Abstract How mountain headwaters and their groundwater‐dependent ecosystems will respond to future climate change remains largely unknown. These challenges stem from the difficulty of gathering relevant hydrological observations and implementing modeling strategies suited to data‐scarce systems. To address this, we present a hydrological modeling framework to quantify and predict changes in groundwater discharge to headwater springs and streams. We applied a process‐based 3D groundwater flow model to a 4 km 2 crystalline alpine catchment in the Saint‐Barthélemy Massif, French Pyrenees, with steep elevation gradients (1,120–2,350 m). In the absence of wells, subsurface hydraulic properties were calibrated using surface data, including stream networks and streamflow measurements, estimating hydraulic conductivity, specific yield, and their depth‐dependent decay. The calibrated model captures the compartmentalized aquifer structure typical of steep crystalline mountain regions, satisfactorily reproducing spring locations, the expansion and contraction of streams and wetlands, and the catchment's associated streamflow dynamics. Forced with IPCC scenarios (RCP2.6, 4.5, 8.5), simulations suggest that by 2040, half of the historical hydrographic network will experience drier low‐flow conditions, with many upstream springs potentially drying. Projected water table declines alter the characteristic response times of the aquifer by reorganizing subsurface flow paths, strongly shaped by the local geomorphology and topography. Ridge‐dominated and steep sub‐catchments are especially vulnerable to hydrological disconnection, threatening biodiversity‐rich downstream peatlands. This replicable framework provides a robust tool to predict changes in water availability in mountain headwaters, offering critical guidance for sustainable land management and adaptive conservation of groundwater‐dependent ecosystems.
Abhervé et al. (2025) studied this question.