Assessing the impact of climate change - and its uncertainty - on snow cover areas by using cellular automata models and stochastic weather generators.

Climate change will modify the spatiotemporal distribution of water resources in the future. Snow availability in alpine systems plays an important role for water dependent ecosystems, water demand supply, tourism, and hydropower. The assessment of the impact of climate change (and its uncertainty) on snow is a key subject in determining suitable adaptation strategies in these systems. In this paper, we propose a new methodology for assessing the impact of climate change on snow cover areas (SCAs). We have developed the Monte Carlo method analysis to combine several approaches to generate multiple input series and propagate them within a previously calibrated SCA cellular automata model. This generates potential future local scenarios from regional climate models. These scenarios are used to generate multiple series by using a stochastic weather generator. The methodology also includes an approach to correct the outputs bias of the stochastic weather generators when it is needed. Finally, the historical and the corrected multiple future weather series are used to simulate the impact on the SCA by using a cellular automata model. It is a novel approach that allows us to quantify the impact and uncertainty of climate change on the SCA. The methodology has been applied to the Sierra Nevada (southern Spain), which is the most southern alpine mountain range in Europe. In the horizon 2071-2100, under the RCP 8.5 emission scenario, we estimate mean reductions of SCA that will move from 42 to 66% from December to February. The reductions are higher for the rest of the year (from March to May reductions of between 47 and 95% and from September to November reductions of between 54 and 100%). These SCA changes may be roughly equivalent to an elevation shift of snow of around 400 m.