WG: Climate Geography
Funded by the German Research Foundation (DFG), 2018 - 2021
This project is part of the DFG Research Unit “Sensitivity of high Alpine geosystems to climate change since 1850” (SEHAG). With study sites in three high Alpine catchments (Martelltal, Kaunertal and Horlachtal), the overarching goal of SEHAG is to investigate to what extent the effects of documented atmospheric changes are detectable in glaciers, hydrology, geomorphic processes, soil and vegetation dynamics, and sediment transport, and how interactions between these geosystem components may enhance or attenuate climate change impacts.
Glaciers are typically situated in complex terrain that strongly interacts with the atmospheric processes which govern the exchange of energy and mass of a glacier with its surroundings. Glaciers therefore experience an expression of larger scale weather and climate patterns that is strongly influenced by the local scale. This local scale, however, is inadequately represented in climate data based on observations, reanalyses and models. In this project, we aim to quantify how much of the total uncertainty of model-based glacier reconstructions stems from inherent glacier model errors, and how much from inadequate forcing data: (i) We will perform dynamical downscaling experiments of an ensemble of global reanalysis data sets using the Weather and Research Forecast Model (WRF) to produce temporally and spatially highly resolved data sets of the atmospheric state over the Alps which are representative of the small-scale climate processes relevant for the geosystem and specifically for glacier dynamics. These data sets will be validated using point observations from weather stations. (ii) With the Open Global Glacier Model (OGGM) we will reconstruct the evolution of glacier mass balance, runoff, ice flow, and geometry of all glaciers in the study catchments, including former glaciers that have now disappeared. The causes of glacier change will be determined through sensitivity studies. (iii) We will quantify to which degree the uncertainty of glacier reconstructions can be reduced by forcing the glacier model with dynamically downscaled atmospheric fields by repeating glacier reconstruction and glacier uncertainty determination using the results of different intermediate steps in the downscaling.
The reconstruction of local scale climate variability and glacier change will improve our ability to attribute any changes in geosystem dynamics (investigated by other SEHAG projects) to atmospheric forcing.
Funded by the Federal Ministry of Education and Research (BMBF), 2017 - 2020
Peripheral glaciers on Greenland, i.e. ice masses that are not dynamically coupled to the ice sheet since they are either entirely detached, or separated from the ice sheet by well-defined ice divides, account for approximately 5% of the ice area on Greenland. They account for even only approximately 1% of the ice volume.
However, because their surface is generally situated at far lower elevations than that of the ice sheet, they are more sensitive to moderate temperature change. Subsequently, their surface mass balance accounts for approximately 25% of the total mass balance of Greenland: during the period 2000 to 2011, the ice sheet was estimated to have contributed 0.6 mm sea-level equivalent (SLE) per year, during the same time the peripheral glaciers were estimated to have contributed 0.2 mm SLE per year. While on long time scales the contribution of the peripheral glaciers to sea level rise is necessarily limited by their comparatively small volume, they are essential for accounting for Greenland’s contribution to sea level rise during the 20th and 21st centuries.
More than one third of the area of peripheral glaciers is contained in tide water glaciers. There are no comprehensive estimates on the rates of frontal ablation (i.e., calving, subaerial and subaqueous frontal melting), but it is safe to assume that these processes add a substantial portion to the surface melt of peripheral glaciers, and thus are substantial also for the total mass change of Greenland. The unsolved problem of quantifying ocean-ice interaction is also of strong relevance on the global scale, since globally, about 40% of the area of glaciers (i.e. outside of the ice sheets in Antarctica and Greenland) is contained in tidewater glaciers, and glaciers were globally estimated to contribute 0.9 mm SLE per year during 2000 to 2011.
In the framework of this project, we will develop, implement, validate, and apply parameterizations for frontal ablation of peripheral Greenland glaciers in the Open Global Glacier Model (OGGM). OGGM is developed in an international collaboration under the lead of Ben Marzeion and able to model the surface mass balance of each of the world’s 200,000 glaciers individually. It is the only global glacier model explicitly calculating ice flow, which is a prerequisite for explicitly accounting for the frontal ablation processes resulting from ocean-ice interactions.
The development and implementation of the parameterizations for frontal ablation will be performed in collaboration with the groups working on ocean-ice interaction of the ice sheet outlet glaciers, with a special emphasis on the scale differences between peripheral and outlet glaciers, and based on the data obtained from the oceanographic and glaciological field work program of the joint project. Data from the remote sensing component of the joint project will be essential for optimization and (cross) validation. Total freshwater production rates from peripheral glaciers will be provided with their pour points to the ocean modeling groups.
Funded by the Federal Ministry of Education and Research (BMBF), 2017 - 2019
Glaciers store less than 1% of the ice on Earth. Nevertheless, melting glaciers are responsible for about one third of current rates of sea-level rise, and glaciers have very likely been more important for 20th century sea-level rise than thermal expansion and mass loss of the Greenland and Antarctic ice sheets. Additionally, glaciers change the seasonal availability of water in many drainage basins and influence geohazards in high mountain settings. Glaciers respond to climate change with a delay of decades to centuries. This implies that part of the future glacier melt is a response to past climate change. Therefore, even is global temperature rise will be limited to 1.5°C above preindustrial values, glaciers can be expected to continue to melt. However, there are no quantitative estimates on the future development of glaciers on the global scale under such low warming scenarios. We will provide these projections and analyse their results with respect to sea-level rise.
Funded by the European Space Agency (ESA), 2017 - 2019
Meltwater from glaciers currently contributes about one third to GMSL and is thus a key component of sea-level rise. The main problem in accurately estimating the contribution of glacier melt is related to the large number of glaciers (ca. 200,000), only a few hundred of which are annually measured in terms of their mass changes. Moreover, their representativeness for the mass changes of the surrounding larger mountain region is only known for a few regions, probably variable over time, and simple extrapolation schemes fail due to their diverse nature. Current best estimates of their global mass change thus rely on a combination of field and remote sensing-based observations. As these are temporarily restricted to the measurement period of the satellites, and as satellite-sampling is spatially incomplete as well, numerical models help to bridge the spatio-temporal coverage and extend the time periods backwards and forwards in time. Only one of these models (Marzeion et al., 2012) is able to extend time-series backwards in time while also considering glacier geometry change. All these models require three key datasets as an input to determine global glacier mass changes: (a) a globally complete dataset of glacier outlines from a known point in time, (b) a digital elevation model (DEM) to derive their area-elevation distribution, and (c) global meteorological datasets covering the intended modelling period. Moreover, independent calibration and validation datasets are required to achieve good quality and to quantify systematic and random errors. The validation of all glacier models currently under development (e.g., considering more processes, optimizing code, etc.) indicate that their limiting factor lies in the initial and boundary conditions. Major future improvements can thus be expected to be related to dataset (a), the quality of the glacier outlines in the global inventory. The only currently available global dataset of glacier outlines is the Randolph Glacier Inventory (RGI), which has been compiled in an ad-hoc community effort for IPCC AR5. It still contains regional shortcomings in quality (e.g. seasonal snow mapped as glaciers), but it is constantly being improved by the community as well as the Glaciers_cci project. This project will focus on two key components to further improve upon current best estimates regarding the sea-level contribution of glacier melt: (i) improvements of the model used to determine a global value, mostly based on re-calibration using additional observational data, and (ii) improvements of the quality and consistency (in a temporal sense) of the glacier inventory used for initialization.
Funded by the German Research Foundation (DFG), 2016 - 2019
Mountain glaciers and ice caps are responsible for almost one third of the current rate of sea-level rise (SLR), even though they store less than 1 % of the global ice mass. The mass redistribution caused by glacier mass change is one important cause of systematic regional differences in SLR. Glaciers will continue to be an important contributor to SLR throughout the 21st century. Past glacier mass loss was recently successfully attributed to natural and anthropogenic causes, while the individual components (e.g. natural forcing vs. internal variability, or greenhouse gases vs. aerosols) were not specified. In the proposed project, we will perform a more specific attribution of past mass losses, and - following the same principle - extend the attribution to the projected mass loss during the 21st century. This will allow a quantification of the predictability of regional SLR caused by glacier mass loss. It will increase the understanding of the sources of uncertainty and will lead to better estimates of glacier related SLR. Specifically, we will investigate the following mechanisms: (i) Due to the long response time of glaciers, mass loss committed by past forcing but not yet realized, is a strong component of glacier mass change on annual to decadal (in some regions even centennial) timescales. This mass loss commitment forms the basis for a relatively strong and spatially fine-grained predictability of future glacier mass change. We will quantify the committed mass losses on the global scale, subsequently leading to a quantification of committed regional SLR. (ii) It has been shown that modes of internal variability (e.g., NAO and IOD) cause a significant fraction to regional glacier mass change on interannual timescales. However, a comprehensive analysis on the global scale is missing. We will perform this analysis, quantifying the uncertainty in mass loss (and SLR) projections associated with internal variability. (iii) Natural forcing of the climate system has a significant effect on the global glacier mass budget (e.g., ~1 mm drop of sea level after each major volcanic eruption through glacier mass gain). As with the internal variability, we will quantify the uncertainty of projections attributable to this mechanism. (iv) It has been shown that the impact of anthropogenic forcing is detectable in past glacier mass change; but so far, no specific attribution to its individual components has been done. Analyzing the impact of individual anthropogenic forcings on the glacier mass budget will allow us to isolate the impact of different emission scenarios on projected glacier mass change and associated SLR. For all these four mechanisms, we will perform hindcast experiments in order to obtain a comprehensive uncertainty assessment. We will then perform ensemble projections in order to test whether the ensemble spread can in fact be explained by the predictability and attribution analysis.