Sabine Attinger, Helmholtz Centre for Environmental Research

Title: …

Felipe De Barros, University of Southern California

Title: Uncertainty, Connectivity, and Transport Extremes in Heterogeneous Aquifers

Quantifying contaminant transport in heterogeneous aquifers remains a central challenge for groundwater management and public health protection. Limited site characterization and multi-scale variability in hydraulic properties introduce significant uncertainty in key quantities, including solute concentrations and arrival times at environmentally sensitive locations. Early arrival times, which define the leading edge of contaminant plumes, are particularly difficult to estimate and are often subject to the greatest uncertainty. Because regulatory decisions are frequently driven by threshold exceedances, characterizing the full probabilistic distribution of these quantities, particularly their extreme tails, is essential. This talk presents a unified perspective linking probabilistic transport modeling with concepts of hydraulic connectivity to better characterize transport extremes. Advances in the stochastic description of solute concentration are highlighted, emphasizing how heterogeneity, local-scale dispersion, and sampling scale shape the statistical distribution of concentration and control exceedance probabilities relevant to risk assessment. Connectivity metrics based on minimum hydraulic resistance and least-resistance paths provide a physically grounded means to identify preferential flow structures governing early arrivals and leading-edge plume behavior. While connectivity is itself subject to uncertainty, it provides a compact and physically interpretable representation of the subsurface structures that control transport extremes. Embedding these connectivity metrics within stochastic frameworks enables more efficient uncertainty quantification and improved site characterization, including targeted data acquisition strategies that focus on regions exerting the greatest control on high-impact transport predictions. By linking subsurface structure to transport extremes, this perspective supports more reliable risk estimates and more informed decision-making in groundwater systems.

Gabrielle De Lannoy, KU Leuven

Title: Satellite-based multivariate land surface data assimilation: from soil moisture and snow updating to impact on streamflow and atmosphere

Process-based land surface models (LSMs) simulate the temporal evolution of a suite of interconnected water- and energy-related variables, driven by meteorological inputs and physical laws. Their strength lies in their internal consistency and nearly complete representation of the system. This enables the use of satellite data assimilation to update observed variables and then rely on the model to transmit that information to unobserved variables, facilitating the development of land surface digital twin systems. Nonetheless, LSMs do not necessarily represent all processes, which means that (i) they may not take up valuable information contained in satellite observations, or (ii) the effect of assimilation on unobserved variables may be less than optimal. This presentation will address soil moisture and snow updating through satellite data assimilation into the Noah-MP LSM. Each of these variables vary on different temporal timescales, have different memories, and have specific control roles in the LSM. This requires tailored approaches to satellite data assimilation. For soil moisture updating, three case studies will be presented: (i) how assimilating Sentinel-1 backscatter interacts with (missed) irrigation modeling to produce optimal water budget estimates over Italy’s Po River basin; (ii) how the irrigation signal contained in coarse-resolution satellite soil moisture products over California can be effectively assimilated when irrigation is not modeled; and (iii) how soil moisture updating and/or explicit irrigation modeling influence atmospheric conditions and streamflow. For snow updating, the presentation will cover (i) the assimilation of Sentinel-1–derived snow depth estimates over the Alps and their subsequent effect on streamflow in the Po River basin, and (ii) recent progress towards improving snow data assimilation through the use of dynamic observation errors.

Marco Dentz, IDAEA-CSIC

Title: Modeling and upscaling of flow and transport processes in karst aquifers from conduit to network scale

The flow and transport behaviors of karst aquifers are determined by a broad range of spatial scales from the pore to the network scale, and the corresponding range of hydrodynamic flow modes from laminar to turbulent. They are important for the understanding of how karst is formed on the one hand, and how karst aquifers transmit and store water and solutes on the other. Those are key items in view of water resources management and
preservation. The disparity of scales and variety of hydrodynamic behaviors make the quantification of karst processes a challenging task. This talk gives an overview of approaches and concepts for the modeling of karst from the conduit to the network scale, and report on recent advances in the framework of the ERC-SYG project KARST. On the karst conduit scale, we discuss the impact of wall geometry on flow parameterization in terms of the Darcy friction factor, the transition from laminar to turbulent flow, and the validity of rough pipe correlations for karst conduits. We explore classical transport description in terms of hydrodynamic (turbulent) dispersion and its limitations for strong wall fluctuations. We then discuss the quantification of laminar and turbulent network-scale flow and transport in synthetic and real network topologies. In this context, we focus on the relation between network structure and the emerging flow and transport patterns, also in the light of karst generation. The impact of the simultaneous occurrence of laminar and turbulent flow on solute dispersion, and the upscaling of these processes are discussed in the light of network topology and heterogeneity in the framework of stochastic modeling using continuous time random walk approaches.

Manuela Girotto, University of California, Berkeley

Title: The Role of Sierra Nevada Mountains in Regulating Central Valley Groundwater Recharge

Climate-driven shifts in snowpack dynamics and intensified hydro-climatic extremes pose significant threats to hydrological systems dependent on snowmelt. A primary example is California’s Central Valley, which relies on groundwater extracted from deep aquifers and surface water transported from the seasonal snowpack of the Sierra Nevada. However, the mechanisms by which snowmelt recharges Central Valley groundwater remain poorly represented in existing hydrological models. This presentation introduces a novel computational framework to quantify mountain snowmelt and recharge rates by bridging the gap between hydrological models, coarse satellite gravimetry observations, and fine-scale physical recharge processes.

Hadi Hajibeygi, TU Delft

Title: Building Confidence in Underground Hydrogen Storage

Underground hydrogen storage (UHS) for energy supply-demand management is a relatively new topic compared to the natural gas storage. It is gaining increasing interest due to the ‘hope’ that hydrogen may indeed be the missing link of ‘scalable’ low-carbon energy system. Successful deployment of UHS depends on reliable performance analysis, among others, which depends on rigorous understanding of the relevant cyclic hydromechanics processes at various scales. To address this, we present a multiscale experimental-numerical framework for UHS, addressing the thermo-chemical properties at molecular scale, trapping mechanisms at micro-meter scale, and its performance and recoverability in continuum reservoir scale. The nonlinear, time-dependent mechanical response of the host rocks is also addressed, with the focus on model construction and parameter calibration under uncertainty, including field validation. Emphasizing the importance of reliable performance assessments under uncertainty and heterogeneity, some key knowledge gaps in this evolving technology will be also addressed.

Delphine Roubinet, Geosciences Montpellier CNRS

Title: …

Alexandre Tartakovsky, University of Illinois Urbana-Champaign

Title: …