Simulating melt, runoff and refreezing on Nordenskiöldbreen, Svalbard, using a coupled snow and energy balance model

Abstract

Abstract. A distributed energy balance model is coupled to a multi-layer snow model in order to study the mass balance evolution and the impact of refreezing on the mass budget of Nordenskiöldbreen, Svalbard. The model is forced with output of a regional climate model (RACMO) and meteorological data from Svalbard Airport. Extensive calibration and initialisation are performed to increase the model accuracy. For the period 1989–2010, we find a mean net mass balance of −0.39 m w.e. a−1. Refreezing contributes on average 0.27 m w.e. a−1 to the mass budget and is most pronounced in the accumulation zone. The simulated mass balance, radiative fluxes and subsurface profiles are validated against observations and are generally in good agreement. Climate sensitivity experiments reveal a non-linear, seasonally dependent response of the mass balance, refreezing and runoff to changes in temperature and precipitation. Output of the climate sensitivity experiments is used in combination with temperature and precipitation time-series to extend mass balance time-series in the past and the future to obtain estimates for the period 1912–2085. It is shown that including seasonality in climate change, with less pronounced summer warming, has a major impact on future mass balance and ELA estimates. Due to compensating effects, the contribution of refreezing hardly changes in a future climate.