Abstract ID: 3.12215 | Accepted as Talk | Talk/Oral | TBA | TBA

The Peruvian Andes contain glaciers which have undergone considerable mass loss over the past three decades and which face significant changes under future warming. Meltwater from snow and glaciers is important for water resources in this region, particularly in the dry season, and changes in the cryosphere could impact runoff regimes, downstream ecosystems and water users. To determine the drivers of past changes in the mountain water cycle, and investigate its likely future response under climate change, we run the fully distributed, hourly glacier-hydrological model TOPKAPI-ETH over the upper Rio Santa catchment in the Cordillera Blanca. We run the model from 1987 to present and then to 2100 in the future, forced in the past by bias-corrected WRF simulations, which are also used for statistical downscaling of 12 CMIP5 model projections which form a future climate ensemble. The ability of the model to replicate past glacier changes is assessed through comparison against historical glacier outlines, and we also evaluate model outputs against glacier mass balance, snow cover and gauged runoff datasets.

We find that past ENSO fluctuations have a significant impact on both glacier mass balance and the hydrological functioning of the catchment. Warmer El Niño air temperatures result in more negative glacier mass balances, but particularly negative mass balances are also found in low precipitation years which are not related to ENSO. Ice melt and on-glacier snowmelt is increased in El Niño years, resulting in an increase in discharge in highly glaciated sub-catchments. However, off-glacier snowmelt is decreased and discharges further downstream do not change significantly due to water loss from increased evapotranspiration. Despite the majority of the glaciers in the catchment facing overall negative cumulative mass balances over the past 30 years, there are exceptions, with positive balances found for the highest elevation glaciers. We also investigate the long term changes in the catchment runoff regime, demonstrating that the timing of ‘peak water’ in the dry season varies per sub-catchment depending on the degree of glacier recession. Our model outputs also allow us to explore the future changing resilience of the water cycle to predicted climate extremes.