Small glaciers (${ \lt }0.5\,\mathrm{km}^2$), glacierets (${ \lt }0.25\,\mathrm{km}^2$) and, in particular, very small glacierets (${ \lt }0.01\,\mathrm{km}^2$), despite being numerous in mountain environments, are underrepresented in scientific inquiry when assessing their response to climate change. We present new insights into the vanishing (no visible surface ice whilst underlain by bedrock or water) of 77 very small glacierets distributed in the Northern and Central Andes of Chile. We also analyse the presumable vanishing (no visible surface ice whilst underlain by regolith) of 244 additional very small glacierets, comprising a total dataset of 321 very small glacierets within the study area, equivalent to the loss of $5.69\times 10^6\,\mathrm{m}^3$ of water equivalent ice volume according to the 2022 Chilean Public Glacier Inventory. Our results show that 45.5% of the sample shrank from individually small glaciers at the beginning of the 21st century, whereas 53.0% of the sample vanished after being fragmented from larger glaciers in the same time span. The observed generalised reduction behaviour and vanishing results after extremely dry conditions at the end of the 2009–2022 Central Andes megadrought. We discuss our results in terms of the minimum area threshold for classifying very small glacierets, and whether their vanishing poses a hydrological impact.

Atmospheric warming is causing glaciers worldwide to retreat and disappear, with repercussions for nearby communities. Effects of glacier retreat have been especially consequential for mountain resorts such as Whistler Blackcomb, Canada, where nearly 60 years of summer glacier use by skiing has been terminated due to the ongoing demise of Whistler, Horstman and Blackcomb Glaciers. Over the past five decades, these three glaciers decreased in area by 95%, 75% and 60%, respectively. Using a degree-day model with downscaled CMIP6 data, we project that Whistler Glacier will disappear within the next decade, whereas Horstman and Blackcomb Glaciers will vanish in the next $20\pm10$ years and $50\pm20$ years, respectively. The large uncertainty in disappearance dates depends primarily on the applied ice-thickness model. Our work demonstrates that ski resorts that rely on glaciers will need to balance current ice-preservation strategies with adaptation in light of ongoing future climate change.

We summarize and reflect on the nearly 30 studies that form this Collection on Vanishing Glaciers as a contribution to the 2025 International Year of Glaciers’ Preservation. Some of the studies in this Collection are extensions of the stories told in the Global Glacier Casualty List, a collaboration between social and physical scientists. They document glaciers in various states of demise in regions across six continents, including Svalbard, Iceland, Norway, Germany, France, Italy, Spain, Austria, Switzerland, New Zealand, Indonesia, USA, Canada, Peru, Bolivia and Chile. Vanishing glaciers affect archaeology, biodiversity, water resources, tourism, culture and glaciology itself, as we grapple with more precise definitions of dwindling ice masses and the loss of long-term monitoring sites. Landscapes are being transformed by rapid deglaciation, serving as a call to action and introducing numerous challenges for those engaged in measuring, monitoring and documenting glacier change.

NASA’s ICESat-2 mission was launched in 2018, carrying a photon-counting laser altimeter, with a primary objective of measuring height changes across Earth’s surface. ICESat-2 has provided measurements of ice surface height between 88º N and S, repeated four times per year, with high vertical accuracy and along-track spatial resolution. Its accuracy and coverage has enabled near-complete recovery of height changes across the ice sheets, capturing subtle changes in the interior, and rapid changes along the dynamic margins with steep slopes and the floating peripheral ice shelves. The ICESat-2 Science Team has developed a suite of algorithms that produce along-track and gridded land ice height products at various levels of processing, all freely available at the National Snow and Ice Data Center. Here, we describe three higher-level land-ice data products derived from ATL06 and their underlying algorithms: along-track height change (ATL11), digital elevation model (ATL14) and gridded surface height change (ATL15). We demonstrate the suitability of each data product for studying different ice sheet regions. We then show height changes for Greenland and Antarctica from ATL15 during the first 6 years of the ICESat-2 mission (October 2018–December 2024), illustrating how ICESat-2 measurements can distinguish the multi-year trends from seasonal fluctuations.

Disappearing glaciers are observed worldwide, but a reliable counting is only available for some regions. This is due to several issues, ranging from inconsistent glacier identification to different criteria applied to decide whether or not a glacier has disappeared. The public perception of a glacier being lost usually has a focus on specific, often well-known glaciers, with a related media attention. In contrast, hundreds of glaciers might disappear in other regions over the same period without any notice. When they are not widely known, this silent disappearance can also happen to scientifically valuable glaciers. For example, the loss of benchmark glaciers with decades of mass balance measurements is also a loss of important information about climate variability in remote high-mountain regions. This study gives an overview of the challenges and different criteria used to determine if a glacier has disappeared and presents recommendations for a proper counting and change assessment.

Land-to-lake glacier terminus transition can alter glacier dynamics and enhance mass loss. However, the magnitude and timing of dynamic changes during this transition remain poorly constrained, especially in hyperhumid regions such as western Patagonia. We analyze this process at four glaciers by acquired bathymetry data combined with multi-temporal remote-sensing observations to track lake development and glacier retreat. Although each glacier shows distinct patterns, three glaciers underwent marked periods of rapid retreat associated with the onset of buoyant conditions at the termini, while the fourth glacier appears to be entering this stage. Results indicate that during rapid retreat, subaqueous ice loss contributed to 8–40% of total mass loss, underscoring its relevance for glacier mass balance. Based on land-lake contact and dynamics behavior, we propose reclassifying Exploradores Glacier from land- to lake-terminating. The analysis is constrained by the limited spatial coverage of bathymetric surveys, particularly at Grosse and Exploradores Glaciers. Nevertheless, the findings underline the importance of improving bed models and buoyant condition analysis to anticipate glacier changes. Many Patagonian glaciers are expected to evolve into lake-terminating systems, making updated inventories and inclusion of subaqueous mass loss critical for hazard assessment and projections of future ice loss.

We assess ongoing regional glacier loss in the Austrian state of Vorarlberg using a set of manually mapped glacier outlines for 2017, 2020, 2022 and 2023. Vorarlberg has lost about half of its glacierized area since a previous inventory in the mid-2000s. In 2017–23, glacier area was lost at an average rate of 5% per year. Area loss rates at individual glaciers have increased over time but show considerable variability between glaciers and subperiods. Of 30 glaciers previously inventoried, 5 have vanished completely since 2017. We discuss mapping differences due to the variable interpretation of the images by multiple observers and the mapping challenges that arise even with very high-resolution (10 cm) imagery. Processes leading up to the complete loss of glacier ice, mainly increased debris cover and fragmentation into very small features, cause inherent uncertainties in documenting the disappearance of mountain glaciers and ice bodies. We considered criteria that might be used to define terminology and found 16 remaining glacier fragments with crevasses indicating past or current ice flow, which could be considered glaciers rather than ice bodies.

The Zugspitzplatt represents a key high-alpine reference landscape in German geography. Located below the Zugspitze, Germany’s highest peak, it still hosts some of the country’s small remaining glacier ice reserves. Owing to its long history of scientific observation, high accessibility, and pronounced sensitivity to climatic change, the Zugspitzplatt has acquired emblematic significance as a national reference site for alpine geomorphological and glaciological research. The observation of the Schneeferner glaciers follows a long-standing tradition of glacier monitoring in this region. As of 2022, one of the two glaciers has completely disappeared, and the remaining one is critically endangered. Here, we present geodetic mass balances since 1892 based on historical maps and own surveys. Measurements of ice thicknesses allow the determination of absolute ice masses and their change. From 52 megatons of ice in 1892, less than one megaton was left in 2023 and its disappearance is inevitable. While this deglaciation will have negligible effects on humankind, it represents a potent visual and emotional indicator of climate change. Given the presence of a touristic hotspot and an environmental research station, the Zugspitzplatt has the potential to serve as a focal point for climate education.

The Oregon Cascades had 35 named glaciers on seven volcanoes in the 1980s, with 34 of those glaciers remaining by 2000. Here, we document the glaciers that fall into the Global Glacier Casualty List categories based on five years of field observations of these 34 glaciers. Five glaciers have disappeared, four have almost disappeared and eight are critically endangered. Thus, half of the Oregon Cascades named glaciers have disappeared, almost disappeared, or reached critically endangered status in the 21st century. Between 1980 and 2024, the May–October ablation season of the Oregon Cascades region warmed at ∼0.3°C per decade, with a 2020–24 mean temperature ∼1.7°C warmer than the 1975–84 mean. In contrast, there was no significant trend in November–April accumulation season precipitation. Given the significant rise in melt-season temperature, we attribute ongoing glacier disappearance in the Oregon Cascades to the warming climate.

Ground-based time-lapse cameras are often used to monitor glacier recession, which is primarily driven by the falling of ice from the glacier front, known as iceberg calving or, more commonly, calving events. Glaciologists can utilise these images by manually identifying calving events, a laborious task that requires the analysis of thousands of images in order to identify image pairs that represent the glacier front before, during and after calving. We present a computer vision based method to filter out images rendered unusable due to weather effects such as fog and precipitation by calculating the number of salient key-points detected in the image using the SIFT (Scale-Invariant Feature Transform) algorithm as an indicator of the visibility of the glacier front and discarding any image with fewer key-points than a defined threshold. We propose the use of SNN (Siamese neural network) and show that it is useful in detecting calving events since it allows to separately calculate features on two images and then merges them together in order to track differences between them thus detecting calving areas. The trained model achieved an overall accuracy of 92%, with 79% of calving events and 93% of non-calving being correctly classified on an unseen test set formed from imagery in the same time period as the training data. The model was able to generalise to new time periods (and therefore small changes in viewpoint and alignment) to some extent with an overall accuracy of 82%, with 27% of calving events and 90% of non-calving being correctly classified.

Patagonia and Tierra del Fuego (Austral Andes) are the most glacierised regions in the Southern Hemisphere, where glaciers have experienced significant mass changes in recent decades. Understanding glacier–climate–water interactions is crucial for addressing future climate challenges. Open-access data play a key role in advancing geoscience research, improving models and assessing the impacts of hazards and sea-level rise impacts. Here, we present QFuego-Patagonia, a free glacier-related GIS dataset and web portal covering Patagonia and Tierra del Fuego, which provides essential geospatial information across four scientific topics: Glaciology, Atmosphere, Terrain Models, and Glacial Geology and Geomorphology. This initiative aims to foster interdisciplinary research and collaboration, synthesise current knowledge and establish an advanced glacier data repository that will be continuously updated as new data and insights become available.

Global Land Ice Measurements from Space (GLIMS), an initiative to build and distribute a database of global glacier data, has recently begun to track glaciers that have recently disappeared. GLIMS provides a definition of “extinct” glaciers for our community, and the final determination of extinction is left to local experts. There are currently 181 glaciers in the GLIMS Glacier Database that are marked as “extinct”, though we recognize that there have been many more reported in the literature. GLIMS welcomes more submissions to make the list more complete.

This study traces the evolution and the fate of two Icelandic glaciers, Hofsjökull eystri in the southeast and Okjökull in the west, from their Little Ice Age maximum (∼1890) to the present. Glacier maximum extents were reconstructed using historical maps, written descriptions, geomorphological evidence, and aerial and satellite imagery. Time series of digital elevation models and a 2025 radio echo sounding survey provide further insight into changes in surface area, volume and bedrock topography. Despite similar latitudes (∼64.4°N) and elevation ranges (∼900–1150 m a.s.l.), the glaciers reside in different climate regimes. Both glaciers covered ∼7 km2 in the 1940s. Since then, Okjökull has been declared vanished and Hofsjökull eystri has lost ∼70% of its area and ∼90% of its volume, with a maximum thickness of 55 m in 2024. Projections suggest that, at the current rate of thinning, Hofsjökull eystri will disappear entirely within the next 30–45 years.

Over the past two decades, the Canadian Arctic Archipelago has undergone significant glacier mass loss, driven primarily by surface melt. This study presents a detailed analysis of supraglacial drainage evolution along Ellesmere Island’s ∼830 km latitudinal extent using satellite imagery, historical aerial photographs and DEMs from 1959 to 2020. Analysis of five glaciers shows that drainage density (Dd) has increased over time, driven by the expansion of perennial rivers, especially at higher elevations. Far northern glaciers exhibit stable, well-developed drainage systems, while southern glaciers show a relatively greater increase in canyon development since 1959. Cold surface ice in the north supports higher Dd, while southern glaciers with extensive sinks (moulins and large crevasses) exhibit stronger surface-to-bed connectivity. Despite increased channelization, sinuosity changes remain statistically insignificant, reflecting dynamic canyon behavior governed by surface slope and meltwater discharge. Results align with modeled increases in melt, especially on southern glaciers where supraglacial systems have expanded most rapidly. Continued equilibrium line altitude rise under future warming is expected to intensify melt and result in the expansion of supraglacial drainage systems up-glacier, particularly for glaciers with large amounts of ice at mid-elevation.

Rock glaciers support alpine biodiversity and may respond more slowly to warming than snow or glaciers. While responses of snow and glaciers to climate change are relatively well understood, a robust assessment of rock glacier environmental niche, future distributions of rock glaciers and potential for development of rock glaciers from current glaciers is lacking. Using process-relevant, high-resolution environmental descriptors, we develop a species distribution model of the topographic, geologic and hydroclimatic niche of rock glaciers that provides novel estimates of potential rock glacier distributions for different climates. We identify mean annual air temperature and headwall area as the dominant controls on rock glacier spatial distributions, with rock glaciers more likely to be found in areas with mean annual temperatures close to −5°C, little rain, northern aspects and broad headwalls. While rock glacier climate equilibration may take hundreds of years, we find that equilibration to present climate will result in a 50% reduction in rock glacier habitat and equilibration to late 21st-century climate under a high-end warming scenario will result in a 99% reduction in rock glacier habitat across the western USA. Under future conditions, we find limited potential for glacier to rock glacier transformation (3% of glacierized area), concentrated in cold, high elevation, moderate precipitation areas.

Melting and calving of glaciers and ice caps in Antarctica and Greenland could potentially contribute significantly to global sea level rise. Updates to existing outlines that provide critical glacier baseline information in both regions could help in the analysis of particular changes in glacier parameters such as area and volume from time-series inventories. Here we synthesize previously established techniques and apply new multi-source datasets to update glacier outlines in selected test areas of Antarctica and Greenland, as well as to reduce uncertainties and errors during the mapping process. The workflow includes mapping glacier boundaries, subdividing glaciers by watersheds and assigning glacier attributes. Complicated glacier scenarios and updating challenges in polar regions are discussed and demonstrated by representative case studies. For the first time in Antarctica, we analyze the effect of terminus types on mapped glacier areas, and in Greenland we compare the differences with glacier mapping results using Landsat OLI and ETM+. With new data sources, the methods described in this study might help to create glacier outlines on a larger scale in Antarctica and Greenland. Although data sources can be substituted, the enormous amount of manual labor required to update glacier inventories remains a significant challenge.

We developed a multi-frequency, multi-Global Navigation Satellite System (GNSS) positioning instrument optimized for autonomous applications in the cryosphere. At lower power requirements and a fraction of the cost and weight compared to commercially available options, this instrument simplifies field usage and associated logistics. In this paper, we assess several baseline aspects of performance in a polar environment relative to geodetic receivers commonly used for glaciological applications. Evaluations of precision and relative accuracy of positioning show millimeter to centimeter-level (‘geodetic-grade’) quality of this instrument, making it a competitive alternative for GNSS glaciological and geophysical applications such as monitoring surface elevation change and ice flow. An array of these instruments, tested in the field on the Greenland Ice Sheet, also demonstrated robustness throughout the polar winter and met power and reliability requirements.

Innovations in geodesy enable widespread analysis of glacier surface elevation change and geodetic mass balance. However, coincident glacier area data are less widely available, causing inconsistent handling of glacier area change. Here we quantify the bias introduced into meters water equivalent (m w.e.) specific geodetic mass balance results when using a fixed, maximum glacier area, and illustrate the bias for five North American glaciers. Sites span latitudes from the northern U.S. Rocky Mountains (48°N) to the Central Alaska Range (63°N) between 1948 and 2021. Results show that fixed (maximum) area treatment subdues the m w.e. mass change signal, underestimating mass balance by up to 19% in our test cases. This bias scales with relative glacier area change and the mass balance magnitude. Thus, the bias for specific geodetic mass balances will be most pronounced across rapidly deglaciating regions. Our analysis underscores the need for temporally resolved glacier area in geodetic mass balance studies.

The Jostedalsbreen ice cap is mainland Europe's largest ice cap and accommodates 20% (458 km2 in 2019) of the total glacier area of mainland Norway. Jostedalsbreen and its meltwater contribute to global sea-level rise and to local water management, hydropower and tourism economies and livelihoods. In this study, we construct a digital terrain model (DTM) of the ice cap from 1966 aerial photographs, which by comparing to an airborne LiDAR DTM from 2020, we compute changes in surface elevation and geodetic mass balances. The area mapped in both surveys cover about 3/4 of the ice cap area and 49 of 82 glaciers. The measured glacier area has decreased from 363.4 km2 in 1966 to 332.9 km2 in 2019, i.e. a change of −30 km2 or −8.4% (−0.16% a−1), which is in line with the percentage reduction in area for Jostedalsbreen as a whole. The mean geodetic mass balance over the 49 glaciers was −0.15 ± 0.01 m w.e. a−1, however, large variability is evident between glaciers, e.g. Nigardsbreen (−0.05 m w.e. a−1), Austdalsbreen (−0.28 m w.e. a−1) and Tunsbergdalsbreen (−0.36 m w.e. a−1) confirming differences also found by the glaciological records for Nigardsbreen and Austdalsbreen.

Semantic segmentation is a critical part of observation-driven research in glaciology. Using remote sensing to quantify how features change (e.g. glacier termini, supraglacial lakes, icebergs, crevasses) is particularly important in polar regions, where glaciological features may be spatially small but reflect important shifts in boundary conditions. In this study, we assess the utility of the Segment Anything Model (SAM), released by Meta AI Research, for cryosphere research. SAM is a foundational AI model that generates segmentation masks without additional training data. This is highly beneficial in polar science because pre-existing training data rarely exist. Widely-used conventional deep learning models such as UNet require tens of thousands of training labels to perform effectively. We show that the Segment Anything Model performs well for different features (icebergs, glacier termini, supra-glacial lakes, crevasses), in different environmental settings (open water, mélange, and sea ice), with different sensors (Sentinel-1, Sentinel-2, Planet, timelapse photographs) and different spatial resolutions. Due to the performance, versatility, and cross-platform adaptability of SAM, we conclude that it is a powerful and robust model for cryosphere research.