Abstract. Watershed characteristics create a mosaic of potential storage zones, and linkages between these result in hydrologic connectivity between groundwaters and surface waters. Watershed storage and connectivity are fundamental controls on stream network dynamics, yet predicting the patterns of hydrologic responses like streamflow remains difficult due to landscape heterogeneity across multiple, interacting spatial scales. Here, we characterize storage and connectivity dynamics at the watershed scale and the hydrogeomorphic feature (HGF) scale using a network of steam and riparian groundwater monitoring wells in three watersheds spanning the Coastal Plain, Piedmont, and Appalachian Plateaus physiographic regions of the southeastern USA. At the watershed scale, we operationalized watershed storage as the slope of streamflow recession at the watershed outlet, and hydrologic connectivity as the relationship between outlet discharge and network length. At the HGF scale, we operationalized storage as the slope of water level recession at seven in-stream monitoring locations, and connectivity as the lateral hydraulic gradient and water level hysteresis relationships between the stream channel and adjacent riparian zone at three locations per watershed. At the watershed scale, we found patterns in storage and connectivity varied across physiographic regions. We also found distinct storage and connectivity dynamics across HGFs, many of which challenged existing conceptualizations for similar geomorphic settings. Notably, channel incision emerged as a key structural control, homogenizing storage and connectivity patterns across otherwise disparate watersheds and producing a common threshold response in watershed storage. Together, these results demonstrate that watershed-scale patterns do not reflect the heterogeneity observed at the HGF scale. Moreover, river corridor structure, particularly channel incision, is a key driver of storage and connectivity dynamics across physiographic settings.

Water is essential for life on Earth, supporting ecosystems, human health, and economic activities. Hydrology relies on observational data, and this paper discusses regional and national datasets for the conterminous United States (CONUS) publicly available as of 2023, focusing on headwaters, defined as first- and second-order streams at 1:24000 scale. It identifies 72 primary and secondary datasets and 11 repositories and argues how better integration and accessibility of hydrological data can improve research. The paper distinguishes between datasets where streamflow was the primary data collection objective and those where it was secondary. This distinction highlights opportunities to consider data from efforts peripheral to hydrology but is still useful for understanding hydrologic conditions. The analysis reveals that out of about 118 000 active and inactive stream observation sites, about 6.6% and 25% are located on first- and second-order streams, respectively. This indicates a substantial data gap for headwater systems, which account for over 77% of stream length in CONUS. Federal agencies manage 72% of hydrologic monitoring sites across all stream orders, but only 34% of these are in headwater systems. Academic institutions operate about 2% of sites, with almost half (48%) in headwater systems, focusing on ecosystem research. State agencies also operate about 2% of sites, primarily on larger systems, with 19% on headwaters. Additionally, 23% of sites are managed by multiple agencies. Spatial patterns further reveal pronounced disparities among physiographic regions. Eastern and coastal provinces show relatively dense monitoring, while central and western regions show sparse coverage. These gaps reflect historical priorities, logistical constraints, funding limitations, and the high cost of continuous instrumentation. To address biases in monitoring networks, data collection could be enhanced with low-cost monitoring, community science, and remote sensing technologies. This study also notes the benefits of long-term monitoring and prioritizing retention of streamgages with longer records.

Headwater streams play critical roles in hydrologic and biogeochemical processes and functions, yet their spatial distribution and land cover context remain poorly understood at continental scales, and no dedicated geospatial dataset exists. Building from a high-resolution conterminous United States (CONUS) hydrography network dataset, we quantified the spatial extent, density, and upstream catchment characteristics of headwater stream segments across the CONUS. We identified approximately 8.4 million kilometers of headwater streams, finding that 77

Abstract The groundwater—surface‐water interface is an important regulator of many biogeochemical processes along river corridors. One of the main drivers of these processes is the pore water travel time along hyporheic flow paths. However, our understanding and ability to predict the spatial‐temporal dynamics of subsurface travel times are limited because current techniques are restricted to a few locations or constant subsurface flows. To overcome these limitations, we designed and field‐tested a small, easy‐to‐build, low‐cost pore water conductance sensor to monitor subsurface travel times in real‐time. Leveraging the simple design and low cost, we built and installed 93 sensors as profiles (here measuring up to 20 cm depth) in a first‐order stream and conducted repeated tracer injections during baseflow conditions and a storm event. We quantified the dynamics of pore water travel times at unprecedented high vertical (cm‐scale), horizontal (dm‐scale) and temporal (minutes) resolution during a period of almost 2 months, moving beyond the usual snapshot view of these processes. We observed small‐scale deviations from the general flow field along a riffle and inconsistent patterns of pore water travel times during the storm event. Measured travel times, combined with targeted pore water sampling of reactive or conservative solutes, allows calculation of solute transformation rates at a high spatial resolution and extent in future sensor applications. These insights could help to understand the relative contribution of the different drivers (e.g., residence time, redox conditions, substrate availability) controlling biogeochemical processes at the groundwater—surface‐water interface and the impact of these drivers on the metabolism and solute fluxes along river corridors.

Slug tracer experiments have greatly advanced our understanding of solute transport in streams. Breakthrough curves (BTCs) from these experiments are biased toward faster flow paths, highlighting the need for alternative tracers to cover longer timescales. The radioactive tracer radon (222Rn) is increasingly used to quantify transit times in subsurface transient storage zones, capturing durations of up to 21 d. However, it remains unclear whether calibrating transient storage models (TSMs) with radon yields longer subsurface timescales of transit times than calibrating them with slug tracers such as sodium chloride (NaCl). To address this, we conducted radon measurements and NaCl slug tracer experiments in Oak Creek (Oregon, USA) and jointly and individually calibrated TSM parameters with both tracers. We applied parameter identifiability analysis and used information theory to evaluate how the two tracers constrain model parameters. Our results show that TSM calibration with both radon and chloride increases parameter information compared to TSM calibration with either tracer alone. This suggests that incorporating radon into calibration improves estimates of solute transport in future studies. However, when calibrating the TSM with only radon measurements, all resulting parameters of the TSM were non-identifiable. This non-identifiability arises because radon activity in streams remains at steady-state and is highly sensitive to the location and amount of groundwater inflow, as well as contributions of flow paths from subsurface transient storage zones. As a result, radon measurements are biased toward longer-timescale flow paths, limiting their applicability to uniquely constrain solute transport parameters in TSM calibration without complementary slug tracers.

We discuss the impact of design and operational parameters on the performance of direct air capture units using solid sorbents by deploying a state-of-the-art process simulation tool combined to black-box optimization. We identity that system designs employing moderate CO2 sorption kinetics and contactors with low length-to-radius ratios yield the best performance in terms of system productivity.

The benthic biolayer is a shallow zone of reactive streambed sediments, widely believed to contribute disproportionately to whole-stream reactions such as aerobic respiration and contaminant transformation. Quantifying the relative contribution of the biolayer to whole-stream reactions remains challenging because it requires that hyporheic zone solute transport and reaction heterogeneity are explicitly captured within a single modeling framework. Here, we use field experiments and modeling to quantify the biolayer's aerobic reactivity relative to other stream compartments. We co-injected and monitored several fluorescent tracers, including the reactive tracer resazurin, into a controlled experimental stream. We characterized reactive transport in the water column and at multiple depths in the hyporheic zone by fitting all data to a new mobile-immobile model, using resazurin-to-resorufin conversion as an indicator of aerobic bioreactivity. Results show that the biolayer converted 8 times more resazurin to resorufin than all other stream compartments, and 80% of this conversion occurred within 2 reach advection times. This hotspot and hot moment behavior is attributed to the biolayer's ability to rapidly acquire, transiently retain, and rapidly degrade stream-borne solutes. The model analysis shows that the majority of raz-to-rru conversion occurs in the biolayer across streams with a wide range of biolayer structural properties, including streams with a biolayer that is less reactive than deeper regions of the hyporheic zone. Together, our results show that the biolayer is a common feature of streams and rivers that should be considered in network-scale models of aerobic reactivity.

We present the design and performance evaluation of a multi-sorbent process for CO2 capture based on vacuum-swing adsorption by studying layered-bed and mixed-bed configurations. We find that the multi-sorbent process achieves improved separation effectiveness and yields a reduced energy usage compared to single-adsorbent processes.

The exchange of stream water and groundwater (hyporheic exchange) plays an important role in hydrological and biogeochemical processes in rivers. Hyporheic flow comprises a distribution of subsurface flow paths characterized by distinct transit times and flow path lengths. Much of the previous research relied on the interpretation of slug tracer experiments that only capture a portion of the overall hyporheic exchange, given their relatively short (minutes to hours) duration. Therefore, there is a need to go beyond the characterization of shorter flow paths in hyporheic research to understand flow paths of the entire transit time distribution. We hypothesize that environmental tracers provide complementary information into longer hyporheic flow paths. Here we derive and compare commonly used transport metrics for hyporheic exchange derived from radon and slug tracer injections and aim to identify combinations of model parameters that predict the concentrations of both tracers along experimental stream reaches. For this purpose, we measured the environmental tracer radon (222Rn), that increases with time along hyporheic flow paths, at several stream sections along Oak Creek, Oregon (USA). We conducted slug tracer (sodium chloride) injections at the same stream sections. We employed a transient storage model that includes radon specific processes such as radioactive decay to ensure comparability in the information acquired on hyporheic exchange from radon and the typically applied slug tracer experiments. We calibrated final stream discharge and hyporheic exchange metrics through a global identifiability analysis and subsequently calculated relevant transport metrics using the refined parameters. Results with the field data will be obtained soon. Hence, this study will contribute to a more holistic understanding of hyporheic flow paths and related processes, such as the biogeochemical turnover processes in rivers.

Pressure-Vacuum Swing Adsorption (PVSA) shows great potential in post-combustion carbon capture. However, accurately modelling it requires a large amount of computational time using detailed process models, and screening large numbers of adsorbents becomes computationally prohibitive. Using data-driven neural network models have great potential to solve this problem and is the focus of this paper. We build and adapt the previously established machine-assisted adsorption process learning and emulation (MAPLE) framework by using the dual-site Langmuir model, expanding the number of features used, and predicting capture costs, to demonstrate the benefits of using a neural network to make fast and accurate assessments of a PVSA process. A detailed mathematical process model was used to generate training data for our neural networks and a case study was performed to compare the performance of our neural network with that of the detailed model. Our results indicate that our neural networks have performance comparable to that of the detailed model with acceptable levels of uncertainties up to around 14% whilst requiring up to 25,200x less computational time. This vast reduction in computational time shows the great potential of this tool in solving process optimisation problems compared to when using traditional process modelling.

The protection of headwater streams faces increasing challenges, exemplified by limited global recognition of headwater contributions to watershed resiliency and a recent US Supreme Court decision limiting federal safeguards. Despite accounting for 77

Hyporheic exchange is critical to river corridor biogeochemistry, but decameter‐scale flowpaths (∼10‐m long) are understudied due to logistical challenges (e.g., sampling at depth, multi‐day transit times). Some studies suggest that decameter‐scale flowpaths should have initial hot spots followed by transport‐limited conditions, whereas others suggest steady reaction rates and secondary reactions that could make decameter‐scale flowpaths important and unique. We investigated biogeochemistry along a 12‐m hyporheic mesocosm that allowed for controlled testing of seasonal and spatial water quality changes along a flowpath with fixed geometry and constant flow rate. Water quality profiles of oxygen, carbon, and nitrogen were measured at 1‐m intervals along the mesocosm over multiple seasons. The first 6 m of the mesocosm were always oxic and a net nitrogen source to mobile porewater. In winter, oxic conditions persisted to 12 m, whereas the second half of the flowpath became anoxic and a net nitrogen sink in summer. No reactive hot spots were observed in the first meter of the mesocosm. Instead, most reactions were zeroth‐order over 12 m and 54 hr of transit time. Influent chemistry had less impact on hyporheic biogeochemistry than expected due to large amounts of in situ reactant sources compared to stream‐derived reactant sources. Sorbed or buried carbon likely fueled reactions with rates controlled by temperature and redox conditions. Each reactant showed different hyporheic Damköhler numbers, challenging the characterization of flowpaths being intrinsically reaction‐ or transport‐limited. Future research should explore the prevalence and biogeochemical contributions of decameter‐scale flowpaths in diverse field settings.

ABSTRACTFloodplains along low‐gradient, meandering river systems contain diverse hydrogeomorphic features, ranging from isolated depressions to hydrologically‐connected channels. These ephemerally‐flooded features inundate prior to river water overtopping all banks, enhancing river‐floodplain connectivity during moderately high flow stages. Predicting when and where ecological functions occur in floodplains requires understanding the dynamic hydrologic processes of hydrogeomorphic features, including inundation and exchange. In this study, we examined storm event‐scale inundation and exchange dynamics along a lowland, meandering river system in central Illinois (USA). We monitored surface water presence/absence, surface water level, and groundwater level across floodplain hydrogeomorphic feature types (i.e., isolated depression, backwater channel, and flow‐through channel). Using these data, we evaluated inundation onset and recession characteristics, drivers of groundwater‐surface water interactions, and direction of hydrologic exchange with the river channel. Surface water presence/absence patterns suggested inundation onset timescales were primarily controlled by microtopography and recession timescales were correlated with floodplain elevation. Employing a novel hysteresis approach for characterising groundwater‐surface water interactions, we observed distinct patterns indicating differences in water sources across hydrogeomorphic units and event characteristics. Finally, differences in hydraulic head along floodplain channels revealed that channels with multiple inlets/outlets (i.e., flow‐through channels) conveyed down‐valley flow and channels with single inlets primarily functioned as sinks of river‐derived water to the floodplain with short source periods. These results highlight the heterogeneity of hydrologic processes that occur along lowland, meandering river‐floodplains, and more specifically, point to the important role hydrogeomorphic features play in controlling dynamic connectivity within the river corridor.

Download This Paper Open PDF in Browser Add Paper to My Library Share: Permalink Using these links will ensure access to this page indefinitely Copy URL Simulating Long-Term Dynamics of Solute Transport in a Mountainous Stream Network Using a Multiscale Transport Model Informed by Machine Learning 37 Pages Posted: 24 Feb 2024 See all articles by Phong LePhong LeGovernment of the United States of America - Oak Ridge National LaboratorySaubhagya RathoreGovernment of the United States of America - Oak Ridge National LaboratoryEthan T. CoonGovernment of the United States of America - Oak Ridge National LaboratoryAdam WardOregon State UniversityRoy Haggertyaffiliation not provided to SSRNScott PainterGovernment of the United States of America - Oak Ridge National Laboratory Abstract The movement of solutes in a watershed is a complex process with multiple interactions and feedbacks across spatial and temporal scales. Understanding how solute moves along diverse hydrologic pathways through watersheds – from hillslopes to stream channels and in and out of the hyporheic zones – is challenging but critically important since these processes integrate and contribute to the biogeochemical functioning of the river corridor up to the river network scale. Here we use results from a long-term network-scale tracer test at the H.J. Andrews experimental forest in western Cascade Mountains, Oregon, USA to inform a multiscale model for transport in stream corridors. The model uses a Lagrangian-based subgrid model to represent the effects of hyporheic exchange flow and advective transport at stream network scales. The spatially and temporally resolved stream discharge needed for the transport model was imputed across the river system by an entity-aware long short-term memory network. Modeled concentrations show good agreements with the observations and exhibit power scaling laws indicative of a very wide range of timescales over which hyporheic exchange flow occurs. Our results demonstrate a data-informed modeling framework that links dynamical processes occurring at small scales to a network context to help understand how changes at reach scale cascade into network-scale effects, providing a useful tool for sustainable river basin management. Keywords: modeling, reactive transport, multiscale, ATS, LSTM, network Suggested Citation: Suggested Citation Le, Phong and Rathore, Saubhagya and Coon, Ethan T. and Ward, Adam and Haggerty, Roy and Painter, Scott, Simulating Long-Term Dynamics of Solute Transport in a Mountainous Stream Network Using a Multiscale Transport Model Informed by Machine Learning. Available at SSRN: https://ssrn.com/abstract=4737922 Phong Le (Contact Author) Government of the United States of America - Oak Ridge National Laboratory ( email ) 1 Bethel Valley Road, P.O. Box 2008, Mail Stop 608Room B-106, Building 5700Oak Ridge, TN 37831United States Saubhagya Rathore Government of the United States of America - Oak Ridge National Laboratory ( email ) 1 Bethel Valley Road, P.O. Box 2008, Mail Stop 608Room B-106, Building 5700Oak Ridge, TN 37831United States Ethan T. Coon Government of the United States of America - Oak Ridge National Laboratory ( email ) 1 Bethel Valley Road, P.O. Box 2008, Mail Stop 608Room B-106, Building 5700Oak Ridge, TN 37831United States Adam Ward Oregon State University ( email ) Bexell Hall 200Corvallis, OR 97331United States Roy Haggerty affiliation not provided to SSRN ( email ) No Address Available Scott Painter Government of the United States of America - Oak Ridge National Laboratory ( email ) 1 Bethel Valley Road, P.O. Box 2008, Mail Stop 608Room B-106, Building 5700Oak Ridge, TN 37831United States Download This Paper Open PDF in Browser Do you have negative results from your research you’d like to share? Submit Negative Results Paper statistics Downloads 0 Abstract Views 8 71 References PlumX Metrics Feedback Feedback to SSRN Feedback (required) Email (required) Submit If you need immediate assistance, call 877-SSRNHelp (877 777 6435) in the United States, or +1 212 448 2500 outside of the United States, 8:30AM to 6:00PM U.S. Eastern, Monday - Friday.

We present the design and performance evaluation of a novel multisorbent process for CO2/N2 separation based on vacuum-swing adsorption (VSA). We study two process configurations: (i) layered-bed processes, wherein two distinct adsorbent materials are arranged in sequential layers within the adsorption bed, and (ii) mixed-bed processes, wherein two distinct adsorbent materials are homogeneously mixed within the adsorption bed. We develop, validate, and deploy a high-fidelity dynamic adsorption column model for the multisorbent process configurations and apply Bayesian optimization to design processes that achieve maximum separation effectiveness in terms of CO2 purity and recovery with an application to postcombustion carbon capture (PCC) on a coal-fired power plant. We find that the multisorbent process configurations achieve improved CO2/N2 separation effectiveness compared to benchmark classical single-adsorbent processes, increasing the CO2 recovery by up to 5% while achieving high CO2 purity. When operating in compliance with widely adopted performance targets for PCC (PuCO2 >= 95%, ReCO2 >= 90%), we find that the multisorbent process configurations reduce the energy usage of the separation by approximately 35%. We use the modeling framework to analyze the subcolumn scale adsorption dynamics and identify that the observed improvements in performance are associated with the positioning of the CO2 adsorption front under optimized operating conditions, leading to favorable dynamic interactions with the operation of the VSA process cycle.

Rivers receive substantial dissolved organic matter (DOM) input from the surrounding land which is transported to the ocean. As this DOM travels through watersheds, it undergoes biotic and abiotic transformations which impact biogeochemical cycles and release CO2 into the atmosphere. While recent research has increased our mechanistic knowledge of DOM development within watersheds, DOM development across broad spatial distances and within divergent biomes is under investigated. Here, we combined DOM characterization, geochemical analyses, and shotgun metagenomics to analyze samples from seven rivers ranging from the U.S. Pacific Northwest to Berlin, Germany. Initial analyses revealed that many DOM properties separated based upon river type (e.g., wastewater, headwater), though paired geochemical analyses indicated that geochemistry often explained some variation. Analyses rooted in meta-metabolome ecology indicated that, at the global scale, DOM was structured overwhelmingly by deterministic selection. When controlling for scale, however, analyses indicated that ecological assembly dynamics were again structure, in part, by river type. Finally, microbial analyses revealed that many riverine microbes from our systems shared core metabolic functional potential while differing in peripheral capabilities in spatially resolved patterns. Further analyses in the carbon degradation potentials of the recovered metagenomically assembled genomes indicated that the sample rivers had strong taxonomically conserved niche differentiation regarding carbon degradation and that the diversity in carbon degradation potential was significantly related to organic matter diversity. Together, these results help us uncover interconnections between the development of DOM, riverine geochemistry, and microbial functional potential. ### Competing Interest Statement The authors have declared no competing interest.

Transitions between dry and wet hydrologic states are the defining characteristic of non-perennial rivers and streams, which constitute the majority of the global river network. Although past work has focused on stream drying characteristics, there has been less focus on how hydrology, ecology and biogeochemistry respond and interact during stream wetting. Wetting mechanisms are highly variable and can range from dramatic floods and debris flows to gradual saturation by upwelling groundwater. This variation in wetting affects ecological and biogeochemical functions, including nutrient processing, sediment transport and the assembly of biotic communities. Here we synthesize evidence describing the hydrological mechanisms underpinning different types of wetting regimes, the associated biogeochemical and organismal responses, and the potential scientific and management implications for downstream ecosystems. This combined multidisciplinary understanding of wetting dynamics in non-perennial streams will be key to predicting and managing for the effects of climate change on non-perennial ecosystems. This Perspective presents a wetting regime framework that is classified by dominant hydrologic mechanisms and highlights the resulting responses of stream biogeochemistry and community ecology.

The movement of solutes in a watershed is a complex process with multiple interactions and feedbacks across spatial and temporal scales. Modeling the dynamics of solute transport along diverse hydrologic pathways within watersheds - from hillslopes to stream channels and in and out of the hyporheic zones - is challenging but critically important, as these processes integrate and contribute to the biogeochemical functioning of the river corridor up to the river network scale. Here we use results from a long-term network-scale tracer test at the H.J. Andrews experimental forest in western Cascade Mountains, Oregon, USA to inform a multiscale framework for transport in stream corridors. The framework uses a Lagrangian-based subgrid model to represent the effects of hyporheic exchange flow and advective transport at stream network scales. The spatially and temporally resolved stream discharge needed for the transport model is imputed across the river system by an entity-aware long short-term memory network. Modeled concentrations show good agreements with the observations and exhibit power scaling laws indicative of a very wide range of timescales over which hyporheic exchange flow occurs. Our results demonstrate a data-informed modeling framework that links dynamical processes occurring at small scales to a network context to help understand how changes at reach scale cascade into network-scale effects, providing a useful tool for sustainable river basin management.