Rianne Giesen
Postdoc at IMAU

The ice cap Hardangerjøkulen in the past, present and future climate

Hardangerjøkulen (60.55°N, 7.43°E) is an ice cap in southern Norway with an area of 73 km2 (Fig. 1). It is the sixth largest glacier on the Norwegian mainland, situated on the northwestern border of the Hardangervidda mountain plateau. The ice cap is located in a transitional zone between the maritime climate at the western coast and the more continental climate further inland. Hardangerjøkulen ranges in altitude from 1020 to 1865 m a.s.l., with more than 80% of its total area located above 1600 m a.s.l..

Map of Hardangerjøkulen

Since 1963, mass balance measurements have been carried out on the largest, westerly draining outlet glacier Rembesdalsskåka (17 km2). The program was initiated by the Norwegian Polar Institute (NP), since 1985 the measurements are carried out by the Norwegian Water Resources and Energy Directorate (NVE) and published annually in the report ‘Glaciological investigations in Norway’. Large interannual variability in winter, summer and net mass balance is found (Fig. 2), with high winter and net balances around 1990 and the most negative net mass balance in the year 2006. Over the period 1963-2007, the mean annual mass balance was slightly positive (+0.10 m w.e.), with an average winter mass balance of +2.10 m w.e. and a summer mass balance of -2.00 m w.e. The unit m w.e. gives the mass change in meters water equivalent, hence snow depth and ice melt are converted to their water equivalent by multiplying by the ratio of snow or ice density and water density. The annual mass balance turnover (the summed absolute values of the winter and summer balance divided by two) of Rembesdalsskåka (2.05 m w.e.) lies in between values for the maritime glaciers (3.7 m w.e. at Ålfotbreen) and the continental glaciers (0.92 m w.e. on Gråsubreen) in southern Norway.

Mass balance Rembesdalsskåka 1963-2007

In October 2000 an automatic weather station (AWS) was installed on Midtdalsbreen, a northeasterly outlet glacier of Hardangerjøkulen (Figs. 1 and 3). This AWS is operated by the Institute for Marine and Atmospheric research Utrecht (IMAU), that operates AWSs at various ice-covered locations in the world. The AWS on Midtdalsbreen is situated at an altitude of 1450 m a.s.l. and measures air temperature, relative humidity, air pressure, wind speed and wind direction, incoming and reflected solar radiation, incoming and outgoing longwave radiation and the distance to the glacier surface at halfhourly intervals. The AWS data collected between October 2000 and September 2006 were analysed and used in a sophisticated surface energy balance model to calculate the local surface energy and mass balance. The mean annual air temperature at the AWS site during this period was -1.4oC with an annual range of 15oC. Annual average wind speed was 6.6 m s-1, with the highest wind speeds in autumn and winter, when storms frequently come in from the Atlantic Ocean. Net solar radiation was found to dominate the surface energy balance in summer, contributing on average 75% of the melting energy. The sensible and latent heat fluxes together account for 35%, while net longwave radiation and the subsurface heat flux are energy sinks of -8% and -2% respectively. Although the melt rate is generally larger during clear-sky conditions, almost 60% of the melt occurs under cloudy skies, because these occur much more frequently.

The AWS sites on Midtdalsbreen and Storbreen


A comparison of the AWS record from Midtdalsbreen with the record from an identical AWS on Storbreen (Fig. 3), a glacier located approximately 120 km north-east of Hardangerjøkulen, revealed interesting similarities and differences. Air temperature measured at the two sites during the period 2001-2006 was highly correlated, even when the seasonal cycle was removed. The most striking difference between the two locations is the different wind climate. Wind speeds are higher on Midtdalsbreen and are most of the time related to the large-scale circulation, while katabatic winds were frequently observed on Storbreen. Because of the lower wind speeds, the turbulent fluxes are smaller on Storbreen and net radiation makes a larger relative contribution to the melting energy (76%) than on Midtdalsbreen (66%). The melt energy is a factor 1.3 larger on Midtdalsbreen. Because the snow accumulation in winter is similar at the two AWS sites, the larger melt on Midtdalsbreen results in an earlier disappearance of the snowpack and more ice melt than on Storbreen.

In the summer of 2005, a second IMAU AWS was operational near the highest point of Hardangerjøkulen (at 1860 m a.s.l., see Fig. 1). Due to serious riming on the sensors, this AWS collapsed in November 2005. The data from this AWS were combined with the information about the local climate obtained from the AWS record on Midtdalsbreen and the mass balance measurements from NVE to construct a spatially distributed mass balance model for the entire ice cap Hardangerjøkulen. By driving the model with meteorological measurements from synoptic weather stations nearest to the ice cap, mass balances were reconstructed back to the year 1905 (using a constant, present-day geometry). The computed net mass balances show the highest interannual variability in the period before the mass balance measurements were initiated, with strongly negative values in the 1930s and 1940s (Fig. 4). Although surface slopes on Hardangerjøkulen are generally gentle, the modelled spatial mass balance distribution does show significant variations that can be associated with topographical effects on the solar irradiance.

Mass balance Rembesdalsskåka 1905-2005

The ice dynamics of Hardangerjøkulen were modelled with a two-dimensional ice-flow model, based on the shallow-ice-approximation. The unknown model parameters for ice deformation and sliding were calibrated in a dynamical way, using the length change records from Midtdalsbreen and Rembesdalsskåka, provided by NVE. First, the sensitivity of the ice cap to mass balance changes was investigated, using a simple mass balance profile derived from NVE mass balance measurements. The ice cap almost entirely disappears when the mass balance is lowered by only 0.2 m w.e., for a 0.5 m w.e. decrease in mass balance, Hardangerjøkulen vanishes within 500 years. This high sensitivity to mass balance changes is a result of the flat upper part of the ice cap: only a small change in the equilibrium-line altitude is needed to turn a significant part of the accumulation area into ablation area.

Finally, the mass balance model is coupled to the ice-flow model. With the coupled model, we can model the response of the ice cap to changes in climate variables like air temperature, precipitation and cloudiness. A simulation for the 20th century gives length changes and mass balance values in good correspondence to the observations. Linear changes in temperature and precipitation were added to the mean climate over the period 1961-1990 to simulate possible future climate change. For south-western Norway, the expected temperature increase between 1961-1990 and 2071-2100 is 3oC, combined with a 10% increase in precipitation. For this scenario, the model produces a rapidly decreasing ice volume, with an almost complete disappearance of Hardangerjøkulen by the year 2100. Even a precipitation increase of 100% cannot fully compensate the effect of a 3oC warming.


The results from the project indicate that Hardangerjøkulen will disappear in a warmer climate, if the temperature rises by 3oC before the year 2100, it will even vanish within a century. Although this scenario cannot simply be translated to other glaciers in southern Norway, the geometry and climate conditions at other ice caps in southern Norway is similar to those for Hardangerjøkulen, suggesting that these ice caps cannot survive in a warmer climate either. The valley glaciers in southern Norway have a different geometry and are generally located at higher altitudes, which could make them less sensitive to climate change than the ice caps. However, reconstructions of the glacier extent of various glaciers and ice caps in southern Norway during the Holocene indicate that all studied glaciers were absent during at least one period in the early-/mid-Holocene, when temperatures were probably lower than the temperatures projected for the end of the 21st century.


My PhD thesis is available for download (18.1 MB).

Thesis R.H. Giesen