Estimation of the future water balance and water resources of the Pyrenees

Mountainous areas are an important source of water resources, especially in the Mediterranean. The PIRAGUA project aims at assessing the water resources of the Pyrenees in the past and in the future. To this aim, different modelling approaches were used in order to assess the water resources of the Pyrenees and their future evolution. 

In this study, statistically downscaled climate scenarios, generated within the CLIMPY project were used in order to force four different modelling tools: SWAT, SURFEX, RECHARGEand GIS-Balan. SWAT is a semi-distributed hydrological model, SURFEX is a distributed physically based land-surface model, RECHARGE is a simple potential recharge estimation method base on water balance model for effective precipitation computation and GIS-Balan is a GIS-based groundwater model. 

With SWAT and GIS-Balan we used a delta-change approach to apply the scenarios, and with SURFEX and RECHARGE we used an analogue methodology, which used the SAFRAN-PIRAGUA gridded dataset of meteorological variables as the observational dataset. This way, we covered many sources of uncertainty, and provided an incomplete, but large, representation of the sources of uncertainty (GCMs, RCPs, downscaling methods and hydrological models) at play.

In this exercise, we found that the resulting uncertainties are rather large for almost all variables except temperature. Temperature will very likely increase more than 4 degrees at the end of the century for the RCP85 scenario. Precipitation changes, however, are quite uncertain, although we should expect decreases on the northern slope of the Pyrenees. On the southern slope, the different projections disagree on the sign of the change. They also agree on increases of precipitation on the eastern basins of the domain (Mediterranean), while it is very likely that there will be less solid precipitation (snowfall) in the future. From here, the uncertainties increase, due to the non-linearity of the hydrological models. In the SWAT approach, aridity does not change on average. However, SURFEX projects a likely increase in aridity all over the domain. In terms of water yield, SWAT presents a drier future on the northern and western slopes, but wetter on the southern and eastern slopes. SASER shows a similar picture but is generally much drier than SWAT in the future. In terms of seasonality, the water yield will decrease mainly in summer, but also in spring and autumn and, according to SASER, also in winter, especially for the RCP85 scenario. The RECHARGE model leads to a general decrease of the potential recharge over the whole domain that could be more severe in the northern and southern part of the Pyrenees than in the Central part. The GIS-Balan models report a clear decrease in total discharge flow of the basins, which is most pronounced in the RCP8.5 scenario.

We hope that these results, although uncertain, will serve to plan for the certainty that changes in the annual means and seasonality of the water cycle are coming, even if we do not know clearly what these changes will look like. 

This work has been funded by the EFA210/16 PIRAGUA project.