Estimation of trends in extreme melt‐season duration at Svalbard

A time series of monthly mean surface temperatures taken at Svalbard airport, Spitzbergen, for the period 19122010 was examined for changes in melt-season length. The annual melt-season length was constructed from daily temperature estimates based on the monthly data using smoothing splines. We argue that the changes in annual melt-season length are linked to variability in regional sea surface temperatures, the mean Northern Hemisphere surface temperature and the North Atlantic Oscillation (NAO) index. A regression model for the melt-season length with these three parameters as predictors, explained about 40% of the observed variance. The annual mean melt season for the period from 1912 to 2010 was estimated to be 108 days, and the linear trend was 0.17 days/year. The risk of having positive extremes in the melt season increased with increasing Northern Hemisphere surface temperature and the regional sea surface temperatures. On the basis of our study of past observations, the 100-year return length of the melt season at Svalbard was predicted to change from the current 95% confidence interval of 131 (108, 138) days to 175 (109, 242) days with 1 degrees C warming of both regional sea surface temperature and the mean Northern Hemisphere surface temperature. Copyright (c) 2011 Royal Meteorological Society