Snow water equivalent in the Alps as seen by gridded data sets, CMIP5 and CORDEX climate models

Terzago, S., von Hardenberg, J., Palazzi, E., and Provenzale, A. The Cryosphere, 11, 1625-1645,, 2017.

Figure 1. The Aletsch Glacier, the largest on the European Alps, seen from Riederalp, Switzerland. Like most glaciers around the world, also the Aletsch is retreating. In 1870 it was 3.2 kilometers longer and 300 meters thicker than it is today. Picture by Silvia Terzago, March 2016.

Modifications in the water cycle, described as the continuous movement and exchange of water in its liquid, solid and vapor forms among all the components of the climate system, are among the effects of increasing temperatures. In mountain regions water coexists in all three phases and the solid one including ice, snow and permafrost – all forming the mountain cryosphere – is important (being an essential water reserve) and vulnerable to the action of climate and environmental changes at the same time. In mountains, warming is occurring as twice as faster than in other regions and the enhanced temperature increase is leading to amplified effects both in the high-altitude ecosystems and downstream.

The temperature increase in mountain areas is reflected into a decrease in snowfall at low- and mid-altitudes, earlier snow melt and shortening of the snow cover duration with implications on the timing of the seasonal runoff and groundwater recharge.

Our paper wants to provide a picture of the current and future conditions of snow depth in the Alpine region, one of the main water sources for European countries. In the absence of a dense network of ground-based stations measuring snow depth all over the Alps, our analysis is carried out considering the best available and accessible snow water equivalent (SNW) data (SNW is a measure of snow depth) from satellite measurements and from reanalyses. Though pointing out the limitations of these datasets and the differences they exhibit with each other, they can be taken as a reference for snow depth in this area against which to compare (and validate) global and regional climate model simulations (GCM, RCM) then used for future climate projections. It is important, in fact, to see how well models reproduce a certain variable over a historical reference period before using them for future projections. We didn’t limit our analysis to SNW but also to its drivers, surface air temperature and precipitation.

We found that only GCMs with resolution equal or finer than 125 Km (these are relatively high resolutions in the GCM world!) are in closer agreement with the ensemble mean of satellite and reanalysis products. Regional climate models, whose resolution is significantly higher (~12 km), are found to overestimate snow water equivalent. It is a common feature of almost all regional and global climate model to exhibit cold bias (they underestimate temperatures, compared to a given reference) and wet biases (they overestimate precipitation, including snowfall). What about future climate projections then? RCMs and higher-resolution GCMs indicate future snow reduction across the whole Alpine chain (less pronounced for RCMs than for GCMs) from about 20% to more than 90% of the current value, the range of variability depending on the considered model and on the season.

Figure 2. Percent change in the Alpine average snow water equivalent (SNW) expected by the mid-21st century, compared to a reference period (1980-2005), in the RCP8.5 scenario (high-emissions). These estimates are provided by the highest-resolution global climate models (GCMs) participating in the Fifth Coupled Model Intercomparison Project (CMIP5, grey boxplot) and by regional climate models (RCMs) of the European Coordinated Regional Climate Downscaling Experiment (EURO-CORDEX, colored symbols).