaries within a 10 km buffer from glacier extent using ArcGIS and ENVI software, normalized difference water index maps, and Google Earth images. The theoretical and methodological basis for all processing steps including glacial lake definition and classification, lake boundary delineation, and uncertainty assessment is discussed comprehensively in the paper. Moreover, detailed information regarding the coding, location, perimeter and area, area error, type, time phase, source image information, and subregions of the located lakes is presented. It was established that 27 205 and 30 121 glacial lakes (size 0.0054-6.46 km(2)) in HMA covered a combined area of 1806:47 +/- 2:11 and 2080:12 +/- 2:28 km(2) in 1990 and 2018, respectively. The data set is now available from the National Special Environment and Function of Observation and Research Stations Shared Service Platform (China): https://doi.org/10.12072/casnw.064.2019.db (Wang et al., 2019a).
|
| [24] |
|
| [25] |
|
| [26] |
|
| [27] |
Understanding the factors that influence catchment runoff response is essential for effective water resource management. Runoff response to precipitation can vary significantly, depending on the dynamics of hillslope water storage and release and on the transmission of hydrological signals through the channel network. Here, we use ensemble rainfall–runoff analysis (ERRA) to characterize the runoff response of 189 Iranian catchments with diverse landscapes and climates. ERRA quantifies the increase in lagged streamflow attributable to each unit of additional precipitation, while accounting for nonlinearities in catchment behavior. Peak runoff response, as quantified by ERRA across Iran, is higher in more humid climates, in steeper and smaller catchments, and in catchments with shallower water tables. The direction and approximate magnitude of these effects persist after correlations among the drivers (e.g., deeper water tables are more common in more arid regions) are accounted for. These findings highlight the importance of catchment attributes in shaping runoff behavior, particularly in arid and semi-arid regions, where climatic variability and groundwater dynamics are crucial factors in sustainable water resource management and effective flood risk mitigation.
|
/
| 〈 |
|
〉 |
