Groundwater flow patterns and subsurface heterogeneity drive critical zone geochemical reactions

In continental subsurface environments, biogeochemical reactions drive nutrient delivery, deep microbial life and mineral weathering, with crucial importance in the critical zone. Current models often simplify groundwater transport, using the residence time approach or hillslope models. However, increasing observations suggest that the nature, location and efficiency of reactions are strongly affected by groundwater 3D flow patterns, chemical gradients and subsurface heterogeneity. Here, we investigate how hydrological and geological structures control where and when biogeochemical reactions occur in the deep critical zone. For this purpose, our approach integrates long-term and widespread local observations in a catchment-scale framework and is based on data from two critical zone observatories of the French OZCAR national network. The first study took place in Guadeloupe (Obsera), where we integrated geophysical, hydrological and geochemical data in a reactive-hydrogeological model to simulate the 3D structure of groundwater flow paths and weathering. We found that the downstream evolution of the river chemistry is controlled by the pattern of hydrogeological circulations and by the depth of the weathering front. Furthermore, the calibrated 3D model allowed the delimitation of areas where weathering occurs and we showed that active weathering is restricted to catchment-areas where downward groundwater flows are deep. The second study focused on the dynamics of dissolved oxygen (DO) in a fractured aquifer at the Ploemeur catchment (Bretagne, France). Deep and intermittent inputs of DO in groundwater were observed, enabling the reaction of DO with dissolved Fe²⁺, in turn sustaining the development of deep microbial communities. In this study, we designed a simple model to simulate jointly the depth-distribution of DO and Fe²⁺ and to investigate the hydrological and geological factors controlling the DO depth-distribution. We found that the reducing capacity of the bedrock and the mean fluid transit time are the main parameters to explain and predict the depth of the oxic-anoxic transition in crystalline environments. In this presentation, we will provide new perspectives to observe and understand the origin of subsurface biogeochemical reactions and we will illustrate key processes that breakdown classical assumptions of reactive groundwater models.