Introduction

Globally, proglacial lakes, including ice-marginal lakes, hold ~0.43 mm of sea-level equivalent1. In Greenland, there are more than 3300 ice-marginal lakes located at the fringe of the ice, where their flow is typically restricted or dammed by ice or moraines2. Over the past three decades, the number of ice-marginal lakes in Greenland has increased, while changes in their size remain less clear1,2,3. These observed changes have been suggested to be associated with changes in ice sheet surface melt3 and retreat of the ice margin2,4. As 10% of the Greenland Ice Sheet (GrIS) and 5% of the Peripheral Glaciers and Ice Caps (PGICs)4 have an ice-marginal lake boundary, they exert an important influence on ice dynamics. Ice-marginal lakes have been shown to accelerate glacier mass loss and terminus retreat through calving5,6 and ice velocities at the GrIS margin are roughly 25% higher for glaciers terminating in lakes compared to those terminating on land4.

The water outflow from ice-marginal lakes can vary from a constant discharge to sudden outburst floods termed jökulhlaups or glacial lake outburst floods (GLOFs). These rapid drainage events have drastic consequences, including alterations in fjord circulations7 and downstream geomorphology8, changes in local ice dynamics9,10 and bedrock displacements11, as well as notable societal impacts12. The societal impacts of GLOFs in Greenland are lesser than other regions (such as the Himalayas and the Swiss Alps) because of the sparse population density, with few settlements in close proximity to the Ice Sheet and/or surrounding ice caps13. Moreover, small-scale studies from Greenland have shown that the dynamics of GLOF events have evolved over time, resulting in changes in timing, frequency, release volume, and/or rerouting of the release water8,10,14,15. Despite recent comprehensive studies of ice-marginal lakes across Greenland utilizing optical and Synthetic Aperture Radar (SAR) satellite imagery, and Digital Elevation Models (DEMs)2,3, the recorded incidents of GLOFs remain low at 153, with only 25 GLOF locations documented16. This count is notably low when considering the substantial number of ice-marginal lakes in Greenland. Moreover, this greatly contrasts the number of reported GLOFs in other glaciated regions globally16,17,18 in large-scale studies, likely reflecting the differing relevance of GLOFs to societal infrastructure and, in some cases, preservation of life. This suggests a scarcity of documentation rather than a scarcity of GLOF events in Greenland, highlighting the need for further investigation and emphasizing that the phenomenon is largely understudied in this region. Typically, GLOFs in Greenland have been monitored at the individual level or within small regional areas, employing in situ observations19 and remote-sensing techniques14,20.

This study presents the first, comprehensive large-scale study of water level fluctuations in ice-marginal lakes across Greenland (Fig. 1). Distinguishing itself from previous large-scale studies relying predominantly on optical and SAR satellite imagery2,3,21, our approach utilizes satellite and airborne altimetry data acquired from 2003 to 2023 (“Data and methods”). We utilize pre-defined lake outlines22 to extract altimetry measurements of all ice-marginal lakes larger than 0.2 km2 (n = 1387). For each lake, we construct a reliable water level time series by applying a simple statistical outlier detection framework, and calculate the largest observed water level difference (dWL) during the observation period (“Data and methods”). All lakes with a dWL exceeding 4 m are manually inspected and categorized into one of three general groups: (i) lakes with GLOF behavior, here defined as an abrupt decrease in water level between two consecutive observations captured within a year, thereby identifying GLOFs as disruptions to the existing trend in water level. (ii) lakes without GLOF behavior but with an overall decrease in water level, and (iii) lakes without GLOF behavior but with an overall increase in water level (“Data and methods”). Existing definitions of GLOFs generally revolve around a sudden release of meltwater without any distinct quantitative characteristics regarding factors such as peak discharge, duration, or magnitude23. In this study, a precise determination of such quantitative parameters remains challenging due to variations in the spatiotemporal resolution of the altimetry data over the studied ice-marginal lakes (“Data and methods”). Finally, we examine observed trends and spatiotemporal patterns of GLOFs across Greenland and investigate potential drivers.

Fig. 1: Overview of ice-marginal lake categories.
figure 1

All lakes included are larger than 0.2 km2 (n = 1387). The symbols indicate the lake drainage category, while the color indicates the observed maximum water level change (dWL). Lake statistics are included for each region, as defined by Mouginot and Rignot45.

Full size image

Results

Out of the 1387 ice-marginal lakes included in the study (Fig. 1), more than 80% (n = 1152) had a minimum of two observations after the statistical filtering (Fig. S1), with an observation defined as the median water level of all altimetry measurements captured within one day. Our results show that 687 ice-marginal lakes have limited surface fluctuation between 0 m and 4 m (Fig. 1 and Fig. S1), indicating that these lakes likely have a consistent and continuous water outflow. Some of the lakes may experience larger water level fluctuations occurring in between observations; however, as ~82% have five or more observations (Fig. S1), we expect this number to be limited. Ice-marginal lakes with limited water level fluctuations are found all around the margin of the GrIS and the PGICs, with the highest relative concentration in the southwest (SW) (~72%) and southeast (SE) (~90%) sector. Conversely, the lowest concentration is observed in the central east (CE) sector (~47%), whereas the central west (CW), northwest (NW), northeast (NE), and north (NO) sectors have similar concentrations, ranging from 61 to 63% (Fig. 1).

Fig. 2: Overview of ice-marginal lakes with GLOF behavior (n = 326).
figure 2

The size of the symbols indicates the minimum drainage magnitude, defined as the difference between the pre- and post-drainage water level with a minimum of 3 m. The drainage magnitude is dependent on when the altimetry observations are obtained in relation to the time of the event. The color indicates the number of observed drainages at each lake.

Full size image

Our findings reveal that 465 ice-marginal lakes experience large fluctuations exceeding 4 m, with an average of 24 observations per lake. We detect 326 lakes exhibiting GLOF behavior (category I), corresponding to more than a quarter of all 1152 ice-marginal lakes with altimetry observations. We detect a total of 541 GLOF events, with 45% of the lakes draining more than once (Fig. 2). As our approach does not capture all occurring GLOFs, this represents a conservative minimum estimate (“Data and methods”). Nonetheless, we still find a notable increase of 301 lakes and 388 GLOF events when compared to existing databases documenting GLOFs in Greenland over the past century16. The highest number of GLOF events are detected in 2019 (n = 178), accounting for one-third of all events, including 75 events from lakes that exclusively drained in 2019 (Fig. 3). During the four years with complete ICESat-2 coverage (2019–2022), we identified a total of 510 events, among which 170 are one-off events from a single lake. Five lakes drained every year, while 29 lakes experienced three events. In addition, 81 lakes drained twice, with 55% of them draining every second year.

Fig. 3: Number of annual GLOFs from 2019 to 2022.
figure 3

The lakes are divided into groups of single event GLOF-lakes (lakes that drain only once) and multiple event GLOF-lakes (lakes that drain more than once).

Full size image

Lakes demonstrating GLOF behavior are observed across all regions in Greenland, and in many areas we find them located in close proximity to one another (Figs. 1 and 2). The NE, NO, and southwest (SW) sectors have the highest number of ice-marginal lakes with GLOF behavior with 101, 78, and 58 lakes, respectively. The highest relative concentration of 43% is observed in the CE sector. By calculating the difference between the pre- and post-GLOF water levels, we get an estimate of the minimum drainage magnitude of each GLOF event (Data and Method). All sectors, except the SE, contain lakes with a drainage magnitude larger than 50 m, with the largest absolute and relative concentration in the SW with 11 lakes corresponding to 19% of all lakes in this sector. In addition, the SW sector also has one of the highest concentrations of lakes with 25–50 m drainage magnitude along with the NE, both with 21%.

Ice-marginal lakes without GLOF behavior, but with a general decrease in water level during the observational period (category II), are mainly found in the northern sectors (NE, NW, and NO), whereas those with a general increase (category III) are located in the NE and partly the CE sector (Fig. 1). In addition, we observe a large number of lakes with an overall water level increase located close to Bredebræ and Storstrømmen in the NE sector (Fig. 1 NE zoom-in).

Discussion

Our analysis of water level time series for 1152 ice-marginal lakes reveals a diverse range of lake behaviors, along with large differences within lakes of the same category. This variability is influenced by several factors, such as lake size, shape and location, catchment area, runoff, damming glacier size, and thickness, and, notably, the density and timing of observations. The latter has proved hugely important for properly gauging the water level variations for a large quantity of lakes on a Greenland-wide scale.

Within the category of lakes exhibiting GLOF behavior, we find that those experiencing recurrent events tend to drain at decreasing water levels over time. While in some instances, this may be linked to the timing of the observations taken prior to the event, we observe a similar pattern for lakes with dense observation coverage, as exemplified by Iluliallup Tasia and Lake Isvand (Fig. 4). Furthermore, upon examining optical satellite images of the selected lakes, we also identify a reduction in the pre-GLOF lake area (Fig. 4). Similar patterns have been observed in studies of individual lakes in Greenland and attributed to a thinning of the damming glacier8,10,14. Understanding whether this constitutes a general trend across Greenland requires a longer and more consistent dataset, which will become available as more data is continuously collected. However, large-scale studies of GLOFs from other glaciated regions detect only a moderate correlation between magnitude and glacier thinning18. The thinning of the damming glacier also influences those lakes, which show no GLOFs and a general decrease in water level during the observational period (category II) (Fig. S7). This can be induced by the lowering elevation of a potential spillway. However, in cases where a substantial decline in water level is observed from the early to the more recent observations (Fig. S7), there is a potential risk of overlooking GLOFs that occur between observations.

Fig. 4: Time-series of water level and lake area changes for luliallup Tasia, Lake Isvand, and Lake Hullet.
figure 4

These lakes have previously been found to produce substantial GLOFs46. Errors bars indicate the standard deviation of all measurements.

Full size image

Lakes displaying a general increase in water level (category III) are likely influenced by the thickening or advance of the damming glacier (Fig. S8), potentially coupled with changes in runoff from the catchment area into the lake. Alternatively, the rise in water level could be due to GLOFs occurring before the acquisition of the first altimetry observations, indicating that we are only obtaining measurements during the subsequent filling period. This was confirmed using optical satellite images for a small cluster of lakes dammed by Budolfi Isstrøm (Fig. 5). Two of the lakes drained simultaneously in late August 2017, while the third drained in the spring or summer of 2018, all prior to the first ICESat-2 observation.

Fig. 5: Evolution of lakes dammed by Budolfi Isstrøm.
figure 5

a Time series of water level and lake area changes. Note the concurrent GLOF of Lake 29377 and 135994 in 2017. b Optical Sentinel-2 images showing the observed maximum and minimum lake area of each lake from 2017 to 2023. Due to ice and snow cover on the lakes, the minimum and maximum lake extent may be difficult to determine. The observed lake changes would not be detected using automated water detection frameworks applied to optical images, which highlights one of the primary advantages of using altimetry data. Note the general agreement between fluctuations in lake area and water level.

Full size image

Given the observed close proximity of various draining lakes (Figs. 1 and 2), simultaneous GLOFs may be relatively common and occur at multiple locations across Greenland. A similar case of simultaneous GLOFs is observed at A.B. Drachmann Glacier, the neighboring glacier of Budolfi Isstrøm, where two lakes drained simultaneously in 2022, following more than 3 years of water level increase (Fig. S2). The simultaneous occurrences of these GLOFs might be initiated by changes in the damming glacier, such as an acceleration in ice velocity leading to the opening of cavities23 or a decrease in (sub)glacial meltwater24, possibly induced by a sudden drop in air temperature8. Alternatively, it may be a cascading effect, where the sudden release of meltwater from one lake triggers an immediate response in others, as it has been observed for supraglacial lakes25. However, the mechanism responsible for the simultaneous GLOFs is believed to be highly localized, as none of the other lakes dammed by Budolfi Isstrøm or A.B. Drachmann Glacier drained during the same period.

Throughout the observational period, we find substantial annual variations in the number of observed GLOFs, with a notable peak in 2019 (Fig. 6). From 2008 to 2018, the GLOF count is limited by the availability of observations. However, following the introduction of ICESat-2 in October 2018, there is a great increase in both the annual number of observations and GLOFs (Fig. 6). Between 2019 and 2022, the number of observations remains almost constant, and thus cannot explain the substantial differences in the annual number of observed GLOFs during this period.

Fig. 6: Evolution of the annual lake level observations, GLOFs, and mean runoff from 2008 to 2023.
figure 6

Runoff is extracted from RACMO2.3p2 at 1 km resolution, with a conservative uncertainty estimate of 20%. The period 2019–2022 has complete ICESat-2 coverage and a nearly constant number of lake observations.

Full size image

The primary factor regulating the fill-up rate of ice-marginal lakes is the inflow of water from melt and precipitation reaching the lakes from the surrounding catchments. Model-derived runoff from the Regional Atmospheric Climate Model (RACMO2.3p2)26 from 2008 to 2022 (Figs. 6 and 7a), estimates large annual variations. Between 2019 and 2022, 2019 stands out as having the highest annual runoff, despite the conservative uncertainty estimate of 20% (“Data and methods”). The extreme runoff in 2019 coincides with the peak in the annual number of GLOFs, whereas from 2020 to 2023 both the number of GLOFs and the mean annual runoff are considerably lower (Fig. 6 and Fig. 7a). Moreover, 2019 also stood out with a warm and dry summer, a record-low albedo, and a high degree of melting on the ice sheet27, with an annual cumulative melt area substantially larger than that observed in 2020, 2021, and 2022 (Fig. S3).

Fig. 7: Variations in runoff and annual number of GLOFs.
figure 7

a annual changes in GLOFs and runoff 2019–2022. Runoff anomaly is relative to a 2000–2022 baseline. b shows the covariance between the annual number of GLOFs and annual mean runoff at a sectoral-level. For cross-sectoral comparison, we normalize the annual GLOFs and runoffs by calculating the ratio of the total amount across all years (2019–2022) within a sector. Thus, the sum of each sector amounts to one on both axis. Note that the SE region is not included in (B) due to the limited amount of annual GLOFs.

Full size image

Previous research3 has suggested increased runoff as a factor driving regional changes in total lake volume along the CW and SW margins from 1987 to 2010, but it has not yet been linked to large-scale variations in GLOFs. Theoretically, larger volumes of runoff accelerate the filling of ice-marginal lakes, causing them to reach their maximum threshold more rapidly, suggesting that the system responds almost instantaneously to changes. Moreover, the additional runoff received during extreme years could result in earlier timing of drainage. Determining whether the lakes experience earlier outbursts would require longer and more temporally consistent time-series data. However, recent large-scale studies of GLOF changes during the past decade from other glaciated regions report earlier outbursts18.

At a regional scale, we find large differences in the covariance between runoff and the annual number of GLOFs, showing both positive and negative trends (Fig. 7b). Despite the SW region having the highest runoff, receiving more than two times the amount of water during an average year (2020) compared with some of the northern regions during extreme melt years (2019) (Fig S5), the extreme runoff in the SW in 2019 did not result in more GLOFs. These observations, coupled with an overall R2 value of 0.37 for all regions combined (Fig. 7b), suggest that factors other than runoff contribute to the fluctuations in GLOF occurrences.

During the past few decades, the GrIS margin has undergone extensive thinning28,29. Similarly, we observe a continuous thinning of the glaciers damming the ice-marginal lakes (Fig. S4). As the ice dams progressively thin, less water is required to initiate drainages10,23, suggesting a potential increase in the annual number of GLOFs. This anticipated trend is not reflected in the Greenland-wide annual number of GLOFs from 2019 to 2022, though it may influence regional patterns, such as those observed in the SW region.

Given the increase in runoff and the gradual ice dam thinning in the past decades (Figs. S6 and S4) and their anticipated future acceleration30,31, we hypothesize that the number of annual GLOFs has grown, and will persist as an upward going trend. As runoff from the GrIS is currently 60% more variable from year-to-year compared to the recent three decades32, we would also expect large interannual variations, as observed in our results from 2019 to 2022. However, determining such trends requires consistent long-term results17, which will become possible in the coming years as more data is continuously collected.

Conclusion

This study presents the first comprehensive view of ice-marginal lake drainages in Greenland. By combining altimetry measurements from ATM, ICESat, and ICESat-2 with data on lake outlines, we determine water level changes of 1152 ice-marginal lakes larger than 0.2 km2 between 2003 and 2023. Compared to commonly used approaches relying on optical imagery, our method is largely independent of clouds, snow, ice, and lake water turbidity. Our findings show that while about 60% of these lakes experience minimal water level fluctuations, 326 lakes have experienced drainage events, leading to a total of 541 observed GLOFs from 2008 to 2022. While the temporal resolution of the altimetry observations does not allow us to measure the exact duration of all events, optical imagery validation gives high confidence (95%) that the mapped events are indeed GLOFs. We find a diverse range of GLOFs, characterized by variations in both the duration of their cyclic behavior and the magnitude of drainage events, with some lakes exhibiting water level changes exceeding 140 m. The majority of the GLOF events occurred during a period of complete ICESat-2 coverage from 2019 to 2022, with a notable peak of 178 GLOFs observed in 2019. This particular year was marked by extreme runoff and record-low albedo.

While our findings of GLOFs present a conservative estimate, they reveal a remarkable 1200% increase compared to existing historical databases, highlighting a substantial underreporting of GLOF events in Greenland. Despite its comprehensive spatial coverage, the temporal resolution of our GLOF dataset is not extensive enough to accurately determine causal relationships to potential drivers or whether the lakes experience earlier outbursts and lower drainage magnitudes. However, new trends or more consistent patterns will undoubtedly emerge as more data is continuously collected. Providing simple explanations for Greenland-wide patterns poses challenges given the inherent complexity of GLOF events and triggering mechanisms may differ among lakes based on their specific characteristics. Moreover, multiple triggering mechanisms have been identified for causing GLOFs within the same lake across different years8. Even though, posing a smaller societal risk in Greenland, GLOFs have severe consequences for downstream ecosystems, geomorphology, and influence ice dynamics, among other factors. Therefore, we advocate for further research into the mechanisms and consequences of these dramatic events, with our results serving as a valuable starting point for future investigations.

Data and methods

Lake outlines and identification of ice-marginal lakes

To accurately estimate changes in water levels of lakes using altimetry measurements, precise and contemporary lake outlines are essential. We use lake outlines from the recently released mapping of Greenland conducted by The Danish Agency for Data Supply and Infrastructure (SDFI)22. The dataset has a high accuracy and internal consistency and is based on high-resolution satellite images (0.5 × 0.5 m) captured between 2018 and 2022, during a time when more than 96% of the altimetry data used in our analysis is captured. The dataset contains more than 150,000 lakes, each assigned a unique lake ID, with 10,073 identified as ice-marginal. This identification is based on the lakes proximity to the GrIS and peripheral glaciers, with the criterion of being situated within a 100-m buffer from the corresponding SDFI ice mask. We remove all lakes with an area smaller than 0.2 km2, as smaller lakes have limited altimetry coverage and consequently contain a high fraction of potential inaccurate observations, leaving a total of 1387 ice-marginal lakes to be included in the study.

Altimetry data

Lake water levels are determined by combining altimetry measurements from four different datasets (Table 1). We used altimetry data products, such as ATL06 and GLAH06, as they are pre-processed, have high accuracy and precision33,34, and have previously been used for studying fluctuations in lake water levels10,14,35. All ICESat-2 data was downloaded from the NSIDC homepage36 between July 1 and July 10, 2023, with the most recent measurements captured on April 16, 2023 (Table 1). The ATL06 is chosen over the ATL13 dataset, which is specific for inland surface water as it has extremely sparse coverage compared to the ATL06 dataset. Contrary, the ATL03 photons dataset exhibits an exceptionally high density of observations, necessitating a substantial computational capacity, likely with minimal additional information.

Table 1 Overview of the types of altimetry data included
Full size table

Estimation of water level time series

All altimetry datasets are merged and spatially filtered using the lake outlines, resulting in a comprehensive dataset of 15,745,605 altimetry measurements spanning all 1387 ice-marginal lakes. We apply a simple outlier detection framework to remove erroneous measurements located inside the lake area and calculate lake-specific water level time series. The framework is designed to be simple while effectively handling the inherent variability present in our measurements and the diverse characteristics of ice-marginal lakes. We observe considerable variation in the number of measurements across lakes, with smaller lakes having fewer measurements and thus being more susceptible to outlier influence compared to larger lakes with more extensive data coverage. In addition, there is a noticeable contrast in measurement density between altimeter datasets, with ATM tracks typically comprising thousands of measurements compared to relatively fewer measurements in the ICESat data. Lastly, some lakes exhibit large fluctuations in water level over various periods, while others remain relatively stable throughout the observation period.

  1. (1)

    For each lake, we calculate the median elevation of each unique day (Observation median) based on all measurements acquired on that day. Subsequently, we determine the median elevation across all unique days (lake median WL) (Fig. 8) to prevent bias due to variations in the number of measurements across different days.

    Fig. 8: Filtering of erroneous observations using the outlier detection framework.
    figure 8

    a Overview of all altimetry lake measurements and the specific steps of the outlier detection framework used for filtering. b Zoom in observations with the new post-filtering observation medians. Errors bars indicate the STD of all observed surface measurements.

    Full size image
  2. (2)

    Next, we calculate the absolute difference from all Observation medians to the lake median WL and calculate the median absolute deviation (MAD) (Fig. 8). We then define an upper and lower threshold using Eqs. (1) and (2), which we use to identify potential outliers.

    $${Upper\; threshold}={lake\; median\; WL}+C* {MAD}$$
    (1)
    $${Lower\; threshold}={lake\; median\; WL}{{{{{\rm{\hbox{-}}}}}}}C* {MAD}$$
    (2)

    This detection process employs a conservative outlier detection approach with C set to 3. We opt for the MAD method as it offers greater robustness in handling large outliers observed in our dataset (Fig. 8).

  3. (3)

    If any Observation median falls outside of the upper or lower bounds, we mark them as potential outliesr. However, we do not filter the observations immediately, but perform additional tests to check if they are actual outliers:

  1. 4.

    If over 30% of measurements during a day are within the threshold, we remove measurements outside and recalculate a new Observation median (Fig. 8).

  2. 5.

    If over 70% fall outside, we test the validity of the measurements. We filter out all measurements if they fail to meet the following criteria: (i) total measurements >3, (ii) Observation STD < 20 m, and (iii) difference between Observation median and lake median WL < 100 m (Fig. 8).

  1. 6.

    Lastly, we perform a final check on all Observation medians varying more than 50 m from the lake median WL, to capture potentially undetected outliers. If the Observation median varies more than 10 m from both the previous and following observation (date) within a 200-day window, the Observation median is removed (Fig. 8). This is based on the assumption that observations obtained within close proximity in time are likely to be more similar.

Following the outlier removal, we recalculate the Observation median water level of all remaining lake measurements and subsequently generate time-series illustrating the variations in water level for each individual lake. Next, we calculate the largest observed water level difference (dWL) (max. water level – min. water level) of each lake to identify lakes with notable water level changes.

For some lakes, the method produces time-series with unusual and abrupt post-drainage water-level fluctuations, as exemplified by the sudden water level increase of ~15 m observed in 2011 at Lake Tininnilik (Fig. S9, orange square) following the 2010 drainage. These instances primarily occur following drainage events when the lake area is reduced. During lower water levels, we include numerous altimetry observations from outside the post drained lake area, i.e., from the surrounding bathymetry/bedrock. Consequently, observations from the same overflight often exhibit a much larger internal variance, whereas observations from different days may differ due to variations in the drained lake bathymetry. In some cases, refining the lake outline for a more precise spatial filtering of the observations using a negative buffer considerably improves the water level time series. Lakes showing a general decrease in water level, which typically coincides with a decrease in lake area, may also be influenced by spatial filtering issues. In addition, if the lake-terminating glacier front has undergone substantial retreat or advance during the observational period, there is a risk that measurements might include elevation observations of the ice rather than the lake water level. However, decreasing the lake size too drastically may end up excluding observations with potentially important information. Striking the right balance in selecting the most appropriate criterion for outlier detection is crucial. Being overly conservative in the statistical outlier removal may also result in the inadvertent exclusion of abrupt water level changes as outliers.

Determining lake characteristics

All lakes with a water level difference larger than 4 m were selected (n = 503) for subsequent analysis as we are interested in lakes with substantial fluctuations. Within this subset, we examine the changes in water level and correct undetected outliers by manually removing them or by minimizing the lake area used for spatial filtering. In rare instances, we refer to optical imagery to confirm whether specific altimetry observations are indeed outliers. Following this additional filtering, we are left with 465 lakes with a water level difference exceeding 4 m.

We classify these 465 lakes based on their water level changes and divide them into three general groups: (i) lakes with GLOF behavior, (ii) lakes without any apparent GLOF behavior but with an overall falling water level during the observational period, and (iii) comparable to (ii) but with an overall rise in water level during the time series. There may be overlaps between categories as lakes experiencing GLOFs might also exhibit rising or falling water levels. However, we assign each lake to only one category due to the relatively short observational time span of the majority of our observations.

For each of the lakes exhibiting GLOF behavior, we manually determine key parameters such as the number of drainages, drainage year, and the pre- and post-drainage water level. For some lakes, such as Lake Hullet (Fig. 4), limited observations make it challenging to discern seasonal variations and precisely estimate the magnitude and timing of GLOFs. Nevertheless, GLOFs occurring in 2020, 2021, and 2022 can be identified. Conversely, for lakes with more frequent observations, such as Iluliallup Tasia and Lake Isvand (Fig. 4), we can more accurately determine the seasonal evolution and GLOF characteristics. We define the magnitude of each GLOF event as the difference between the pre- and post-drainage water level. The drainage magnitude serves as a valuable indicator for assessing the scale of these events. However, as it depends on the timing of the observations it should be interpreted as a minimum magnitude. In addition, it does not provide a direct measurement of the volume of released water, as the metric is dependent on the lake’s bathymetry. Consequently, ice-marginal lakes with extensive surface areas and drainage magnitudes of less than 10 m can still discharge substantial volumes of water, while smaller lakes with larger drainage magnitudes may release less volume.

Impact of spatiotemporal resolution on lake characterization

An important factor for accurately assessing changes in ice-marginal lake water levels and determining lake characteristics is the spatiotemporal resolution of the altimetry observations. In comparison to the relatively sparse spatiotemporal resolution of ATM and ICESat observations, ICESat-2 has a precise repeat track of 91 days37. This translates to a theoretical minimum observation frequency of three months for all ice-marginal lakes measured by the altimeter. However, gaps in the ALT06 dataset can result in extended periods between consecutive observations, particularly notable for small lakes, with some receiving only one observation annually (Fig. 9, Lake 23502 and 23853). Conversely, for larger lakes covered by multiple tracks, observations may be obtained less than two weeks apart during specific periods (Fig. 9, Lake 30591, 29659, and 29179). Therefore, higher temporal resolution enables more precise detection of potential GLOFs, and an overall increase or decrease in water levels, compared to lakes with lower temporal coverage. This is evident for Lake 23502, 28904, and 23853 (Fig. 9) where only one or very few observations are indicative of the GLOFs. Had any of these observations not been captured, no GLOF would have been identified. Contrary, for Lake 30591, 29659, and 29179 (Fig. 9) multiple points could potentially be removed, while still enabling identification of the GLOF. Moreover, for lakes with few observations, the specific timing of these is extremely important for detecting changes. A notable gap in observations during melt season, is more likely to result in a missed GLOF than one during winter since lakes mostly fill up during summer months and are consequently more prone to draining. Moreover, the duration of the drainage and fill-up cycle may also influence our ability to detect GLOFs in lakes with sparse observations. Water level changes of lakes experiencing multi-year fill-up periods are more easily detected, compared to lakes with an annual GLOF cycle and a shorter fill-up period.

Fig. 9: Water level time series of Lake 23502, 28904, 23853, 30591, 29659, and 29179.
figure 9

All six ice-marginal lakes exhibit GLOF behavior while showing varying temporal observation densities.

Full size image

When comparing the generated water-level time series with findings from comprehensive, lake-specific studies8,10, it expectedly shows that our large-scale approach does not detect all drainage events, nor the specific timing of these (Fig. S9). This is attributed to GLOFs occurring during periods when altimetry observations are limited, such as before the launch of ICESat-2 (e.g., Fig. S9, Lake Tininnilik), and due to a lack of observations, particularly evident for smaller lakes (e.g., Fig. S9, Russel Glacier Ice-dammed lake). This emphasizes the trade-off between detailed site-specific investigations and analyzing a broader dataset spanning numerous lakes across Greenland. Moreover, it highlights that our results represent a conservative minimum estimate of GLOF events. While undetected GLOFs may have an impact on the regional and annual number of observed GLOFs, they ultimately strengthen the overarching conclusions of our study, underscoring that GLOFs have been largely underreported in Greenland.

Accuracy assessment of the water level time series

To validate our findings, we conduct a comparison between the water level fluctuations derived from altimetry data and observed changes in lake area in optical satellite images for 110 lakes that drain. As some of the 110 lakes drained more than once during the observation period our validation includes a total of 184 GLOF events. For each of these events, we manually examined optical satellite images to confirm whether a drainage in the altimetry-based water level coincided with alterations in the lake area. For validation we rely on optical images from multiple satellites such as Planet Scope, Sentinel-2, and Landsat 8-9, offering nearly daily coverage. In instances where the lake is obscured by clouds, exhibits a frozen surface layer, or experiences a GLOF event during winter, we search for unobstructed images of the lake captured closest in time before and after the supposed GLOF. Although, in rare cases, there may be several months between the images, our primary concern is whether the event occurred during the time indicated by our altimetry observations of GLOF behavior, rather than the specific timing of the event. Out of the 184 validated events, we successfully confirm 95% of these as being GLOFs. The temporal resolution of the altimetry observations limits our ability to determine the exact duration of the GLOF or whether the release of water occurs over a few hours, weeks, or, in the worst cases, several months (Fig. 9). However, the optical validation also highlights that these drainages events are indeed GLOFs, as water level changes are observed to occur rapidly within a few days. Moreover, at lakes with more detailed mapping of changes in the lake area (Figs. 4 and 5), we observed a high agreement with the observed fluctuations in water level and the timing of the GLOFs. Combined with the high accuracy of the altimeter measurements and the large percentage of successfully confirmed drainages, this provides us with a high level of confidence in our results. In, instances where we erroneously document a drainage event using our altimetry record, this can be attributed to various factors, such as erroneous altimetry observations, frontal fluctuations of the damming glaciers, or the presence of large floating icebergs in the lakes. As our validation method solely addresses cases in which we identify a GLOF based on altimeter measurements that did not occur, so-called false positives, we lack a precise measure of the number of undetected GLOFs. These false negatives mainly occur at lakes with low temporal observation resolution, as accurately identifying drainage events relies on the spatiotemporal resolution of our observations, as discussed in the section above.

Automated water detection frameworks applied to optical images, commonly used in studies of ice-marginal lakes2,3, rely heavily on cloud-free images with distinct water body reflectance. Thus, these approaches would likely not have been able to capture the observed changes in Fig. 5 as the lakes are often covered by ice or snow, which is not uncommon for ice-marginal lakes located in the northern regions. This limitation emphasizes one of the primary advantages of using altimetry data for large-scale, Greenland-wide studies of lake changes, as it is largely independent of clouds, snow, ice, and lake water turbidity.

Annual runoff

The mean annual runoff is determined at Greenland-wide and at regional scale using the Regional Atmospheric Climate Model (RACMO2.3p2) with a 1 km resolution26, which is statistically downscaled from the RACMO2.3p2 at 5.5 km. The model is forced by ERA-40 (1958–1978), ERA-Interim (1979–1989), and ERA5 (1990-2022) climate reanalysis data. The downscaled 1 km version more accurately models high runoff rates observed at the GrIS margins, and overall, meltwater runoff is generally well-reproduced in the RACMO2.3p2, however, with a small negative bias compared to observed meltwater discharge26. The RACMO2.3p2 has an average runoff bias of 5% between 1976 and 2016. However, Noël et al.26 use a conservative uncertainty of 20% for annual basin-wide estimates, as this reflects the bias during extreme runoff years. Following a similar approach, we also adopt a 20% uncertainty. Considering that we calculate average annual runoff values at a Greenland-wide and sectoral level (NO, NE, NW, CE, CW, SE, SW), encompassing multiple basins, a 20% uncertainty is considered extremely conservative.

Surface melt extent

The annual cumulative surface melt extent of the GrIS is extracted from the Greenland Melt Analysis Spreadsheet38 available via The National Snow and Ice Data Center. Daily data on melt area from 1 April 1979 to the present is derived at 25 km resolution using a combination of multiple microwave sensors: The Special Sensor Microwave Imager/Sounder (SSMIS), the Special Sensor Microwave/Imager (SSM/I), and the Scanning Multichannel Microwave Radiometer (SMMR).

Ice dam elevation

The changes in ice dam elevations are calculated using two altimetry-derived elevation datasets of the GrIS covering a combined period from 2011 to 2022. The dataset by Khan 202339 contains annual (April to April) elevation change rates from 2011 to 2020 at 1 km resolution. The PRODEM40 dataset contains annual summer (June 1st to September 30th) elevations of the ice sheet marginal zone between 2019 and 2022 at a 500 m resolution41. For all lakes exhibiting GLOF behavior, we identify the centroid point and extract the annual elevation value from the closest elevation/dH pixel, located 1500 m from the lake centroid to avoid erroneous observations from the frontal, fluctuating part of the ice margin. Finally, we calculate the median elevation change rate and the associated uncertainty for each year (Fig. S4).