ABSTRACT. What is the depth and frequency of rainfall infiltration? How deep to plant roots penetrate into the soils? How deep is the groundwater table? Infiltration brings the acidic and thus chemically aggressive surface fluids into contact with the basic regolith and bedrocks, and by doing so it controls weathering rates. It also determines the wetted soil depth and hence influences plant rooting depths, the latter determines the depth of the Earth’s crust penetrated and altered by terrestrial life. It also controls the infiltration depth and the water table depth through ET consumption, and a suite of soil microbial processes. The water table depth defines the redox boundary and hence a range of biogeochemical reactions. Its depth also determines its hydraulic connection with the rainfall infiltration depth, hence completing the flushing of weathering products into groundwater and streams. Thus these three mutually dependent hydrologic depths shape the plumbing system of the Earth’s Critical Zone and its structure, function and evolution. Observation syntheses are presented to illustrate the mechanisms, and a high-resolution and dynamic inverse model is used to explore the global patterns in the co-evolution of the three depths.

ABSTRACT. We propose a new active DTS method to characterize groundwater and surface water exchanges over relatively large distances with an unprecedented high spatial resolution. By monitoring thermal response in streambed sediments using fiber optic cable deployed in streambed sediments, we offer to quantify directly flow velocities and to characterize the complexity of exchange processes by estimating the spatial groundwater flows distribution along the stream and determining the spatial correlation of velocities.

ABSTRACT. An integrated hydrological model coupling a low-dimensional subsurface model with a 2D diffusive wave model is applied to the Rohrschollen Island to investigate and quantify the effect of the hydraulic restoration project on surface-subsurface interactions. The model is calibrated using piezometric measurements and further validated using surface-subsurface interactions patterns derived through thermal infrared imaging. The comparison between pre-restoration and post-restoration scenarios demonstrates the effect of restoration on infiltration and exfiltration processes.

ABSTRACT. The use of integrated surface and subsurface hydrologic models (ISSHM) to understand the flow exchange processes occurring in stream-wetland areas may give important insights to researchers and water resource managers. Monitoring and quantifying the surface water–groundwater interaction are non-resolved issues. Several techniques have been proposed based on the use of heat as a tracer. An ISSHM of a stream-wetland area in central Jutland (Denmark) is developed to create a benchmark model to better constrain the applicability of such techniques.

ABSTRACT. Here we use an integrated hydrologic model to systematically evaluate two widespread trends impacting hydrologic systems, unsustainable groundwater pumping and increasing temperatures. We evaluate behavior using a high-resolution simulation of the majority of the contiguous US. Using the continental scale model, we illustrate spatial patterns in the sensitivity of land surface dynamics to shallow groundwater perturbations. Our results highlight the utility of integrated modeling tools for diagnosing system behavior over large heterogeneous domains at high spatial and temporal resolution.

ABSTRACT. Natural systems are heterogeneous with sharp interfaces between contrasting compartments. These Critical Reactive Interfaces (CRIs) occupy small area extent yet dominate biogeochemical dynamics at the domain scale. Here the concepts and drivers of CRI will be illustrated through three systems: a heterogeneous medium where the fast approach to dissolution equilibrium drives the CRI formation; a shallow aquifer where microbe-mediated iron reduction primarily occur at the high permeability-low permeability interfaces; a watershed where CRIs is shaped by topography and land-water conjunctions.

ABSTRACT. It remains a major challenge to predict water quality and solute export out of watersheds because interactions between hydrological conditions and biogeochemical reactions are often poorly understood at the watershed scale. Field measurements suggest that distinct concentration-discharge (CQ) relations among DOC, nutrients, metals, base cations (i.e. Na, Ca, Mg) and chloride. This study aims to develop a bio-reactive transport code to understand watershed biogeochemistry at the watershed scale.

ABSTRACT. The investigation of carbon and nutrient cycling in terrestrial systems is driving the need for integrated hydrology models that include coupled surface/subsurface reactive transport. A fully coupled surface/subsurface hydrology model can be advantageous in hydrological investigations, especially for capturing the strong interaction between surface and subsurface processes. Including this unique capability in coupled fine-scale integrated models provides the opportunity to study: contaminant transport involving surface-subsurface interactions, tracer hydrographs, transport of isotopes, biogeochemical processes in the hyporheic zone.

ABSTRACT. To quantify subsurface exports of nutrient, nitrogen, and carbon to the river system at the mesoscale, we develop an upscaling technique incorporating residence time and meander geometry into the reactive transport modeling framework. Overall, we present hillslope-floodplain-meandering channel interactions and their influence on groundwater exports at the mesoscale.

ABSTRACT. Issues involving the formulation and closure of a new generation of two-fluid-phase flow models are examined. A general formulation is shown and issues involved with closure are presented. Microfluidic experiments and lattice-Boltzmann simulations are presented to provide the needed information. Closure relations for capillary pressure and momentum transfer resistance are considered. Finally, we report on missing model components needed to complete the formulation of a next generation of two-phase flow models.

ABSTRACT. We imaged dynamic flow of oil and brine in porous media at steady state by micro-CT, at a resolution of 3.5 µm during the co-injection into Estaillades carbonate. An experimental method based on X-ray differential imaging was used to examine how changes in flow rate impact the pore-scale distribution of fluids during co-injection. We present and quantify the impact of pore-scale heterogeneity on intermittency by comparing the experiment on Estaillades carbonate with that conducted on Bentheimer sandstone.

ABSTRACT. In this work, to study two-phase flow in a fractured rock, we generate a hybrid pore network that represents the fracture, the surrounding matrix and their corresponding connectivity. A dynamic pore network modeling approach is then utilized to simulate various displacement processes in such network by incorporating pore-scale displacement mechanisms. The presented network model is validated against experimental results of a two-phase flow study on a fractured Berea sandstone core used to generate the hybrid network.

ABSTRACT. We characterise drainage, imbibition and trapping in several heterogeneous sandstone rock cores from the mm to m scale. Rich experimental data sets for CO2-brine systems across multiple fractional flows, total flow rates and drainage-imbibition cycles provide validation for numerical simulations. `Digital cores' are characterised with rate invariant, intrinsic multiphase flow properties for both primary drainage and imbibition cycles, allowing accurate prediction of mm-m scale experimental saturations, equivalent relative permeabilities and trapping characteristics, for direct use in field scale modelling efforts.

ABSTRACT. In this paper, we seek to reproduce the development of fractures in an extensional context by capturing some pattern inherited from mechanical processes. The approach combines a 3D stochastic point process generation of disc shaped flaws with a predetermined orientation, and rules for simulating their growth until arrest is prescribed. The pseudo-mechanical status of each element, with regard to neighboring elements is continuously saved, thus allowing further interactions in case a sudden effective stress change is superimposed.

ABSTRACT. We use dfnWorks to model micro fracture flow at centimeter-scale DFNs, and study the effect of micro fractures on diffusive transport. The LTDE experiment served as a motivation to research. Micro-fracture network is combined with continuum volume mesh, where DFN represents the fractured layer of the sample, and the continuum mesh corresponds to matrix, where transport is driven by diffusion only. Simulation results allow to study the effect of micro-fractures presence on tracer and determine the role of advective term.

ABSTRACT. We present a graph based method to identify the primary subnetworks in a three-dimensional fracture networks. The discrete structure of fracture networks naturally lends its self to using graphs to represent and characterize fracture network. The method incorporates topological, geometric, and hydrological properties of the fracture networks into the graphs as edge weights. By finding the edge-disjoint shortest paths through the network we identify subnetworks where the fastest transport occurs without user controlled parameters or running flow and transport simulations.

ABSTRACT. Reactivation of fractures is a strongly coupled problem involving disparate physical processes such as fluid flow, temperature and rock deformation. We present a coupled mixed dimensional model, where the main focus is on the deformation of fractures. We will discuss numerical solutions of friction models for the reactivation of pre-existing fracture networks, and the treatment of the non-linearities introduced by the slip. Numerical examples include a 3D fracture network inspired by realistic data.

ABSTRACT. Joint heat and solute tracer tests allow to add diffusion and conduction information to the solute advection-dispersion and help imaging preferential pathways in heterogeneous aquifers. We perform a joint interpretation of heat and solute tracer tests combining deterministic modeling and Bayesian Evidential Learning. The results show a strong influence of the water viscosity. The stochastic simulations highlight the influence of spatial and parameter uncertainty on the resulting breakthrough curves, stressing the need for realistic uncertainty quantification.

ABSTRACT. We present some of the key results stemming from the development of a comprehensive approach targeting the quantification of environmental risks associated with hydraulic fracturing. Our analyses are framed in the context of the EU Horizon 2020 project FracRisk.

Using exemplary generic risk scenarios during and after fracking operations, we employ a modeling strategy by combining a forward modeling step with a probabilistic characterization of selected target model output based on global sensitivity metrics.

ABSTRACT. The objective of our study is to investigate the “inversion-approach” for the study of the influence of the breach geometry and location to water depth in a bi-dimensional inundated area. This method consists in the definition of a “safety criterion” (“as the water level in a given location ”) and the analysis of all the uncertain input parameters combination permitting to respect the safety objective. With this aim, we used the “Stepwise Uncertainty Reduction” algorithm implemented in the Prométhée workbench.

ABSTRACT. The source of a contaminant is localized with batch-sequential Bayesian optimization approaches. The problem is illustrated with 2D synthetic cases which display sharp transmissivity contrasts and specific connectivity patterns. The resulting objective functions are highly non-linear and display multiple local minima. The contaminant source is identified as the minimum of the objective function, which is localized by derivative-free global optimization algorithms relying on Gaussian Process models and on Expected Improvement criteria.

ABSTRACT. The floodplain hydrology is strongly related to the interactions between the shallow groundwater and the surface water. Recent studies point out the need for combining multiple measurement methods in order to quantify their interactive fluxes. The groundwater heads, together with temperature measurements or hydrochemical analyses, etc., will be performed in the study sites. Further, we will develop a coupled local and catchment-scale groundwater model. The coupled model will enable multi-data inversion and uncertainty quantification within an affordable computational time.

ABSTRACT. Due to land-use change and climate variability, increase of streamflow intermittency and water scarcity are very likely to occur. Most of our knowledge about runoff generation refers to perennial catchments. This study aims to provide a generalizable understanding of the response of intermittent catchments to rainfall variability by analyzing scenarios in an intermittent rural catchment using the Integrated Surface-Subsurface Hydrological Model CATchment HYdrology (CATHY). Outcomes of the study will help detect key rainfall statistics driving streamflow in intermittent catchments.

ABSTRACT. CATHY-Pesticide is a physically based coupled surface-subsurface flow and reactive solute transport model. Advective exchanges between surface and subsurface are managed with a switching boundary condition strategy. A mixing module which represents the solute mobilisation from the top soil to surface runoff has been implemented. CATHY-Pesticide evaluation, based on comparison with data from a vineyard hillslope and on a global sensitivity analysis, highlights its robustness and ability to represent solute transfers at the hillslope scale.

ABSTRACT. Hydrological models typically applied to predict climate change impacts in mountainous areas include only simplified representations of groundwater processes. This is problematic because mountains often present complex geological arrangements that can strongly influence overall catchment functioning. Integrated models can theoretically quantify the role of the subsurface, whilst still simulating surface processes. However, there have been few applications at catchment scale in mountains terrain. Here, we present the development and multi-objective evaluation of an integrated snow/surface-water/groundwater model in a Swiss catchment.

ABSTRACT. We propose a simplified model to describe the flow of a water table and its supply from the surface. This is done for a very wide and thin aquifer. The model consist in coupling two simplified problems describing respectively the quasi-horizontal flow in the water table and the quasi-vertical flow in the ``unsaturated'' part of the aquifer. A particular attention is given to treat the case of an overflowing aquifer. Numerical scheme and simulation are presented.

ABSTRACT. Hyporheic zones can serve many roles throughout the year and may provide a critical control point on watershed scale exports. We apply loose numerical coupling to represent groundwater-surface water connectivity, biogeochemical cycling, and microbial growth in a hyporheic zone test case. Microbial growth is included in the numerical code MIN3P to represent the physically based feedback when pore-spaces are “bioclogged”. A Bayesian MCMC approach is used to constrain the river GPP model to data collected at the East River, Colorado.

ABSTRACT. Understanding ion transport through clays and clay membranes is important for many geochemical and environmental applications. Ion transport is affected by electrostatic forces exerted by charged clay surfaces. Such effects can be modeled by the Donnan approach. Here we introduce a new, comparatively simple way to represent Donnan equilibria in transport simulations, based on including charged surfaces as immobile ions in the balance equation.

ABSTRACT. Particle-tracking algorithms were first developed for conservative solutes. The method is appealing for its speed and accuracy. The method was extended to simple reactions through calculation of particle collision probability and a birth/death process. Real reactions were elusive until the method was extended to variable mass particles and multiple species on each particle. These improvements allow arbitrary reactions on particles, including immobile or solid species. We demonstrate with examples of acid mine drainage and calcite replacement with dolomite.

ABSTRACT. Calcite dissolution is a mass transfer limited process which suggests the importance of simulating transport effects on evolving geometry at a pore scale. In this study we simulated different flow rate conditions with reactions and moving pore wall. The results suggest that flow rate controls the shape of the pore. Fast moving fluid gets less buffered so produces uniform pore shapes as compare to slow moving fluid. These characteristic pore shape yields different relationship between conductivity and pore volume change.

ABSTRACT. The dissolution of a stressed porous media induces compaction caused by mechanical weakening. Pore-scale simulations show that permeability enhancement is suppressed by stress concentration and compaction in the less dissolved (hence stiffer) outlet region. Since this region is also less conductive, its compaction provides a bottleneck effect. This mechanism strongly affects the initial stages of wormholing (high Damkohler), reducing transport heterogeneity and promoting wormhole competition. Near breakthrough, the effect of stress becomes more apparent for more uniform dissolution (low Damkohler).

ABSTRACT. In the present work, we show how large-scale classical molecular dynamics simulations can be used to help interpret the phenomenon of nanoscale diffusion within fine-grained sedimentary material. All-atom simulations containing multiple discrete clay particles are utilized to understand the roles of dry bulk density (porosity) and pore water chemistry (charge-balancing cation and pore water salt composition) on the microstructural, mechanical and transport properties of the overlying clay matrix. Special emphasis is placed on comparing our results with available experimental data.

ABSTRACT. We study the velocity statistics of a pore scale Stokes flow simulation on a digitized Berea sandstone sample. With the use of particle tracking methods, we compute accurately the Lagrangian and Eulerian velocity statistics. After having observed that the velocity process is ergodic and that the velocity statistics become stationary after a few pores, we propose two different Markov models based on velocity distributions which are able to reproduce pore-scale results and to predict higher scales transport and velocity evolution.

ABSTRACT. We present measurements of solute hydrodynamic dispersivity in sandstone and carbonate rock cores by means of unidirectional single- and multi-phase pulse-tracer tests conducted over a range of Péclet numbers (Pe = 20-400) and with the simultaneous imaging of solute transport by 3D Positron Emission Tomography (PET). The latter enables monitoring the spatial and temporal evolution of the full tracer plume non-invasively, thus providing an unprecedented level of observational detail on both spreading and mixing processes in rocks.

ABSTRACT. Reactive transport is of importance for geological, environmental and industrial applications. We review recent developments on the pore-scale simulation of reactive transport on micro-CT images of rocks. The impact of reactive flow on rock petrophysical and mechanical properties is demonstrated. The application of Quantitative Evaluation of Minerals by SCANning electron microscopy (QEMSCAN) for studying reaction processes in rocks containing multiple minerals is demonstrated. The incorporation of mineral particle migration into reaction processes is described.

ABSTRACT. The research deals with the RTM of the aqueous alteration of a block of fractured nuclear glass in the scope of two approaches: discrete and homogenized. The workflow, allowing to pass to homogenised representation is put in place and comprises diverse steps of the characterisation of the fracture network, the upscaling of the major flow/transport and geochemical parameters and ultimately, modeling of the glass leaching by the coupled reactive transport code in conjunction with the geochemical model.

ABSTRACT. Naturally occurring fractures in porous medium are considered joint surfaces that can withstand tectonic stresses due to friction by asperities in the fracture walls. Hydraulic stimulation by the injection of fluids in the fracture network decreases can overcome the frictional resistance, leading to sliding along the fracture surfaces. Here we study how the shearing will alter the local stress fields, and can potentially trigger formation of new fractures by presentation the numerical modeling of fracture propagation in this setting.

ABSTRACT. We derive the relationships that link the general elastic properties of rock masses to the geometrical properties of fracture networks, with a special emphasis to the case of frictional crack surfaces. We extend the well-known elastic solutions for free-slipping cracks to fractures whose plane resistance is defined by an elastic fracture (shear) stiffness k_s and a stick-slip Coulomb threshold. The results were applied to power-law fracture size distributions, which are likely relevant to geological cases.

ABSTRACT. Purely stochastic DFN models, beyond the very limited number of data they are based on, may be irrelevant by not reproducing the fracture-to-fracture interaction, and the resulting network topology. We propose here a new process-based DFN model from simplified mechanical rule dependent of the allegedly known remote stress field and local stress perturbations for nucleation, growth and arrest of fractures, leading to correlated networks.

ABSTRACT. We present progress on Discrete Fracture Network (DFN) flow modeling, including realistic advanced DFN spatial structures and local fracture transmissivity properties, through an application to the Forsmark and Laxemar site in Sweden. In particular, we present new DFN models, where fractures result from a growth process defined by simplified kinematic rules for nucleation, growth and fracture arrest. The differences with DFN models where fracture are randomly distributed in space are very important in terms connectivity, permeability and flow structure (channeling).

ABSTRACT. We address the issue of Uncertainty Quantification analyses for flow simulations in Discrete Fracture Networks with stochastic geometrical parameters. We propose an approach based on Multi Level Monte Carlo ideas used in conjunction with a well assessed solver allowing to circumvent the need for conforming meshes. Robustness and efficiency of the solver represent a key point in the effective application of UQ strategies based on MLMC ideas, as they allow to tackle, with very coarse meshes, complex geometrical configurations.

ABSTRACT. For computationally expensive models, which are common in hydrology, surrogate response surfaces are often employed to increase the number of samples used in approximating the solution of a stochastic inverse problem. We formulate a method for adaptively creating a special class of surrogate response surfaces with stochastic error in mind. Using these two levels of surrogates, the surrogates are adaptively refined. Three types of refinement strategies are presented and combined in an iterative adaptive surrogate construction algorithm.

ABSTRACT. Given its importance in characterizing nearshore and riverine flows, there is great interest in estimation of bathymetry using indirect observations. There have been several efforts to combine data assimilation via ensemble-based Kalman filters with video observations of surface flow properties. Here, we consider two recent low-rank approximation based schemes using both continuous observations taken from stationary platforms as well as snapshot data taken from aerial platforms. To evaluate the methods’ accuracy and efficiency, we compare with established ensemble-based Kalman filters.

ABSTRACT. A stochastic wavelet-based data-driven forecasting framework (WSDDFF) is developed for the purposes of forecasting uncertain water resources processes. The stochastic part of the model explicitly accounts for uncertainty in input data, input variable selection, parameters, and model output; the wavelet part is used to characterize the multi-scale nature of water resources processes. WSDDFF is shown to provide more accurate and reliable forecasts than benchmark (non-wavelet-based) models on a real-world urban water demand forecasting experiment in Montreal, Canada.

ABSTRACT. We apply the multilevel Monte Carlo (MLMC) method combined with the Karhunen-Loève approximation of the input random fields and the maximum entropy method to approximate the density and distribution functions of the concentrations of radioisothopes. The distribution function and quantiles are essential for safety analysis of a deep radioactive waste repository. Coupling of fractures and continuum is used in the transport model. We shall present the methods, developed software, and results from practical application.

ABSTRACT. Parameter specification of distributed hydrological models is challenging whether using automatic calibration procedures or empirical approaches based on relations between parameters and spatial observations. A novel approach relying on hydrological signatures, i.e. characteristic indices derived from data that can be related to hydrological processes and the model parameters is presented. When evaluating a model, the advantage of comparing signatures is that discrepancies between simulations and observations are diagnostically meaningful. Results are illustrated with the J2K model on a 2250km² catchment.

ABSTRACT. The advent of high-resolution measurements of topographic and (vertical) vegetation features using areal LiDAR are enabling us to resolve micro-scale (~1m) landscape structural characteristics over large areas. Availability of hyperspectral measurements is further augmenting these LiDAR data by enabling the biogeochemical characterization of vegetation and soils at unprecedented spatial resolutions (~1-10m). Such data have opened up novel opportunities for modeling Critical Zone processes and exploring questions that were not possible before. We show how an integrated 3-D model at ~1m resolution can enable us to resolve micro-topographic and ecological dynamics and their control on hydrologic and biogeochemical processes. We address the computational challenge of such detailed modeling by exploiting hybrid CPU and GPU computing technologies. We show results of moisture, biogeochemical, and vegetation dynamics from studies in the Critical Zone Observatory for Intensively managed Landscapes (IMLCZO) in the Midwestern United States. http://hydrocomplexity.net/

ABSTRACT. The model QUALity-NETwork is a deterministic high-resolution model able to simulate hydrological and biogeochemical processes in drainage networks at the regional scale, with water temperature explicitly determined. Hourly variations computed on a sub-catchment of the Loire River (France) prone to eutrophication helped disentangle the complex interactions existing between hydrological and biological processes during low-flows, such as phytoplankton limitation by phosphorus availability, large phosphorus recycling through the microbial loop, or nutrient flush during storm events.

ABSTRACT. We investigate point source nutrient discharges from urban wastewater treatment plants (UWWTPs) in Germany to understand their location and distributions along river networks. In addition, we explore spatial patterns of specific nutrient loadings discharged from UWWTPs to river networks. Our research findings were used to develop a parsimonious model for nutrient spiraling and the implication of spatial patterns of algal blooms. Our data-modeling analyses can be used to improve assess the ecological vulnerability of river networks at catchment scales.

ABSTRACT. Over the past three decades, our understanding of drivers and processes controlling transport, accumulation, and transformations along river corridors has significantly matured. It is clear now that “rivers are not pipes,” and that many of the key reactions controlling water quality take place in exchange zones where riverine water is in close contact with geochemically and microbially-active sediments. This talk presents a view of the past, present, and future of river corridor science.

ABSTRACT. We compare observed concentration-discharge relationships of nutrients in German and US catchments with results of an explorative stochastic model. The observational data exhibit archetypical dilution, enrichment, and constant C-Q patterns. Our model indicates that the dominant driver of emergent C-Q patterns was structured heterogeneity of solute sources. We conclude that efforts to improve stream water quality and ecological integrity in intensely managed catchments should lead away from landscape homogenization by introducing structured source heterogeneity.

ABSTRACT. The PESHMELBA model represents water and pesticide transfer and fate at the catchment scale and takes explicitly into account the influence of each landscape feature including surface elements as plots and discontinuities (ditches, hedgerows…). It is based on separated modules with different levels of complexity, standing for each process or ensemble of processes occuring at the local scale of one element. All modelling units are gathered and coupled thanks to the OpenPALM coupler in order to reach the catchment scale.

ABSTRACT. Increasing anthropogenic phosphorus (P) delivery to surface freshwaters is a major cause of eutrophication. To design efficient mitigation strategies, it is important to improve our understanding of large-scale interactions between human activities, hydrology, and processes affecting P cycling during its transit from land to ocean. Therefore, we implemented a mechanistic description of P transfers and transformations in river networks into a global hydro-biogeochemical modeling framework. We focus on the fate of dissolved inorganic P, most likely to affect algal proliferation.

ABSTRACT. We investigate both experimentally and theoretically the properties of A+B$\rightarrow$C reaction fronts during a radial injection of A into B at a constant flow rate $Q$. We show that, in such a radial geometry, the total amount of product varies as $Q^{-1/2}$ for a given volume of injected reactant and the front position as $Q^{1/2}$ for a given time, paving the way to a flow control of the amount and spatial distribution of the reaction front product.

ABSTRACT. Conventional reactive transport modelling approaches do not consider mixing-dependent reaction rates. We quantified the impact of pore-scale mixing on the biodegradation rate from batch experiments, and analyzed its impact on the field-scale fate and transport of dissolved organic carbon (DOC) from simulations. The results show that different mixing conditions can lead to significant differences in the effectiveness of the biodegradation, and that the effects of mixing can represent a dominant source of uncertainty in model predictions.

ABSTRACT. In this work we show the effect of anisotropy structure on plume entropy and reactive mixing in helical flows using conservative and reactive transport simulations. We identify optimal geometric configurations maximizing mixing and reactions, and yielding enhancement factors up to 15 times the outcomes of analogous simulations in homogeneous media. Compound-specific diffusive/dispersive properties were still relevant despite the enhanced plume dilution in helical flows with important consequences for reactive mixing.

ABSTRACT. This study aims to advance the understanding of the interfacial phenomena and instability in density-driven convection with dissolution in porous media. Pore-scale simulation, linear stability analysis and laboratory experiment are conducted in the current approach. The results demonstrate that interfacial instability is triggered by the density ratio between two miscible fluids, suppressed by the heterogeneous surface reaction between the fluid and solid phases, and influenced by the porosity (permeability) evolution due to the dissolution.

ABSTRACT. We consider pore-scale models for reactive flow and transport of charged species in a thin strip. Reactions leading to a non-negligible deposited layer on the boundary of the strip, i.e. the moving interface between the fluid and the deposited (solid) layer is explicitly taken into account. We derive an upscaled model by averaging in the transversal direction. We focus on the cases, where the convective and/or electroosmotic transport dominate the diffusive transport. Finally, numerical computations are complementing the theoretical results.

ABSTRACT. Unsupervised Machine Learning (ML) is a powerful technique for analyses of reactive transport data observed at contamination sites as well as model simulations representing complex biogeochemical reactions. We present an application of a Sparse Non-negative Tensor Factorization ML to reveal the temporal and spatial features in reactants/product concentrations. The field data represent fate and migration of chromium contamination in a regional aquifer at the LANL site. The simulations are based on reactive-diffusion model representing complex mixing processes.

ABSTRACT. Reactive subsurface flows are heavily influenced by advective spreading due to flow field heterogeneities. We present a new particle-based model, that relies on stochastic velocity processes similar to CTRW models. Unlike existing approaches, however, we analytically derive our model from first-order perturbation theory and subsequently generalize it for highly heterogeneous formations. The model thus provides a macrodispersion parametrization that is formally consistent with classical results, but also remains accurate for highly heterogeneous formations displaying strong non-Fickian behavior.

ABSTRACT. Dissolution-driven convection in partially miscible systems has regained much interest in the context of CO2 sequestration. We show, both experimentally and numerically, that chemical reactions can accelerate or slow down the development of density-driven convection. Furthermore, they can enhance the steady-state flux of dissolving CO2 as they consume it and can induce more intense convection than in the non-reactive case. This result is important to quantify the storage rate of CO2 dissolving into the host oil or aqueous phase.

ABSTRACT. Creeping non-newtonian flows in porous media are common, for instance polymer-improved aquifer remediation techniques. Which macro-scale equation is relevant? Most engineering macro-scale models feature Darcy's law corrected with an apparent viscosity or permeability. Transitions in rheological behavior, anisotropy induced by rheology may render these approaches irrelevant. We solved numerically the flow problem in various real and artificial porous structures and results show that an apparent viscosity may be sufficient only in the case of isotropic and sufficiently disordered porous media.

ABSTRACT. During a rain event, water infiltrates into the ground where it flows slowly towards a river. The time scale and the geometry of this flow control the chemical composition and the discharge of the river. With a laboratory aquifer and a combination of complex analysis and numerical methods we try to understand the impact of the depth of the aquifer on the discharge of the river.

ABSTRACT. We developed a numerical method to solve the coupled water and heat transfer in variably-saturated freezing soils. The proposed model well reproduced the hydrothermal dynamics of frozen soils over Tibetan Plateau. The comparison of different soil thermal conductivity parameterizations indicated that de Vries’s parameterization performed better than other methods. The analysis of water/vapor fluxes assured the phenomenon of the upward liquid water fluxes towards the freezing front and furthermore highlighted the crucial role of vapor fluxes during soil freezing/thawing cycles.

ABSTRACT. Harmonic averaging of the coefficient k(p) has been blamed for nonphysical locking and lagging of pressure solutions of the Generalized Porous Medium Equation. This degenerate nonlinear parabolic equation has applications to plasma heat transfer, gas and groundwater flow. The numerical issues also manifest themselves in spurious temporal oscillations, even with none in space. For the continuous coefficient problem, we use a Modified Equation Analysis approach. A shock tracking and level set approach is utilized for the discontinuous case.

ABSTRACT. We present results from three-dimensional fully resolved simulations of the transition to turbulence in fluids injected in fracture apertures or near rough walls. We consider wall-roughness with different amplitudes, in both pipes and fracture apertures, and see how the critical flow rate, at which the fluid becomes turbulent, changes with the roughness characteristics. In general, we observe a non-trivial decrease in the point of transition with the roughness amplitude. Finally, we shall comment on possible geometries which might suppress turbulence.

ABSTRACT. We investigate local aspects and heterogeneities of porous medium morphology on the inertial flow, and relate them to the relevant mechanisms of momentum transfer. The objective of the work is to explore the effects of randomness in the morphological structure of porous media on inertial and viscous forces which control flow behaviour in porous media.

ABSTRACT. We study advective particle motion in three-dimensional fracture networks characterized by power-law fracture length and aperture distributions. We investigate (i) the relation between fracture geometry, aperture distribution and the flow velocity, (ii) the statistics of particle velocities and (iii) their impact on observed large scale transport in terms of particle breakthrough curves at different distances from the inlet. We model Lagrangian velocities as Markov chains, which serves as the basis for a predictive time-domain random walk model.

ABSTRACT. The permeability tensor of a fractured rock is computed numerically for a set of geomechanically realistic three-dimensional discrete fracture patterns. These patterns are generated with a finite element-based discrete fracture propagation simulator, in which deformation and flow are numerically computed. These detailed multi-fracture growth simulations study permeability as a function of the interaction of fractures and the mechanical effects of pattern evolution on the distribution of apertures in response to in situ stresses.

ABSTRACT. An NMM model was developed for fully coupled analysis of THM processes in porous rock with discrete fractures using non-conforming mesh. Fluid flow is simulated by a zero-dimensional model. Mechanically, fractures are treated as discontinuities with open, closed or sliding states. The THM couplings including direct (pore-volume effect, convective heat flux and thermal strain) and indirect couplings through material properties are simultaneously solved. The model was verified step-by-step and applied for geothermal reservoir stimulations.

ABSTRACT. Analysis of flow and transport considering fracture networks with internal heterogeneity described by different multivariate normal distributions is presented. A coherent triad of fields with conserved statistical properties but which greatly differ in connectivity structure are considered. By numerical modelling of multiple scales in a stochastic setting the relative impact of texture type and correlation length is quantified against network topological measures, and thereby key thresholds for cases where flow dispersion is controlled by internal heterogeneity are identified.

ABSTRACT. The generation of quality meshes of Discrete Fracture Networks is still challenging, especially for dense and large-scale models. We will present a methodology for this purpose, even in the case of complex geometries. This methodology has been implemented in BLSURF_FRAC software, which provides the user with a choice of planar meshers. Application examples will be shown, with flow simulations computed on meshes generated by two different planar meshers, BL2D and BAMG.

ABSTRACT. Integrating advanced simulation techniques and data analysis tools in GIS may provide contribution to management of conjunctive use of ground- and surface-water. The FREEWAT platform is a free and open source interface, QGIS-integrated, for simulating the hydrologic cycle, with specific attention to groundwater, coupling the power of GIS geo-processing and post-processing tools in spatial data analysis with that of process-based simulation models. Modules integrated within the FREEWAT platform and examples of real case study applications are provided.

ABSTRACT. Computational modeling of wave and current interaction with structures is becoming a reality for engineering analyses traditionally conducted through physical testing. While advances in computational methods have been of primary importance in achieving a robust, accurate, and efficient modeling capability, better algorithms alone are insufficient for realizing the full impact of computational fluid-structure interaction analyses. In this presentation we will present our recent experience attempting to overcome several software engineering barriers through domain-specific extension modules and domain specific languages.

ABSTRACT. Environmental issues on water resources and water quality are challenging multidisciplinary questions. We argue that progresses in software architecture can provide a shift in the development of hydrological model for simulation, prediction and use of existing information. In this paper, we present a web-based domain-specific environment for designing simulation processes, editing simulation codes, and automatically deploying them over distributed containers.

ABSTRACT. There is an increasing need in environmental science to address so-called collaborative research complexity, that of bringing together: multiscale simulations and observations; a wide range of software tools; teams with different levels of technical skills; researchers from different science disciplines. This paper discusses virtual data labs and their potential to overcome this complexity, seeking approaches that support real innovation and that embrace new data science methods. The paper culminates in a series of ongoing research challenges in virtual data labs.

ABSTRACT. The Adaptive Hydraulics (AdH) Suite is a U.S Department of Defense software package for solving the shallow water, transport and groundwater equations. Recently, a decision was made to refactor the software and use modern software engineering concepts to streamline future additions and error trapping. This presentation will focus on the hierarchical data structure implemented in AdH to separate various aspects of the software and allow developers to concentrate on their areas of interest.

ABSTRACT. In this talk, we will present a vision that promotes a unique approach combining engineering and scientific models to enable informed decision on the basis of open and scientific knowledge, a broader engagement of society for addressing sustainability concerns, and incorporate those decisions in the control loop of smart CPS. We will introduce a research roadmap to support this vision that emphasizes the socio-technical benefits of modeling, especially in computational sciences.