In the context of climate change, forests will face more frequent and intense periods of water deficit. How forests react to this climatic stress will depend, to a large extent, on their access to the soil water resource. For a long time, the so-called “available water content” (AWC, in millimeters of water) has been identified as a key ecosystem parameter, modulating the forest response to water shortage. Yet, AWC is typically determined over shallow depths (e.g. 1-m), and does not consider that in many cases, trees can access deep water resources, the depth and quantity of which is currently not known. Recent works developed by our consortium have introduced the concept of Total Available Water (TAW) to trees, a concept that adds “deep water” to the AWC estimate. Deployed at scales from the forest site to Metropolitan France and supported by an interdisciplinary consortium, the TAW-tree project aims (1) to quantify the TAW reserve in forests through a combination of geophysical (WP1) and ecophysiological (WP2) approaches, (2) to upscale TAW at regional scale using remote sensing (WP3) in order to (3) quantify the influence of TAW on the functioning, growth and vulnerability of temperate and Mediterranean forests facing climate change (WP4). Our working hypotheses are (hypothesis 1) that AWC generally underestimates TAW, often in a considerable way in forests, (hypothesis 2) that the variations of TAW in a particular forest drive a large part of the inter-tree differences in their response to water shortage, (hypothesis 3) that TAW, and particularly its deep component, has a critical role in the functioning and vulnerability of forests exposed to heat and drought stresses and that it changes the forests’ contribution to groundwater discharge.