Search

Knowledge of the ice thickness distribution of glaciers and ice caps is an important prerequisite for many glaciological and hydrological investigations. A wealth of approaches has recently been presented for inferring ice thickness from characteristics of the surface. With the Ice Thickness Models Intercomparison eXperiment (ITMIX) we performed the first coordinated assessment quantifying individual model performance. A set of 17 different models showed that individual ice thickness estimates can differ considerably - locally by a spread comparable to the observed thickness. Averaging the results of multiple models, however, significantly improved the results: on average over the 21 considered test cases, comparison against direct ice thickness measurements revealed deviations on the order of 10 ± 24% of the mean ice thickness (1σ estimate). Models relying on multiple data sets - such as surface ice velocity fields, surface mass balance, or rates of ice thickness change - showed high sensitivity to input data quality. Together with the requirement of being able to handle large regions in an automated fashion, the capacity of better accounting for uncertainties in the input data will be a key for an improved next generation of ice thickness estimation approaches.

Glaciers across the globe react to the changing climate. Monitoring the transformation of glaciers is essential for projecting their contribution to global mean sea level rise. The delineation of glacier-calving fronts is an important part of the satellite-based monitoring process. This work presents a calving-front extraction method based on the deep learning framework nnU-Net, which stands for no new U-Net. The framework automates the training of a popular neural network, called U-Net, designed for segmentation tasks. Our presented method marks the calving front in synthetic aperture radar (SAR) images of glaciers. The images are taken by six different sensor systems. A benchmark dataset for calving-front extraction is used for training and evaluation. The dataset contains two labels for each image. One label denotes a classic image segmentation into different zones (glacier, ocean, rock, and no information available). The other label marks the edge between the glacier and the ocean, i.e., the calving front. In this work, the nnU-Net is modified to predict both labels simultaneously. In the field of machine learning, the prediction of multiple labels is referred to as multi-task learning (MTL). The resulting predictions of both labels benefit from simultaneous optimization. For further testing of the capabilities of MTL, two different network architectures are compared, and an additional task, the segmentation of the glacier outline, is added to the training. In the end, we show that fusing the label of the calving front and the zone label is the most efficient way to optimize both tasks with no significant accuracy reduction compared to the MTL neural-network architectures. The automatic detection of the calving front with an nnU-Net trained on fused labels improves from the baseline mean distance error (MDE) of 753±76 to 541±84 m. The scripts for our experiments are published on GitHub (https://github.com/ho11laqe/nnUNet%5fcalvingfront%5fdetection , last access: 20 November 2023). An easy-access version is published on Hugging Face (https://huggingface.co/spaces/ho11laqe/nnUNet%5fcalvingfront%5fdetection , last access: 20 November 2023). [ABSTRACT FROM AUTHOR]

In order to assess future glacier evolution and meltwater runoff, accurate knowledge on the volume and the ice thickness distribution of glaciers is crucial. However, in situ observations of glacier thickness are sparse in many regions worldwide due to the difficulty of undertaking field surveys. This lack of in situ measurements can be partially overcome by remote-sensing information. Multi-temporal and contemporaneous data on glacier extent and surface elevation provide past information on ice thickness for retreating glaciers in the newly deglacierized regions. However, these observations are concentrated near the glacier snouts, which is disadvantageous because it is known to introduce biases in ice thickness reconstruction approaches. Here, we show a strategy to overcome this generic limitation of so-called retreat thickness observations by applying an empirical relationship between the ice viscosity at locations with in situ observations and observations from digital elevation model (DEM) differencing at the glacier margins. Various datasets from the European Alps are combined to model the ice thickness distribution of Alpine glaciers for two time steps (1970 and 2003) based on the observed thickness in regions uncovered from ice during the study period. Our results show that the average ice thickness would be substantially underestimated (∼ 40 %) when relying solely on thickness observations from previously glacierized areas. Thus, a transferable topography-based viscosity scaling is developed to correct the modelled ice thickness distribution. It is shown that the presented approach is able to reproduce region-wide glacier volumes, although larger uncertainties remain at a local scale, and thus might represent a powerful tool for application in regions with sparse observations. [ABSTRACT FROM AUTHOR]

This study investigates strategies for calibration of surface mass balance (SMB) models in the Monte Sarmiento Massif (MSM), Tierra del Fuego, with the goal of achieving realistic simulations of the regional SMB. Applied calibration strategies range from a local single-glacier calibration to a regional calibration with the inclusion of a snowdrift parameterization. We apply four SMB models of different complexity. In this way, we examine the model transferability in space, the benefit of regional mass change observations and the advantage of increasing the complexity level regarding included processes. Measurements include ablation and ice thickness observations at Schiaparelli Glacier as well as elevation changes and flow velocity from satellite data for the entire study site. Performance of simulated SMB is validated against geodetic mass changes and stake observations of surface melting. Results show that transferring SMB models in space is a challenge, and common practices can produce distinctly biased estimates. Model performance can be significantly improved by the use of remotely sensed regional observations. Furthermore, we have shown that snowdrift does play an important role in the SMB in the Cordillera Darwin, where strong and consistent winds prevail. The massif-wide average annual SMB between 2000 and 2022 falls between -0.28 and -0.07 m w.e. yr -1 , depending on the applied model. The SMB is mainly controlled by surface melting and snowfall. The model intercomparison does not indicate one obviously best-suited model for SMB simulations in the MSM. [ABSTRACT FROM AUTHOR]

In order to assess future glacier evolution and melt-water runoff, accurate knowledge on the volume and the ice thickness distribution of glaciers is crucial. However, in-situ observations of glacier thickness are sparse in many regions worldwide due to challenging field surveys. This lack of in-situ measurements can be partially overcome by remote sensing information. Multi-temporal and contemporaneous data on glacier extent and surface elevation provide past information on ice thickness for retreating glaciers in the newly deglacierized regions. Yet, these observations are concentrated near the glacier snouts, which is disadvantageous because it is known to introduce biases in ice thickness reconstruction approaches. Here, we show a strategy to overcome this generic limitation of so-called "retreat" thickness observations by applying an empirical relationship between the ice viscosity at locations with in-situ observations and observations from DEM-differencing at the glacier margins. Various datasets from the European Alps are combined to model the ice thickness distribution of Alpine glaciers for two timesteps (1970 & 2003) based on observed thickness in regions uncovered from ice during the study period. Our results show that the average ice thickness would be substantially underestimated (~40%) when relying solely on thickness observations from previously glacierized areas. Thus, a transferable topography-based viscosity scaling is developed to correct the modeled ice thickness distribution. It is shown that the presented approach is able to reproduce region-wide glacier volumes, while larger uncertainties remain at a local scale, and thus might represent a powerful tool for application in regions with sparse observations. Additionally, we derive a volume of 125.4±24.7 km2 in the 1970s for glaciers in the Swiss and Austrian Alps. [ABSTRACT FROM AUTHOR]

Glaciers in the Tien Shan mountains contribute considerably to the fresh water used for irrigation, households and energy supply in the dry lowland areas of Kyrgyzstan and its neighbouring countries. To date, reconstructions of the current ice volume and ice thickness distribution remain scarce, and accurate data are largely lacking at the local scale. Here, we present a detailed ice thickness distribution of Ashu-Tor, Bordu, Golubin and Kara-Batkak glaciers derived from radio-echo sounding measurements and modelling. All the ice thickness measurements are used to calibrate three individual models to estimate the ice thickness in inaccessible areas. A cross-validation between modelled and measured ice thickness for a subset of the data is performed to attribute a weight to every model and to assemble a final composite ice thickness distribution for every glacier. Results reveal the thickest ice on Ashu-Tor glacier with values up to 201 ± 12 m. The ice thickness measurements and distributions are also compared with estimates composed without the use of in situ data. These estimates approach the total ice volume well, but local ice thicknesses vary substantially. [ABSTRACT FROM AUTHOR]

Ice cliffs within a glacier represent a challenge for the continuity equations used in many glacier models by interrupting the validity of input parameters. In the case of Gourdon Glacier on James Ross Island, Antarctica, a ∼300–500 m high, almost vertical cliff, separates the outlet glacier from its main accumulation area on the plateau of the island. In 2017 and 2018 we conducted ice thickness measurements during two airborne ground penetrating radar campaigns in order to evaluate differences to older measurements from the 1990s. The observed differences are mostly smaller than the estimated error bars. In comparison to the in situ data, the published "consensus ice thickness estimate" strongly overestimates the ice thickness at the outlet. We analyse three different interpolation and ice thickness reconstruction methods. One approach additionally includes the mass input from the plateau. Differences between the interpolation methods have a minor impact on the ice discharge estimation if the used flux gates are in areas with a good coverage of in situ measurements. A much stronger influence was observed by uncertainties in the glacier velocities derived from remote sensing, especially in the direction of the velocity vector in proximity to the ice cliff. We conclude that the amount of in situ measurements should be increased for specific glacier types in order to detect biases in modeled ice thickness and ice discharge estimations. [ABSTRACT FROM AUTHOR]

The two Patagonian icefields cover an area of 17194km2 and they have been estimated to accommodate an ice volume of 5561km3. They are therefore the largest connected ice entities in the southern hemisphere, apart from the Antarctic ice sheet. The last distributed estimate of glacier ice thickness was solely informed by glacier geometry (Carrivick et al., 2016). Moreover, the inferred thickness fields were only loosely compared to few available measurements. The approach deliberately ignored the fast glacier flow characteristic for these icefields as well as the singular climatic setting. The Andes constitute a formidable obstacle for the tropospheric flow, which results in the formation of one of the most extreme climatic divides in the world with super humid conditions in the west and dry conditions in the east. I expect that these unique climatic and ice-dynamic conditions are imprinted in the basal topography but it remains unclear to what extent the total ice volume estimate is affected. I therefore applied a mass-conserving approach for mapping glacier ice thickness that accounts for the ice-dynamic and climatic state as well as recent changes in the glacier geometries. Moreover, the approach can readily assimilate available thickness measurements.In a first attempt to map the thickness distribution of the Northern Patagonian Icefield, the approach could be constrained by more than 160'000 individual thickness measurements. These results confirm the previous geometry-based volume estimate. Yet the spatial thickness distribution differs substantially. Differences are observed both in areas with and without observational constraints and they have severe implications for estimating the ice discharge through the many marine- and lake-terminating outlet glaciers in this area. [ABSTRACT FROM AUTHOR]

We present an ice-free topography of the Svalbard archipelago based on a two-step massconserving approach for mapping glacier ice thickness. For this reconstruction, almost 73,000 individual thickness measurements were assimilated. The considered surveys covered a total length of ~7,600 km. The approach is further informed by surface mass balance from a regional climate model and several products inferred by satellite remote sensing: a recent digital elevation model, surface elevation changes and glacier surface velocities. For Svalbard, we infer a total ice volume of 6,573 km3 . Aggregation of the values of an associated error-estimate map provides upper and lower ice-volume bounds of 5,274 and 8,555 km3. These bounds are partly exhausted by the last two estimates from volume-area scaling, which reported 5,350 and 9,089 km3 . Thickness values from a recent distributed reconstruction sum up to an elevated ice volume of 8,123 km3 (Huss and Farinotti, 2012). Part of this difference is explained by, on average, 50 % higher ice-cliffs of the marine terminating glaciers on Svalbard. [ABSTRACT FROM AUTHOR]