Abstract ID: 3.10647 | Accepted as Talk | Talk/Oral | TBA | TBA

Global warming exacerbates water scarcity by increasing the frequency of hydrological droughts. These droughts can be buffered by glaciers, which mitigate runoff variability across multiple timescales. However, as glaciers retreat, the reliability of this unique natural water reserve becomes increasingly uncertain, potentially intensifying drought stress. While the fields of hydrological modelling and glacier change modelling have advanced considerably, they are often treated independently, resulting in high levels of uncertainty. To address this, we leverage recent advances in regional and global datasets and propose a hybrid modelling approach that combines Long Short-Term Memory (LSTM) neural networks with process-based glacier modelling. Unlike traditional glacio-hydrological models that require parameterisations and basin-specific calibration, the proposed approach can leverage data-driven relationships between meteorological inputs, basin characteristics, and streamflow. We present a proof of concept in the Andes that examines the role of glaciers in the propagation of meteorological to streamflow drought. Furthermore, we explore potential changes in this influence by employing “what-if” storyline scenarios that incorporate projected glacier states. Our hybrid modelling approach improves the representation of the complex interactions between glacier dynamics and streamflow droughts. Preliminary results indicate that glaciers play a crucial role in buffering streamflow droughts, delaying and reducing drought severity under historical climate conditions. However, under “what-if” storyline scenarios, this buffering capacity declines considerably, leading to more frequent and prolonged drought events. The storyline-based analysis further suggests that the most pronounced changes occur in basins with a high dependency on glacier meltwater, particularly in arid regions. The results underscore the potential of hybrid modelling in enhancing our understanding of glacier-drought interactions, offering valuable insights for climate change communication and informing adaptation strategies in high mountain regions such as the Andes.