Search

Glaciers in the tropical Andes of Peru and Bolivia are important but rapidly declining water resources. Precise knowledge of their extent is thus mandatory for calculation of their volume, area changes and mass balance. Due to wrongly mapped seasonal snow, the glacier outlines currently available from the widely used RGIv6 are often too large, resulting in errors for change assessment and volume estimation. Apart from snow cover, also frequent cloud cover and shadows cast by the steep terrain make glacier mapping in this region challenging.For this study, we have mapped all glaciers in Peru and Bolivia using cloud-free Landsat TM scenes from 1998 and Sentinel-2 scenes from 2020. In both years seasonal snow off glaciers was largely absent. Glacier extents were mapped with a standard band ratio (red/SWIR) and a scene specific threshold value. Wrongly classified lakes and missing debris cover were manually corrected, the latter also by using the very high-resolution satellite images available in the ESRI Basemap. The Copernicus DEM GLO-30 was used to derive new drainage divides and topographic information for each glacier.In total, we mapped 3586 glaciers larger than 0.01 km2 covering an area of 1747 km2 in 1998. This is 419 km2 or 20% less than the 2166 km2 in RGIv6. As glacier outlines in RGIv6 are from 2000 to 2009 and most area change studies found continuous and strong area decrease over this period, a ‘back-calculation’ of all glacier areas to the year 1998 with an annual shrinkage rate of 1% gives a mean area overestimation of 27% for 1998. This value varies regionally and is smaller in the Cordillera Blanca and much larger (>50%) in other regions. The related modelled glacier volumes for these regions are thus also overestimated. From 1998 to 2020 glaciers have lost 23% of their area (-1% per year).

Glaciers are crucial sources of freshwater in particular for the arid lowlands surrounding High Mountain Asia. In order to better constrain glacio-hydrological models, annual, or even better, seasonal information about glacier mass changes is highly beneficial. In this study, we test the suitability of very high-resolution Pleiades DEMs to measure glacier-wide mass balance at annual and seasonal scales in two regions of High Mountain Asia (Muztagh Ata in Eastern Pamir and parts of Western Nyainqêntanglha, South-central Tibetan Plateau), where recent estimates have shown contrasting glacier behavior. We find that the average annual mass balance in Muztagh Ata between 2020 and 2022 was -0.11 ±0.21 m w.e. a-1, suggesting the continuation of a recent phase of slight mass loss following a prolonged period of balanced mass budgets previously observed. The mean annual mass balance in Western Nyainqêntanglha for the same period was highly negative (-0.60 ±0.15 m w.e. a-1 on average), suggesting increased mass loss rates. The 2022 winter (+0.21 ±0.24 m w.e.) and summer (-0.31 ±0.15 m w.e.) mass budgets in Muztag Ata and Western Nyainqêntanglha (-0.04 ±0.27 m w.e. [winter]; -0.66 ±0.07 m w.e. [summer]) suggest winter and summer accumulation-type regimes, respectively. We support our findings by implementing a Sentinel-1–based Glacier Index to identify the firn and wet snow areas on glaciers and characterize accumulation type. The good match between the geodetic and Glacier Index results demonstrates the potential of very high-resolution Pleiades data to monitor mass balance at short time scales and improves our understanding of glacier accumulation regimes across High Mountain Asia.

The ongoing glacier shrinkage in the Alps requires frequent updates of glacier outlines to provide an accurate database for monitoring, modelling purposes (e.g. determination of run-off, mass balance, or future glacier extent), and other applications. With the launch of the first Sentinel-2 (S2) satellite in 2015, it became possible to create a consistent, Alpine-wide glacier inventory with an unprecedented spatial resolution of 10 m. The first S2 images from August 2015 already provided excellent mapping conditions for most glacierized regions in the Alps and were used as a base for the compilation of a new Alpine-wide glacier inventory in a collaborative team effort. In all countries, glacier outlines from the latest national inventories have been used as a guide to compile an update consistent with the respective previous interpretation. The automated mapping of clean glacier ice was straightforward using the band ratio method, but the numerous debris-covered glaciers required intense manual editing. Cloud cover over many glaciers in Italy required also including S2 scenes from 2016. The outline uncertainty was determined with digitizing of 14 glaciers several times by all participants. Topographic information for all glaciers was obtained from the ALOS AW3D30 digital elevation model (DEM). Overall, we derived a total glacier area of 1806±60 km2 when considering 4395 glaciers >0.01 km2. This is 14 % (−1.2 % a−1) less than the 2100 km2 derived from Landsat in 2003 and indicates an unabated continuation of glacier shrinkage in the Alps since the mid-1980s. It is a lower-bound estimate, as due to the higher spatial resolution of S2 many small glaciers were additionally mapped or increased in size compared to 2003. Median elevations peak around 3000 m a.s.l., with a high variability that depends on location and aspect. The uncertainty assessment revealed locally strong differences in interpretation of debris-covered glaciers, resulting in limitations for change assessment when using glacier extents digitized by different analysts. The inventory is available at https://doi.org/10.1594/PANGAEA.909133 (Paul et al., 2019).

Glacier monitoring has been internationally coordinated for more than 125 years. Despite this long history there is no unambiguous answer to the popular question: which are the world’s largest glaciers?In this study, we present a first scientific assessment of the largest glaciers in the world – distinct from the two ice sheets in Greenland and Antarctica – and in the 19 regions used for the current Randolph Glacier Inventory. Ranking glaciers by size is non-trivial since it depends on how an individual glacier is defined and mapped. It is also important to differentiate between individual glaciers and glacier complexes, which are contiguous glaciers that meet at ice divides and might form an ice cap or ice field.We find that the largest glacier complexes cover areas larger than ten thousand square kilometres, whereas the largest individual glaciers cover up to several thousand square kilometres. The world’s largest glaciers and glacier complexes are located on the Antarctic Peninsula, on sub-Antarctic Islands, in the Arctic, and in Patagonia. As such, the largest glacier complexes cover areas the size of smaller countries (e.g., Switzerland or Austria) or of smaller US states (e.g. New Jersey or South Carolina), but are still orders of magnitudes smaller than the Greenland and Antarctic Ice Sheets.In addition, we show that the ranking of glaciers requires not only clear definitions but depends on the availability, quality, and consistency of digital glacier outlines at global scale. Corresponding additional metadata are required in the available inventories to fully automate a glacier ranking by area, and to extend such a study to rankings by length, volume/mass, and other parameters.

The Randolph Glacier Inventory (RGI) is a globally complete collection of digital glacier outlines, excluding the two ice sheets. It has become a pillar of glaciological research at global and regional scales for estimates of recent and future glacier changes, glacier mass balance, glacier contribution to sea-level rise, among others. The latest RGI version (V6) was released in July 2017.Here, we present a new version of the RGI (version 7.0), which is our best estimate of global glacier outlines around the year 2000. Unlike previous versions which were compiled by an ad-hoc manual process using different sources, RGI7.0 is generated directly from the Global Land Ice Measurements from Space (GLIMS) glacier database, ensuring full traceability of single outlines to their original authors. The dataset is generated automatically with Python scripts parsing the GLIMS database and selecting outlines according to community decisions (based on data availability, quality and closeness to the year 2000). Prior to its release, the dataset was available for open review from the scientific community, and further refined as necessary.About 70% of the outlines (30% of the total area) in RGI7.0 are obtained from new inventories that were submitted to GLIMS since the last release of RGI6.0 by different groups around the world. This led to considerable quality improvements especially in High Mountain Asia, Northern Canada, northern Greenland, Caucasus and Middle East, South America and New Zealand. RGI7.0 includes updated topographical and geometrical glacier attributes generated with a new community software. The new RGI generation process is open-source, fully reproducible and easily adaptable, making future updates straightforward to generate.

Svalbard is famous for its numerous surge-type glaciers as well as for the harsh weather conditions of a highly maritime Arctic island, making regular observations of its glaciers challenging. However, the rapid changes of glacier geometry require a frequent update of their extent to perform accurate glacier-specific calculations such as their mass balance or contribution to sea level. The last inventory for Svalbard has been compiled by Nuth et al. (2013) from about 40 satellite scenes acquired by three different sensors (ASTER, Landsat, SPOT) on 30 unique days over a period of 10 years. Accordingly, any change assessment or other time dependent calculations are difficult to perform and a temporarily more consistent dataset is urgently required.In this study we present the results of a new glacier inventory for Svalbard that has been derived from two Sentinel-2 swaths acquired for the main island within 3 days of 2017 and on 1 day in 2016 from Landsat 8 for Nordaustlandet. The images had overall very good snow conditions but in some regions late seasonal snow was hiding glaciers. Glacier mapping under local clouds in the very north and south could be performed by using further scenes from 2017 processed with GEE. We applied a simple red/SWIR band ratio to map clean ice and corrected wrong classifications (sea ice, lakes) or missing parts (debris cover) manually. New drainage divides and topographic parameters were derived from the ArcticDEM. The new inventory counts 3136 glaciers >0.01 km2 covering an area of 32,948 km2. Of these, glaciers < 1 km2 cover 1.3% of the area but nearly 44% of the number whereas glac-iers >10 km2 cover 91% of the area and 10% by number. Compared to the previous inventory we have 1468 glaciers more and 2.5% area less. However, when excluding the 2025 glaciers 2, we only identified 1111 glaciers, i.e. 557 less than in the previous inventory. The differences are mostly due to newly considered entities, different drainage divides, glacier retreat and advance/surging. By excluding surge-type glaciers, a more meaningful determination of climate-related area changes can be performed. The presentation will discuss the differences of the new inventory to the RGI dataset, the specific glacier mapping challenges and our approach to solve them.

B Robson was supported by the Meltzer foundation and a University of Bergen grant. S MacDonell was supported by CONICYT-Programa Regional (R16A10003) and the Coquimbo Regional Government via FIC-R(2016)BIP 40000343. D. Hölbling has been supported by the Austrian Science Fund through the project MORPH (Mapping, Monitoring and Modeling the Spatio-Temporal Dynamics of Land Surface Morphology; FWF-P29461-N29). N Schaffer was financed by CONICYT-FONDECYT (3180417) and P Rastner by the ESA Dragon 4 programme (4000121469/17/I-NB). Rock glaciers are an important component of the cryosphere and are one of the most visible manifestations of permafrost. While the significance of rock glacier contribution to streamflow remains uncertain, the contribution is likely to be important for certain parts of the world. High-resolution remote sensing data has permitted the creation of rock glacier inventories for large regions. However, due to the spectral similarity between rock glaciers and the surrounding material, the creation of such inventories is typically conducted based on manual interpretation, which is both time consuming and subjective. Here, we present a novel method that combines deep learning (convolutional neural networks or CNNs) and object-based image analysis (OBIA) into one workflow based on freely available Sentinel-2 optical imagery (10 m spatial resolution), Sentinel-1 interferometric coherence data, and a digital elevation model (DEM). CNNs identify recurring patterns and textures and produce a prediction raster, or heatmap where each pixel indicates the probability that it belongs to a certain class (i.e. rock glacier) or not. By using OBIA we can segment the datasets and classify objects based on their heatmap value as well as morphological and spatial characteristics. We analysed two distinct catchments, the La Laguna catchment in the Chilean semi-arid Andes and the Poiqu catchment in the central Himalaya. In total, our method mapped 108 of the 120 rock glaciers across both catchments with a mean overestimation of 28%. Individual rock glacier polygons howevercontained false positives that are texturally similar, such as debris-flows, avalanche deposits, or fluvial material causing the user's accuracy to be moderate (63.9–68.9%) even if the producer's accuracy was higher (75.0–75.4%). We repeated our method on very-high-resolution Pléiades satellite imagery and a corresponding DEM (at 2 m resolution) for a subset of the Poiqu catchment to ascertain what difference image resolution makes. We found that working at a higher spatial resolution has little influence on the producer's accuracy (an increase of 1.0%), however the rock glaciers delineated were mapped with a greater user's accuracy (increase by 9.1% to 72.0%). By running all the processing within an object-based environment it was possible to both generate the deep learning heatmap and perform post-processing through image segmentation and object reshaping. Given the difficulties in differentiating rock glaciers using image spectra, deep learning combined with OBIA offers a promising method for automating the process of mapping rock glaciers over regional scales and lead to a reduction in the workload required in creating inventories. Publisher PDF

Svalbard is dominated by large (often calving) glaciers and ice caps with a strong contribution to global sea-level rise. Due to many surge-type glaciers, large changes of glacier extents are common and determination of their mass balance requires a regular update of their outlines. However, frequent cloud cover prevents accurate repeat mapping. In consequence, the last glacier inventory for Svalbard was compiled from satellite scenes acquired over a period of 11 years, making change assessment and other applications difficult. Due to long-lasting seasonal snow and confusion with large perennial snow patches, the minimum size of this inventory has been set to 1 km2.Here we present a new glacier inventory for Svalbard that has been compiled at 10 m resolution from two Sentinel-2 scenes that were acquired only two days apart. Sea ice, ice-bergs, lakes and turbid water were wrongly classified as glaciers by the applied band ratio method and manually removed. Debris cover, snow and ice under some clouds but also polluted (very dark) clean ice was not mapped as thresholds were optimized to get snow and ice in shadow properly mapped. These missing regions were manually added. Snow patches were removed with a 5 by 5 majority filter applied to the binary glacier map and a minimum size of 0.05 km2. Outlines from the previous inventory as available in the RGI were used to guide the corrections. After careful comparison, we used the Arctic DEM to derive surface drainage divides and topographic attributes for all glaciers.The largest challenges for accurate glacier delineation are discrimination of debris-covered glaciers from peri-glacial debris and rock glaciers, handling of attached seasonal or perennial snowfields, and identifying disintegrating tongues of down-wasting and often debris-covered ice masses remaining after a surge. Compared to the previous inventory, the large area gains and losses of surge-type glaciers are remarkable, but area differences result also from a different interpretation of debris-covered glaciers, inclusion of snow-filled couloirs and several new glaciers that were excluded in the previous inventory.

Creating glacier inventories from satellite images and a digital elevation model (DEM) has become quasi standard. Besides the specific challenges for glacier mapping, also the selection of the ‘best’ DEM can be difficult. When using it to derive surface drainage divides and topographic information for each glacier, one has to consider the date of acquisition, artefacts, spatial completeness (data voids) and resolution. In general, using different DEMs gives different drainage divides and thus other glacier sizes. Moreover, due to widespread glacier retreat and rapid surface lowering, topographic information from older DEMs is increasingly biased towards too high values.In this study we analyse seven freely available DEMs for the Arctic region of Svalbard: ALOS AW3D30, two National Elevation Datasets (NEDs), Arctic DEM, TanDEM-X (90 and 30 m products) and the ASTER GDEM2. All individual DEM tiles were mosaicked and re-projected bilinearly to UTM 33 N. Comparisons of topographic data are performed for three test regions: a) stable terrain (off glaciers), b) glaciers in rough topography, and c) flat glaciers and ice caps.Overlay of drainage divides indicate large area differences on flat ice caps and small ones in rough topography, where mountain ridges are distinct. On the other hand, different spatial resolution results in large differences in rough topography but plays only a minor role for flat topography. Only 2 m elevation differences on stable terrain in flat valley bottoms were detected between the ALOS DEM (79.9m) and the two NEDs (77.9 m). No differences were found between the TanDEM-X 90 / 30 m and the Arctic DEM (all 109. 9 m). The ellipsoid-geoid difference is thus ~30 m in this region.Mean elevations of glaciers with flat topography or ice caps differ only slightly, but in steeper topography they reach 6 to 8 m. These differences are also due to the different resolution of the DEMs. In all test regions, only small gaps are detected in the Arctic DEM and artefacts are especially present in the ALOS DEM. For this region the ‘best’ DEM is the TanDEM-X DEM.

The Randolph Glacier Inventory (RGI) is a globally complete collection of digital glacier outlines, excluding the two polar ice sheets. It has become a pillar of glaciological research at global and regional scales, among others for estimates of recent and future glacier changes, glacier mass balance, and glacier contribution to sea-level rise. After its creation in 2012, the dataset’s further development has been coordinated by an IACS Working Group (WG) until 2019. This new WG (2020 - 2023) expands the scope of the previous one with new and updated objectives.The latest RGI version (V6) was released in July 2017, and several new glacier outline datasets have been generated by the community since then. In the past, the RGI was updated by an ad-hoc manual process, which was effective but labor-intensive. One of the main objectives of the WG is to automate this process as much as possible by incorporating RGI generation tools into the Global Land Ice Measurements from Space (GLIMS) glacier database. Furthermore, the RGI (as of version 6) needs further improvements to remain useful to the wider scientific community. Examples include data quality (wrong/outdated outlines, ice divides) but also the quality and availability of glacier attributes (hypsometry, glacier type, ...). Additionally, there is a demand for consistent historic glacier outlines (e.g. from the mid-1980s or earlier) to facilitate validation of glacier evolution models or transient mass balance calculations. With this WG, we strive to continuously improve and update the RGI, as well as to lay out a long-term plan for sustainable continuation of the RGI beyond the end of this WG.In this presentation, we will discuss the current status and future of the RGI, and will engage with the community to encourage participation and feedback.

Rock glaciers and other ice-debris landforms (I-DLs) are an important part of the debris-transport system in high mountains and their internal ice could provide a relevant contribution to water supply especially in dry regions. Recent research has shown that I-DLs are abundant in High Mountain Asia, but knowledge about their occurrence and characteristics is still limited.We are therefore investigating I-DLs in the Poiqu basin (~28°17´N, 85°58´E) – central Himalaya/southern Tibetan Plateau using remote sensing aided by field observations. We use very high-resolution stereo Pleiades data from the contemporary period and stereo Corona and Hexagon data from the 1970s to generate digital elevation models, applied satellite radar interferometry based on ALOS-1 PALSAR and Sentinel-1 SAR data and feature tracking using Sentinel-2 and the Pleiades data. Generated DEMs allowed us to create a hillshade to support identification, to derive their topographical parameters and to investigate surface elevation changes. I-DLs were identified and classified based on their characteristic shape, their surface structure and surface movement. Field observationssupported the identification of the landforms.We found abundant occurrence of rock glaciers (with typical characteristics like lobate-shaped forms, ridges and furrows as well as steep fronts) but also significant movements of both former lateral moraines and debris-slopes in permafrost area. Preliminary results revealed the occurrence of more than 350 rock glaciers covering an area of about 21 km2. About 150 of them are active. The largest rock glacier has an area of 0.5 km2 and three have an area of more than 0.3 km2. The rock glaciers are located between ~3715 m and ~5850 m with a mean altitude of ~5075 m a.s.l.. The mean slope of all rock glaciers is close to 17.5° (min. 6.8°, max. 37.6°). Most of the rock glaciers face towards the Northeast (19%) and West (18.5%). Surface elevation changes between the 1970s and 2018 show no significant changes but indicate slight elevation gain at the front of active rock glaciers caused by their downward movements.Work will be continued to generate an inventory of all I-DLs in the study area including information about their activity and surface elevation changes.

Mapping snow cover (SC) on glaciers at the end of the ablation period provides a possibility to rapidly obtain a proxy for their equilibrium line altitude (ELA) which in turn is a metric for the mass balance. Satellite determination of glacier snow cover, derived over large regions, can reveal its spatial variability and temporal trends. Accordingly, snow mapping on glaciers has been widely applied using several satellite sensors. However, as glacier ice originates from compressed snow, both have very similar spectral properties and standard methods to map snow struggle to distinguish snow on glaciers. Hence, most studies applied manual delineation of snow extent on glaciers. Here we present an automated tool, named ‘ASMAG’ (automated snow mapping on glaciers), to map SC on glaciers and derive the related snow line altitude (SLA) for individual glaciers using multi-temporal Landsat satellite imagery and a digital elevation model (DEM). The method has been developed using the example of the Ötztal Alps, where an evaluation of the method is possible using field-based observations of the annual equilibrium line altitude (ELA) and the accumulation area ratio (AAR) measured for three glaciers for more than 30 years. The tool automatically selects a threshold to map snow on glaciers and robustly calculates the SLA based on the frequency distribution of elevation bins with more than 50% SC. The accuracy of the SC mapping was about 90% and the SLA was determined successfully in 80% of all cases with a mean uncertainty of ±19 m. When cloud-free scenes close to the date of the highest snowline are available, a good to very good agreement of SC ratios (SCR)/SLA with field data of AAR/ELA are obtained, otherwise values are systematically higher/lower as useful images were often acquired too early in the summer season. However, glacier specific differences are still well captured. Snow mapping on glaciers is impeded by clouds and their shadows or when fresh snow is covering the glaciers, so that more frequent image acquisitions (as now provided by Sentinel-2) would improve results.

Since the end of the Little Ice Age around 1850, glaciers in Europe have strongly retreated. Thanks to early topographic surveys in Switzerland, accurate maps are available, which enable us to trace glacier changes back in time. The earliest map for all of Switzerland that is usable for a detailed analysis is the Dufour map from around 1850 with subsequent topographic maps on a ~ 20 year interval. Despite the large public and scientific interest in glacier changes through time, this historic dataset has not yet been fully utilized for topographic change assessment or visualization of historic glacier extents. In this study, we use eleven historical topographic maps and more recent digital datasets for the region of Zermatt to analyze geometric changes (length, area and volume) of Findelengletscher as well as for creating animations of glacier evolution through time for use in public communication. All maps were georeferenced, the contour lines digitized, and digital elevation models (DEMs) created and co-registered. Additional digital data like the SRTM X-band DEM and high resolution laser scanning data were used to extend the analysis until 2010. Moreover, one independent DEM from aerial photogrammetry was used for comparison. During the period 1859–2010, Findelengletscher lost 3.5 km of its length (6.9 km in 2010), 4.42 ± 0.13 km2 of its area (15.05 ± 0.45 km2 in 2010) and 1.32 ± 0.52 km3 of its volume. The average rate of thickness loss is 0.45 ± 0.042 m yr− 1 for the 151 years period. Four periods with high thickness change from − 0.56 m ± 0.28 yr− 1 (1859–1881), − 0.40 ± 0.08 m yr− 1 (1937–1965), − 0.90 ± 0.31 m yr− 1 (1995–2000) and − 1.18 ± 0.02 m yr− 1 (2000–2005) have been identified. Small positive thickness changes were found for the periods 1890–1909 (+ 0.09 ± 0.46 m yr− 1) and 1988–1995 (+ 0.05 ± 0.24 m yr− 1). During its retreat with intermittent periods of advance, the glacier separated into three parts. The above changes are demonstrated through an animation (available from the supplementary material), which has been created to inform the general public.

Knowledge about the coverage and characteristics of glaciers in High Mountain Asia (HMA) is still incomplete and heterogeneous. However, several applications, such as modelling of past or future glacier development, run-off, or glacier volume, rely on the existence and accessibility of complete datasets. In particular, precise outlines of glacier extent are required to spatially constrain glacier-specific calculations such as length, area, and volume changes or flow velocities. As a contribution to the Randolph Glacier Inventory (RGI) and the Global Land Ice Measurements from Space (GLIMS) glacier database, we have produced a homogeneous inventory of the Pamir and the Karakoram mountain ranges using 28 Landsat TM and ETM+ scenes acquired around the year 2000. We applied a standardized method of automated digital glacier mapping and manual correction using coherence images from the Advanced Land Observing Satellite 1 (ALOS-1) Phased Array type L-band Synthetic Aperture Radar 1 (PALSAR-1) as an additional source of information; we then (i) separated the glacier complexes into individual glaciers using drainage divides derived by watershed analysis from the ASTER global digital elevation model version 2 (GDEM2) and (ii) separately delineated all debris-covered areas. Assessment of uncertainties was performed for debris-covered and clean-ice glacier parts using the buffer method and independent multiple digitizing of three glaciers representing key challenges such as shadows and debris cover. Indeed, along with seasonal snow at high elevations, shadow and debris cover represent the largest uncertainties in our final dataset. In total, we mapped more than 27 800 glaciers >0.02 km2 covering an area of 35 520±1948 km2 and an elevation range from 2260 to 8600 m. Regional median glacier elevations vary from 4150 m (Pamir Alai) to almost 5400 m (Karakoram), which is largely due to differences in temperature and precipitation. Supraglacial debris covers an area of 3587±662 km2, i.e. 10 % of the total glacierized area. Larger glaciers have a higher share in debris-covered area (up to >20 %), making it an important factor to be considered in subsequent applications (https://doi.org/10.1594/PANGAEA.894707).

In this work, we present a framework to integrate physically based hydroclimatological models and remote sensing products, by exploiting their advantages and overcoming their limitations, for an improved understanding and estimation of the alpine glacier accumulation and ablation processes. The capability of remote sensing to well represent the spatial variability of the snow cover over the glaciers is used to correct possible errors in the model simulations, thus obtaining a more reliable estimation of annual glacier mass balance. The proposed approach is tested on the glaciers in the Rofen Valley (Austria) by employing the AMUNDSEN model for accumulation and ablation processes simulation and Landsat-5/7/8 data for glacier zone mapping from 1998 to 2016.

The knowledge about the coverage and characteristics of glaciers in High Mountain Asia is still incomplete and heterogeneous. However, several applications such as modelling of past or future glacier development, runoff or glacier volume, rely on the existence and accessibility of complete datasets. In particular, precise outlines of glacier extent are required to spatially constrain glacier-specific calculations such as length, area and volume changes or flow velocities. As a contribution to the Randolph Glacier Inventory (RGI) and the GLIMS glacier database, we have produced a homogeneous inventory of the Pamir and the Karakoram mountain ranges using 28 Landsat TM and ETM+ scenes acquired around the year 2000. We applied a standardized way of automated digital glacier mapping and manual correction using coherence images from ALOS-1 PALSAR-1 as an additional source of information, separated the glacier complexes into individual glaciers using drainage divides derived by watershed analysis from the ASTER GDEM2, and separately delineated all debris-covered areas. Assessment of uncertainties was performed for debris-covered and clean-ice glacier parts using the buffer method and independent multiple digitizing of three glaciers representing key challenges such as shadows and debris cover. Indeed, along with seasonal snow at high elevations, shadow and debris cover represent the largest uncertainties in our final dataset. In total, we mapped more than 27'400 glaciers >0.02 km2 covering an area of 35'287 ± 1209 km2 and an elevation range from 2260 m to 8600 m with regional median glacier elevations varying from 4150 m (Pamir Alai) to almost 5400 m (Karakoram); this being largely due to differences in temperature and precipitation. The coverage of glaciers by debris is on average ~5 %, but glaciers >5 km2 have often a much higher area share (>10 %), making it an important factor to be considered in subsequent applications. Location of dataset: http://www.geo.uzh.ch/~nmoelg/glacier_inventory.zip

Precise information about the size and spatial distribution of glaciers is needed for many research applications, for example water resources evaluation, determination of glacier specific changes in area and volume, and for calculation of the past and future contribution of glaciers to sea-level change. However, mapping glacier outlines is challenging even under optimal conditions due to time consuming manual corrections of wrongly classified pixels. In the last decades, advantages in computer technologies have led to the development of object-based-image analysis (OBIA), an image classification technique that can be seen as an alternative to the common pixel-based image analysis (PBIA). In this study we compare the performance of OBIA with PBIA for glacier mapping in three test regions with challenging mapping conditions. In both approaches, a ratio image was created to map clean snow and ice while thermal and slope information was used to assist in the identification of debris-covered ice. The mapping results of OBIA have overall a ~ 3% higher quality than PBIA, in particular in the processing of debris-covered glaciers where OBIA has a 12% higher accuracy. The post-processing possibilities in OBIA (e.g., the application of a processing loop and neighborhood analysis) are especially powerful to improve the final classification. This leads also to a reduction of the workload for the manual corrections, which are still required to achieve a sufficient accuracy.

Deriving glacier outlines from satellite data has become increasingly popular in the past decade. In particular when glacier outlines are used as a base for change assessment, it is important to know how accurate they are. Calculating the accuracy correctly is challenging, as appropriate reference data (e.g. from higher-resolution sensors) are seldom available. Moreover, after the required manual correction of the raw outlines (e.g. for debris cover), such a comparison would only reveal the accuracy of the analyst rather than of the algorithm applied. Here we compare outlines for clean and debris-covered glaciers, as derived from single and multiple digitizing by different or the same analysts on very high- (1 m) and medium-resolution (30 m) remote-sensing data, against each other and to glacier outlines derived from automated classification of Landsat Thematic Mapper data. Results show a high variability in the interpretation of debris-covered glacier parts, largely independent of the spatial resolution (area differences were up to 30%), and an overall good agreement for clean ice with sufficient contrast to the surrounding terrain (differences ∼5%). The differences of the automatically derived outlines from a reference value are as small as the standard deviation of the manual digitizations from several analysts. Based on these results, we conclude that automated mapping of clean ice is preferable to manual digitization and recommend using the latter method only for required corrections of incorrectly mapped glacier parts (e.g. debris cover, shadow).

Heavy rainfall in June 2013 triggered flash flooding and landslides throughout the Indian Himalayan state of Uttarakhand, killing more than 6000 people. The vast majority of fatalities and destruction resulted directly from a lake outburst and debris flow disaster originating from above the village of Kedarnath on June 16 and 17. Here, we provide a systematic analysis of the contributing factors leading to the Kedarnath disaster, both in terms of hydrometeorological triggering and topographic predisposition. Topographic characteristics of the lake watershed above Kedarnath are compared with other glacial lakes across the north-western Himalayan states of Uttarakhand and Himachal Pradesh, and implications for glacier lake outburst hazard assessment in a changing climate are discussed. Our analysis suggests that the early onset of heavy monsoon rainfall (390 mm, June 10–17) immediately following a 4-week period of unusually rapid snow cover depletion and elevated streamflow was the crucial hydrometeorological factor, resulting in slope saturation and significant run-off into the small seasonal glacial lake. Between mid-May and mid-June 2013, snow-covered area above Kedarnath decreased by around 50 %. The unusual situation of the lake being dammed in a steep, unstable paraglacial environment but fed entirely from snowmelt and rainfall within a fluvial dominated watershed is important in the context of this disaster. A simple scheme enabling large-scale recognition of such an unfavourable topographic setting is introduced. In view of projected 21st century changes in monsoon timing and heavy precipitation in South Asia, more emphasis should be given to potential hydrometeorological triggering of lake outburst and debris flow disasters in the Himalaya.

In this study, we investigated the use of synthetic aperture radar (SAR) polarimetry (Pol-SAR) and a supervised classification technique, support vector machine (SVM), for the classification of bare soil, ice, and snow, over the Ortles–Cevedale massif, (Eastern Italian Alps). We analyzed the importance of topographic correction on the backscattering and polarimetric SAR signature and the advantage of quad-pol with respect to dual-pol data. When backscattering values only are employed, the incidence angle used as input feature of the SVM classifier assures the best classification accuracy, 9.9% higher than the accuracy obtained with cosine corrected ${\gamma ^0}$ backscattering. The introduction of polarimetric features and decomposition parameters (such as Cloude–Pottier or Touzi decomposition parameters) increases the classification accuracy by 5.2% with respect to the backscattering case. The simulation of RADARSAT-2 data as Sentinel-1 like for dual-pol data shows a decrease of accuracy equal to 7.8% with respect to the fully polarimetric case (93.5%). The first Sentinel-1 image acquired on our test area was also employed for classification. We then tested the capability of C-band SAR to detect accumulation and ablation zones of the glaciers under the winter dry snow by setting up a multi-incidence angle and fully polarimetric SVM classifier, exploiting ascending and descending RADARSAT-2 data. In this case, the accuracy increased by 14.7% combining different geometric acquisitions (88.9%) with respect to the single geometry case. Finally, from the resulting classification maps, we extracted the snowline altitude for a sample of three glaciers, using both optical and SAR data, comparing the different products.

Glacier inventories provide essential baseline information for the determination of water resources, glacier-specific changes in area and volume, climate change impacts as well as past, potential and future contribution of glaciers to sea-level rise. Although Greenland is heavily glacierised and thus highly relevant for all of the above points, a complete inventory of its glaciers was not available so far. Here we present the results and details of a new and complete inventory that has been compiled from more than 70 Landsat scenes (mostly acquired between 1999 and 2002) using semi-automated glacier mapping techniques. A digital elevation model (DEM) was used to derive drainage divides from watershed analysis and topographic attributes for each glacier entity. To serve the needs of different user communities, we assigned to each glacier one of three connectivity levels with the ice sheet (CL0, CL1, CL2; i.e. no, weak, and strong connection) to clearly, but still flexibly, distinguish the local glaciers and ice caps (GIC) from the ice sheet and its outlet glaciers. In total, we mapped ~ 20 300 glaciers larger than 0.05 km2 (of which ~ 900 are marine terminating), covering an area of 130 076 ± 4032 km2, or 89 720 ± 2781 km2 without the CL2 GIC. The latter value is about 50% higher than the mean value of more recent previous estimates. Glaciers smaller than 0.5 km2 contribute only 1.5% to the total area but more than 50% (11 000) to the total number. In contrast, the 25 largest GIC (> 500 km2) contribute 28% to the total area, but only 0.1% to the total number. The mean elevation of the GIC is 1700 m in the eastern sector and around 1000 m otherwise. The median elevation increases with distance from the coast, but has only a weak dependence on mean glacier aspect.

A reconstruction method of historical glacier topographies and possible uses of these results are demonstrated in this article. This reconstruction was accomplished for 310 Alpine glaciers in South Tyrol, Italy. These glaciers are featured with a wealth of different historical (e.g. paintings, photographs and historical maps) and recent data sources (airborne laser scan based digital terrain model and digital orthophotos) that allow the reconstruction of the Little Ice Age maximum extension. These sources are among the best historical and recent documents of glaciers for the mid-nineteenth century. The results of this reconstruction visualize the ongoing climate change in a comprehensive way. The area changes between the time of the Little Ice Age maximum extent (around the year 1850) and the recent glaciation in 2006 amounts to a loss of 182.4 km2 or almost 66%. In the same time, the calculated mean equilibrium line altitude for all South Tyrolean glaciers rose approximately by 160 m.

The cryosphere of mountain regions is highly sensitive to climate change. This is particularly evident in region wide retreat of glaciers and reduced snow coverage in mountain areas. Snow cover is an important parameter in the glacier mass but also energy balance as it controls the energy fluxes on the glaciers surface and protects the glacier from melt. Monitoring snow cover and the snow line altitude of a glacier through remote sensing data near the end of the ablation season may be considered roughly representative for the equilibrium line altitude on Alpine glaciers which is a useful indicator of the annual mass balance. The aim of this study is to develop an automated tool to retrieve information related to glacier mass balance based on multitemporal satellite imagery and an already existing glacier inventory. The method is developed in the Otztal Alps where reliable in-situ data are available for several glaciers for a period longer than 30 years. The results show a clear rise of the snow line altitude, which is in agreement with the equilibrium line altitude observed in the field. Nevertheless, the retrieved snow lines by remote sensing imagery are generally too low (> 200 m difference). Part of this underestimate is likely due to the mapping method or the usage of a present DEM which is not representative for historical snow line altitudes. After a proper validation of the current methodology, it can subsequently transferred to other regions in the world.

A widespread loss of glacier area and volume has been observed in the European Alps since the 1980s. In addition to differences among various regions of the Alps, different responses to climate change characterize neighboring glaciers within the same region. In this study we describe the glacier changes in the Ortles-Cevedale group, the largest glacierized area in the Italian Alps. We analyze the spatial variability, the drivers, and the main factors controlling the current loss of ice in this region, by comparing mean elevation changes derived from two digital terrain models (DTMs), along with glacier extents and snow-covered areas derived from Landsat images acquired in 1987 and 2009, to various topographic factors. Glacier outlines were obtained using the band ratio method with manual corrections. Snow was classified from a near-infrared image after topographic correction. The total glacierized area shrank by 23.4 ± 3% in this period, with no significant changes in the mean altitude of the glaciers. In 2009 the snowline was 240 m higher than in the 1960s and 1970s. From the snow-covered area at the end of summer 2009, which fairly represents the extent and local variability of the accumulation areas in the 2000s, we estimate that approximately 50% of the remaining glacier surfaces have to melt away to re-establish balanced mass budgets with present climatic conditions. The average geodetic mass budget rate, calculated for 112 ice bodies by differencing two DTMs, ranged from −0.18 ± 0.04 to −1.43 ± 0.09 m w.e. a−1, averaging −0.69 ± 0.12 m w.e. a−1. The correlation analysis of mass budgets vs. topographic variables emphasized the important role of hypsometry in controlling the area and volume loss of larger glaciers, whereas a higher variability characterizes smaller glaciers, which is likely due to the higher importance of local topo-climatic conditions.

[1] The recently finalized inventory of Greenland's glaciers and ice caps (GIC) allows for the first time to determine the mass changes of the GIC separately from the ice sheet using space-borne laser altimetry data. Corrections for firn compaction and density that are based on climatic conditions are applied for the conversion from volume to mass changes. The GIC which are clearly separable from the icesheet (i.e., have a distinct ice divide or no connection) lost 27.9 ± 10.7 Gt a−1 or 0.08 ± 0.03 mm a−1 sea-level equivalent (SLE) between October 2003 and March 2008. All GIC (including those with strong but hydrologically separable connections) lost 40.9 ± 16.5 Gt a−1 (0.12 ± 0.05 mm a−1 SLE). This is a significant fraction (~14 or 20%) of the reported overall mass loss of Greenland and up to 10% of the estimated contribution from the world's GIC to sea level rise. The loss was highest in southeastern and lowest in northern Greenland.

Satellite data provide a large range of information on glacier dynamics and changes. Results are often reported, provided and used without consideration of measurement accuracy (difference to a true value) and precision (variability of independent assessments). Whereas accuracy might be difficult to determine due to the limited availability of appropriate reference data and the complimentary nature of satellite measurements, precision can be obtained from a large range of measures with a variable effort for determination. This study provides a systematic overview on the factors influencing accuracy and precision of glacier area, elevation change (from altimetry and DEM differencing), and velocity products derived from satellite data, along with measures for calculating them. A tiered list of recommendations is provided (sorted for effort from Level 0 to 3) as a guide for analysts to apply what is possible given the datasets used and available to them. The more simple measures to describe product quality (Levels 0 and 1) can often easily be applied and should thus always be reported. Medium efforts (Level 2) require additional work but provide a more realistic assessment of product precision. Real accuracy assessment (Level 3) requires independent and coincidently acquired reference data with high accuracy. However, these are rarely available and their transformation into an unbiased source of information is challenging. This overview is based on the experiences and lessons learned in the ESA project Glaciers_cci rather than a review of the literature.