Although invisible, groundwater plays an essential role for society as a source of drinking water or for ecosystems but is also facing important challenges in terms of contamination. Characterizing groundwater reservoirs with their spatial heterogeneity and their temporal evolution is therefore crucial for their sustainable management. In this paper, we review some important challenges and recent innovations in imaging and modeling the 4D nature of the hydrogeological systems.

Located in the French Southern Alps, the Cerveyrette valley constitutes a watershed of about 100 km². Cyclicality in the stable isotope compositions of the river waters recorded over two years allowed us to estimate a time lag of three to four months between precipitations and their sampling at the discharge point of the watershed. We thus show that the transfer time from mountain-accumulated snow toward the low-altitude areas is a sensitive variable responding to the current climate warming.

Rivers are known to record changes in tectonic or climatic variation through long adjustment of their longitudinal profile slope. Here we describe such adjustments in experimental landscapes and show that they may result from the sole effect of intrinsic geomorphic processes. We propose a new model of river evolution that links long profile adjustment to cycles of river widening and narrowing. This result emphasizes the need to better understand control of lateral erosion on river width.

Efficient flood mapping methods are needed for large-scale, comprehensive identification of flash flood inundation hazards caused by small upstream rivers. An evaluation of three automated mapping approaches of increasing complexity, i.e., a digital terrain model (DTM) filling and two 1D–2D hydrodynamic approaches, is presented based on three major flash floods in southeastern France. The results illustrate some limits of the DTM filling method and the value of using a 2D hydrodynamic approach.