Uplift · Upscaling reactive transport in unsaturated media: from the pore to the vadose zone

The vadose zone is the unsaturated region connecting the land surface to groundwater, acting as a filter for nutrients and contaminants and controlling their access to aquifers. Reactive transport in unsaturated porous media is of central importance to our understanding of nutrient cycles and applications such as soil and groundwater remediation. Geological media are heterogeneous across scales, and desaturation leads to complex patterns of air and water and preferential flow channels, enhancing structural heterogeneity. Resolving all scales is unfeasible, severely limiting the capacity of current models to predict field-scale reactivity and the spatial distribution of dissolved reactant plumes in the vadose zone. Uplift proposes a shift in perspective from empirical parameterization to the development of upscaled models rooted in a firm understanding of pore-scale dynamics. The underlying hypothesis is that a framework capable of quantifying subscale mixing limitations, combined with a coarse field-scale description of flow, can significantly improve current models. This requires a better understanding of pore-scale dynamics under unsaturated conditions, along with a theoretical framework capable of capturing their impact at larger scales. Uplift aims to address these challenges by connecting stochastic models of transport to statistical pore-scale features, and combining them with a novel population dynamics framework for reactive transport connecting effective reaction rates to delays due to transport limitations. To support the theoretical developments and complement available data, resolved simulation tools developed in the project will be employed to assess the role of degree of saturation, pore-scale structure, and transient forcings on phase distribution statistics and solute mixing. The proposed framework and its numerical implementation are expected to result in new reactive transport models for the vadose zone with significantly improved predictive power.