Evaluate the impact of hydrodynamic pressure on hydrologic exchange fluxes and resident time for a large-scale river section over a long-term period

Quantifying hydrologic exchange fluxes (HEF) and subsurface water residence times (RT) are important for managing the water quality and ecosystem health in dynamic river corridor systems. Laboratory-scale experiments and models have shown that hydrodynamic pressure variations on the riverbed induced by dynamic river flows can strongly impact HEFs and RTs. In this study, the impacts of hydrodynamic pressure on HEFs and RT for a 30 km section of the Columbia River in Washington State over a three-year period were quantitatively evaluated using a coupled transient three-dimensional (3D) multi-phase surface and subsurface water flow transport approach. Based on comparisons between model simulations with and without considering hydrodynamic pressure, we found that hydrodynamic pressure increase the net HEFs by 7% of flowing into river from subsurface domain, and also leads to a reduction of the area with long RT, and increase of area with short RT.

Machine Learning Analysis of Hydrologic Exchange Flows and Transit Time Distributions in a Large Regulated River

Hydrologic exchange between river channels and adjacent subsurface environments is a key process that influences water quality and ecosystem function in river corridors. High-resolution numerical models were often used to resolve the spatial and temporal variations of exchange flows, which are computationally expensive. In this study, we adopt Random Forest (RF) and Extreme Gradient Boosting (XGB) approaches for deriving reduced order models of hydrologic exchange flows and associated transit time distributions, with integrated field observations (e.g., bathymetry) and hydrodynamic simulation data (e.g., river velocity, depth). The setup allows an improved understanding of the influences of various physical, spatial, and temporal factors on the hydrologic exchange flows and transit times. The predictors also contain those derived using hybrid clustering, leveraging our previous work on river corridor system hydromorphic classification. The machine learning-based predictive models are developed and validated along the Columbia River Corridor, and the results show that the top parameters are the thickness of the top geological formation layer, the flow regime, river velocity, and river depth; the RF and XGB models can achieve 70% to 80% accuracy and therefore are effective alternatives to the computational demanding numerical models of exchange flows and transit time distributions. Each machine learning model with its favorable configuration and setup have been evaluated. The transferability of the models to other river reaches and larger scales, which mostly depends on data availability, is also discussed.

Groundwater Inflows to the Columbia River Along the Hanford Reach and Associated Nitrate Concentrations

Healthy river ecosystems require the interaction of many physical and biological processes to maintain their status. One physical process supporting biogeochemical cycling is hydrologic exchange (i.e., hydrologic exchange flows, HEFs) between relatively fast-flowing channel waters and slower-flowing surface and subsurface waters (lateral and vertical). Land uses adjacent to rivers have the potential to alter the water quality of off-channel surface and subsurface waters, and HEFs therefore have the potential to deliver solutes associated with river-adjacent land uses to rivers. HEFs can be nonpoint, diffuse sources of pollution, making the ultimate pollution source difficult to identify, especially in large rivers. Here, we seek to identify HEFs in the Columbia River near Richland, WA by looking for anomalies in temperature and electrical conductivity (EC) along the bed of the river in February, June, July, August, and November 2018. These are ultimately the “ends” of HEFs as they are locations of subsurface inflow to the river. We found these anomalies to be a combination of warmer or colder and higher (but not lower) EC than river water. We identified a majority of warm anomalies in February and July 2018, and majority cold anomalies in June, August, and November 2018. High-EC anomalies were found mostly in February, August, and November. Combined, we observe a shift from warm, high EC anomalies dominating in February to equivalent EC, warm anomalies in June, to equivalent EC, cool anomalies dominating July. In August, we also measured dissolved nitrate (NO3-) in-situ to determine whether anomalies were associated with increased NO3- loading to the river, especially along the eastern shoreline, which is dominated by agricultural land use. Inflows along the eastern shoreline have greater concentrations of nitrate than river water (up to 10 mg N–NO3-/L). This research demonstrates that HEFs are temporally and spatially dynamic transferring heat and solutes to rivers.

High-Performance Simulation of Dynamic Hydrologic Exchange and Implications for Surrogate Flow and Reactive Transport Modeling in a Large River Corridor

Hydrologic exchange flows (HEFs) have environmental significance in riverine ecosystems. Key river channel factors that influence the spatial and temporal variations of HEFs include river stage, riverbed morphology, and riverbed hydraulic conductivity. However, their impacts on HEFs were often evaluated independently or on small scales. In this study, we numerically evaluated the combined interactions of these factors on HEFs using a high-performance simulator, PFLOTRAN, for subsurface flow and transport. The model covers 51 square kilometers of a selected river corridor with large sinuosity along the Hanford Reach of the Columbia River in Washington, US. Three years of spatially distributed hourly river stages were applied to the riverbed. Compared to the simulation when riverbed heterogeneity is not ignored, the simulation using homogeneous riverbed conductivity underestimated HEFs, especially upwelling from lateral features, and overestimated the mean residence times derived from particle tracking. To derive a surrogate model for the river corridor, we amended the widely used transient storage model (TSM) for riverine solute study at reach scale with reactions. By treating the whole river corridor as a batch reactor, the temporal changes in the exchange rate coefficient for the TSM were derived from the dynamic residence time estimated from the hourly PFLOTRAN results. The TSM results were evaluated against the effective concentrations in the hyporheic zone calculated from the PFLOTRAN simulations. Our results show that there is potential to parameterize surrogate models such as TSM amended with biogeochemical reactions while incorporating small-scale process understandings and the signature of time-varying streamflow to advance the mechanistic understanding of river corridor processes at reach to watershed scales. However, the assumption of a well-mixed storage zone for TSM should be revisited when redox-sensitive reactions in the storage zones play important roles in river corridor functioning.