Geology topics

Yong Zhang

Publications and source records attributed to Yong Zhang.

13 recordsLinked to original sources

USGS2025

Modeling seawater intrusion along the Alabama coastline using physical and machine learning models to evaluate the effects of multiscale natural and anthropogenic stresses

Seawater intrusion threatens groundwater resources in coastal regions, including southern Baldwin County, Alabama, where the freshwater-saltwater interface dynamics remain poorly understood. To address this gap, this study uses combined physics-based and machine-learning models to quantify seawater intrusion caused by natural (storm surges) and anthropogenic (human activities) perturbations. The long short-term memory network and wavelet analysis were used to assess vertical aquifer vulnerabilities, revealing that the shallow part of the Coastal lowlands aquifer system (CL1) in the southern Baldwin County region is more susceptible to sea level rise and groundwater extraction than deeper aquifers. Based on these findings, a cross-sectional numerical model (physics approach) for the CL1 aquifer was developed to evaluate tidal and storm surge effects, using Tropical Storm Claudette (June 2021) as a case study. Results showed that tidal fluctuations had a minimal impact on the saltwater-freshwater interface location, whereas storm surges caused substantial inland movement, with effects lasting for nine months. The steady-state version of the three-dimensional (3D) physical model predicted seawater intrusion across the entire area, and convolutional neural network-based modeling further validated the model results. The 3D physical model was also applied to a smaller area to assess human impact on the saltwater interface due to two groundwater pumping scenarios (± 50% of the baseline pumping rate). Results revealed that a 50% increase in groundwater withdrawals caused seawater to advance ~ 320 m inland, whereas a 50% reduction led to a ~ 270-meter retreat. This study highlights the vulnerability of Alabama’s shallow coastal aquifers to seawater intrusion due to storm surges and human activities, and demonstrates that combining physics-based models with machine learning approaches can improve groundwater predictions, though its accuracy depends on the availability of site-specific data.

Alabama↗

USGS2025

Analyzing multi-year nitrate concentration evolution in Alabama aquatic systems using a machine learning model

Rising nitrate contamination in water systems poses significant risks to public health and ecosystem stability, necessitating advanced modeling to understand nitrate dynamics more accurately. This study applies the long short-term memory (LSTM) modeling to investigate the hydrologic and environmental factors influencing nitrate concentration dynamics in rivers and aquifers across the state of Alabama in the southeast of the United States. By integrating dynamic data such as streamflow and groundwater levels with static catchment attributes, the machine learning model identifies primary drivers of nitrate fluctuations, offering detailed insights into the complex interactions affecting multi-year nitrate concentrations in natural aquatic systems. In addition, a novel LSTM-based approach utilizes synthetic surface water nitrate data to predict groundwater nitrate levels, helping to address monitoring gaps in aquifers connected to these rivers. This method reveals potential correlations between surface water and groundwater nitrate dynamics, which is particularly meaningful given the lack of water quality observations in many aquifers. Field applications further show that, while the LSTM model effectively captures seasonal trends, limitations in representing extreme nitrate events suggest areas for further refinement. These findings contribute to data-driven water quality management, enhancing understanding of nitrate behavior in interconnected water systems.

Alabama↗

USGS2023

Long short-term memory models to quantify long-term evolution of streamflow discharge and groundwater depth in Alabama

Long short-term memory (LSTM) models have been shown to be efficient for rainfall-runoff modeling, and to a lesser extent, for groundwater depth forecasting. In this study, LSTMs were applied to quantify the spatiotemporal evolution of surface and subsurface hydrographs in Alabama in the Southeastern United States, where water sustainability has not been fully quantified across spatiotemporal scales. First, the surface water LSTM model with extensive dynamic (precipitation and other weather variables) and static (basin characteristics) inputs predicted the main characteristics of streamflow for six years at 19 gauged basins in Alabama. The model tended to underestimate extremely high streamflow but adding drainage density as an input feature slightly improved the predictions of extreme events. Second, to predict the groundwater depth evolution, a groundwater LSTM (GW-LSTM) model was proposed and applied using static inputs capturing the aquifers' hydrogeological properties and dynamic inputs of meteorological information. Three precipitation scenarios were also explored to evaluate the groundwater hydrograph evolution in the next two decades. The GW-LSTM model predicted the general trend of daily groundwater depth fluctuations (at 21 wells distributed across Alabama from 1990 to 2021) including most extremely high groundwater levels, and recovered groundwater depth for locations withheld from model training and validation. This study, therefore, extended the application of LSTMs in quantifying the spatiotemporal evolution of surface water and groundwater, two manifestations of a single integrated resource.

Alabama↗

USGS2021

Co-transport of biogenic nano-hydroxyapatite and Pb(II) in saturated sand columns: Controlling factors and stochastic modeling

Biogenic nano-hydroxyapatite (bio-nHAP) has recently gained great interest in many domains, especially in the remediation of heavy metal-contaminated soil, due to its high reactivity, low cost, and eco-friendly nature. The co-transport and reaction of bio-nHAP with Pb(II) in saturated porous media, however, are not well understood. This work investigated the effects of ionic strength (IS), ionic composition (IC), dissolved organic matter (DOM), and flow velocity on transport-reaction dynamics of Pb(II) and bio-nHAP by combining column breakthrough experiments and model simulations. Results showed that the mobility of Pb(II) was significantly enhanced with increasing IS/IC but less affected by flow velocity during the transport-reaction process of bio-nHAP and Pb(II) in the saturated sand column; while the transport of bio-nHAP was restricted by increasing IS/IC but facilitated by increasing velocity. IC, IS, and velocity only slightly affected the reaction kinetics between Pb(II) and bio-nHAP, likely due to the fast reaction rate between Pb(II) and bio-nHAP and precipitation of pyromorphite. The transport dynamics of bio-nHAP and Pb(II) were significantly changed by DOM, and this effect depended strongly on the type of DOM with different molecular weights. Breakthrough curves of Pb(II) and bio-nHAP exhibited apparent “anomalous”, sub-diffusive transport behaviors, which could be well quantified by a novel tempered fractional derivative bimolecular reaction equation (T-FBRE). Our findings highlighted the accurate simulation of the co-transport and reaction of bio-nHAP with Pb(II) using T-FBRE and had a great benefit for risk assessment and remediation strategy development for Pb(II) contaminated soil.

Chemosphere↗

USGS2021

Time-fractional flow equations (t-FFEs) to upscale transient groundwater flow characterized by temporally non-darcian flow due to medium heterogeneity

Upscaling groundwater flow is a fundamental challenge in hydrogeology. This study proposed time-fractional flow equations (t-FFEs) for upscaling long-term, transient groundwater flow and propagation of pressure heads in heterogeneous media. Monte Carlo simulations showed that, with increasing variance and correlation of the hydraulic conductivity ( K ), flow dynamics gradually deviated from Darcian flow and exhibit sub-diffusive, time-dependent evolution which can be separated into three major stages. At the early stage, the interconnected high- K zones dominated flow, while at intermediate times, the transverse flow due to mixed high- and low- K zones caused delayed rise of the piezometric head. At late times when flow in the relatively high- K domains reached stability, cells with very low- K continued to block the entry of water and generate “islands” with low piezometric head, significantly extending the temporal evolution of the piezometric head. The elongated water breakthrough curve cannot be quantified by the flow equation with an effective K , the space-fractional flow equation, or the multi-rate mass transfer (MRMT) flow model with a few rates, motivating the development of t-FFEs assuming temporally non-Darcian flow. Model applications showed that both the early and intermediate stages of flow dynamics can be captured by a single-index t-FFE (whose index is the exponent of the power-law probability density function of the random operational time for water parcels), but the overall evolution of flow dynamics, especially the enhanced retention of flow at later times, required a distributed-order t-FFE with variable indexes for different flow phases that can dominate flow dynamics at different stages. Therefore, transient groundwater flow in aquifers with spatially stationary heterogeneity can be temporally non-Darcian and non-stationary, due to the time-sensitive, combined effects of interconnected high- K channels and isolated low- K deposits on flow dynamics (which is the hydrogeological mechanism for the temporally non-Darcian flow and sub-diffusive pressure propagation), whose long-term behavior can be quantified by multi-index stochastic models.

Water Resources Research↗

USGS2018

Comparison of time nonlocal transport models for characterizing non-Fickian transport: From mathematical interpretation to laboratory application

Non-Fickian diffusion has been increasingly documented in hydrology and modeled by promising time nonlocal transport models. While previous studies showed that most of the time nonlocal models are identical with correlated parameters, fundamental challenges remain in real-world applications regarding model selection and parameter definition. This study compared three popular time nonlocal transport models, including the multi-rate mass transfer (MRMT) model, the continuous time random walk (CTRW) framework, and the tempered time fractional advection–dispersion equation (tt-fADE), by focusing on their physical interpretation and feasibility in capturing non-Fickian transport. Mathematical comparison showed that these models have both related parameters defining the memory function and other basic-transport parameters (i.e., velocity v and dispersion coefficient D ) with different hydrogeologic interpretations. Laboratory column transport experiments and field tracer tests were then conducted, providing data for model applicability evaluation. Laboratory and field experiments exhibited breakthrough curves with non-Fickian characteristics, which were better represented by the tt-fADE and CTRW models than the traditional advection–dispersion equation. The best-fit velocity and dispersion coefficient, however, differ significantly between the tt-fADE and CTRW. Fitting exercises further revealed that the observed late-time breakthrough curves were heavier than the MRMT solutions with no more than two mass-exchange rates and lighter than the MRMT solutions with power-law distributed mass-exchange rates. Therefore, the time nonlocal models, where some parameters are correlated and exchangeable and the others have different values, differ mainly in their quantification of pre-asymptotic transport dynamics. In all models tested above, the tt-fADE model is attractive, considering its small fitting error and the reasonable velocity close to the measured flow rate.

Water↗

USGS2016

Bounded fractional diffusion in geological media: Definition and Lagrangian approximation

Spatiotemporal Fractional-Derivative Models (FDMs) have been increasingly used to simulate non-Fickian diffusion, but methods have not been available to define boundary conditions for FDMs in bounded domains. This study defines boundary conditions and then develops a Lagrangian solver to approximate bounded, one-dimensional fractional diffusion. Both the zero-value and non-zero-value Dirichlet, Neumann, and mixed Robin boundary conditions are defined, where the sign of Riemann-Liouville fractional derivative (capturing non-zero-value spatial-nonlocal boundary conditions with directional super-diffusion) remains consistent with the sign of the fractional-diffusive flux term in the FDMs. New Lagrangian schemes are then proposed to track solute particles moving in bounded domains, where the solutions are checked against analytical or Eularian solutions available for simplified FDMs. Numerical experiments show that the particle-tracking algorithm for non-Fickian diffusion differs from Fickian diffusion in relocating the particle position around the reflective boundary, likely due to the non-local and non-symmetric fractional diffusion. For a non-zero-value Neumann or Robin boundary, a source cell with a reflective face can be applied to define the release rate of random-walking particles at the specified flux boundary. Mathematical definitions of physically meaningful nonlocal boundaries combined with bounded Lagrangian solvers in this study may provide the only viable techniques at present to quantify the impact of boundaries on anomalous diffusion, expanding the applicability of FDMs from infinite do mains to those with any size and boundary conditions.

Water Resources Research↗

USGS2016

Regional oxygen reduction and denitrification rates in groundwater from multi-model residence time distributions, San Joaquin Valley, USA

Rates of oxygen and nitrate reduction are key factors in determining the chemical evolution of groundwater. Little is known about how these rates vary and covary in regional groundwater settings, as few studies have focused on regional datasets with multiple tracers and methods of analysis that account for effects of mixed residence times on apparent reaction rates. This study provides insight into the characteristics of residence times and rates of O 2 reduction and denitrification (NO 3 − reduction) by comparing reaction rates using multi-model analytical residence time distributions (RTDs) applied to a data set of atmospheric tracers of groundwater age and geochemical data from 141 well samples in the Central Eastern San Joaquin Valley, CA. The RTD approach accounts for mixtures of residence times in a single sample to provide estimates of in-situ rates. Tracers included SF 6 , CFCs, 3 H, He from 3 H (tritiogenic He), 14 C, and terrigenic He. Parameter estimation and multi-model averaging were used to establish RTDs with lower error variances than those produced by individual RTD models. The set of multi-model RTDs was used in combination with NO 3 − and dissolved gas data to estimate zero order and first order rates of O 2 reduction and denitrification. Results indicated that O 2 reduction and denitrification rates followed approximately log-normal distributions. Rates of O 2 and NO 3 − reduction were correlated and, on an electron milliequivalent basis, denitrification rates tended to exceed O 2 reduction rates. Estimated historical NO 3 − trends were similar to historical measurements. Results show that the multi-model approach can improve estimation of age distributions, and that relatively easily measured O 2 rates can provide information about trends in denitrification rates, which are more difficult to estimate.

California↗

USGS2015

Rapid removal of nitrobenzene in a three-phase ozone loaded system with gas-liquid-liquid

This study explores the removal rate of nitrobenzene (NB) using a new gas-liquid-liquid (G-L-L) three-phase ozone loaded system consisting of a gaseous ozone, an aqueous solvent phase, and a fluorinated solvent phase (perfluorodecalin, or FDC). The removal rate of NB was quantified in relation to six factors including 1) initial pH, 2) initial NB dosage, 3) gaseous ozone dosage, 4) free radical scavenger, 5) FDC pre-aerated gaseous ozone, and 6) reuse of FDC. The NB removal rate is positively affected by the first three factors. Compared with the conventional gas-liquid (water) (G-L) two-phase ozonation system, the free radical scavenger (tertiary butyl alcohol) has much less influence on the removal rate of NB in the G-L-L system. The FDC loaded ozone acts as an ozone reservoir and serves as the main reactive phase in the G-L-L three-phase system. The reuse of FDC has little influence on the removal rate of NB. These experimental results suggest that the oxidation efficiency of ozonation in the G-L-L three-phase system is better than that in the conventional G-L two-phase system.

Chemical Engineering Communications↗

USGS2015

Peclet number as affected by molecular diffusion controls transient anomalous transport in alluvial aquifer-aquitard complexes

This study evaluates the role of the Peclet number as affected by molecular diffusion in transient anomalous transport, which is one of the major knowledge gaps in anomalous transport, by combining Monte Carlo simulations and stochastic model analysis. Two alluvial settings containing either short- or long-connected hydrofacies are generated and used as media for flow and transport modeling. Numerical experiments show that 1) the Peclet number affects both the duration of the power-law segment of tracer breakthrough curves (BTCs) and the transition rate from anomalous to Fickian transport by determining the solute residence time for a given low-permeability layer, 2) mechanical dispersion has a limited contribution to the anomalous characteristics of late-time transport as compared to molecular diffusion due to an almost negligible velocity in floodplain deposits, and 3) the initial source dimensions only enhance the power-law tail of the BTCs at short travel distances. A tempered stable stochastic (TSS) model is then applied to analyze the modeled transport. Applications show that the time-nonlocal parameters in the TSS model relate to the Peclet number, P e . In particular, the truncation parameter in the TSS model increases nonlinearly with a decrease in P e due to the decrease of the mean residence time, and the capacity coefficient increases with an increase in molecular diffusion which is probably due to the increase in the number of immobile particles. The above numerical experiments and stochastic analysis therefore reveal that the Peclet number as affected by molecular diffusion controls transient anomalous transport in alluvial aquifer–aquitard complexes.

Journal of Contaminant Hydrology↗

USGS2014

Accuracy of travel time distribution (TTD) models as affected by TTD complexity, observation errors, and model and tracer selection

Analytical models of the travel time distribution (TTD) from a source area to a sample location are often used to estimate groundwater ages and solute concentration trends. The accuracies of these models are not well known for geologically complex aquifers. In this study, synthetic datasets were used to quantify the accuracy of four analytical TTD models as affected by TTD complexity, observation errors, model selection, and tracer selection. Synthetic TTDs and tracer data were generated from existing numerical models with complex hydrofacies distributions for one public-supply well and 14 monitoring wells in the Central Valley, California. Analytical TTD models were calibrated to synthetic tracer data, and prediction errors were determined for estimates of TTDs and conservative tracer (NO 3 −) concentrations. Analytical models included a new, scale-dependent dispersivity model (SDM) for two-dimensional transport from the watertable to a well, and three other established analytical models. The relative influence of the error sources (TTD complexity, observation error, model selection, and tracer selection) depended on the type of prediction. Geological complexity gave rise to complex TTDs in monitoring wells that strongly affected errors of the estimated TTDs. However, prediction errors for NO3− and median age depended more on tracer concentration errors. The SDM tended to give the most accurate estimates of the vertical velocity and other predictions, although TTD model selection had minor effects overall. Adding tracers improved predictions if the new tracers had different input histories. Studies using TTD models should focus on the factors that most strongly affect the desired predictions.

California↗

USGS2014

Linking aquifer spatial properties and non-Fickian transport in mobile-immobile like alluvial settings

Time-nonlocal transport models can describe non-Fickian diffusion observed in geological media, but the physical meaning of parameters can be ambiguous, and most applications are limited to curve-fitting. This study explores methods for predicting the parameters of a temporally tempered Lévy motion (TTLM) model for transient sub-diffusion in mobile–immobile like alluvial settings represented by high-resolution hydrofacies models. The TTLM model is a concise multi-rate mass transfer (MRMT) model that describes a linear mass transfer process where the transfer kinetics and late-time transport behavior are controlled by properties of the host medium, especially the immobile domain. The intrinsic connection between the MRMT and TTLM models helps to estimate the main time-nonlocal parameters in the TTLM model (which are the time scale index, the capacity coefficient, and the truncation parameter) either semi-analytically or empirically from the measurable aquifer properties. Further applications show that the TTLM model captures the observed solute snapshots, the breakthrough curves, and the spatial moments of plumes up to the fourth order. Most importantly, the a priori estimation of the time-nonlocal parameters outside of any breakthrough fitting procedure provides a reliable “blind” prediction of the late-time dynamics of subdiffusion observed in a spectrum of alluvial settings. Predictability of the time-nonlocal parameters may be due to the fact that the late-time subdiffusion is not affected by the exact location of each immobile zone, but rather is controlled by the time spent in immobile blocks surrounding the pathway of solute particles. Results also show that the effective dispersion coefficient has to be fitted due to the scale effect of transport, and the mean velocity can differ from local measurements or volume averages. The link between medium heterogeneity and time-nonlocal parameters will help to improve model predictability for non-Fickian transport in alluvial settings.

Journal of Hydrology↗

USGS2013

The impact of medium architecture of alluvial settings on non-Fickian transport

The influence of heterogeneous architecture of alluvial aquifers on non-Fickian transport is explored using the Monte Carlo approach. More than two thousand high-resolution hydrofacies models representing seven groups of alluvial settings are built to test the effects of varying facies proportions, mean length and its anisotropy ratio, juxtapositional tendencies, and sub-facies heterogeneity. Results show that the volumetric fraction (P(Z)) of floodplain layers classified by their thicknesses Z controls the non-Fickian tailing of tracer transport at late times. A simple quantitative relationship S BTC ≈S P(Z) /2-1 is built based on a multi-rate mass transfer analysis, where S BTC is the slope of the power-law portion of tracer breakthrough curve, and S P(Z) denotes the slope of the power-law portion of the distribution of P(Z) which can be measured, e.g., in core logs. At early times, the mean length of hydrofacies affects the non-Fickian tailing by controlling the channeling of flow in high-permeability non-floodplain materials and the sequestration in surrounding low-permeability floodplain layers. The competition between channeling and sequestration generates complex pre-asymptotic features, including sublinear growth of plume mean displacement, superlinear growth of plume variance, and skewed mass distribution. Those observations of the influence of medium heterogeneity on tracer transport at early and late times may lead to development of nonlocal transport models that can be parameterized using measurable aquifer characteristics.

Advances in Water Resources↗