Geology topics

Kenneth C. Carroll

Publications and source records attributed to Kenneth C. Carroll.

9 recordsLinked to original sources

USGS2026

Localization of spatiotemporally heterogeneous subsurface flows using autoencoder-based deep learning framework for time-lapse self-potential tomography

Self-potential (SP) monitoring has emerged as a valuable method for characterizing subsurface hydrogeological features and processes due to its sensitivity to fluid-induced electrokinetic effects. Despite advancements in SP inversion, challenges remain in imaging groundwater dynamics from SP activities due to complex hydrological settings and transient noise. In this study, a deep learning autoencoder (AE)-based framework is proposed for the spatiotemporal localization of subsurface fluid movement from time-lapse SP tomography. Temporal segments of time-lapse numerical inversions were first derived from long-term SP monitoring conducted from a floodplain site in Oak Ridge, Tennessee, known for active hyporheic exchange. Subsequently, AE models based on vision transformer (ViT), convolutional long short-term memory (ConvLSTM), convolutional neural network, and temporal convolutional network were individually trained and compared on the SP tomography segments for reconstruction performance. Finally, the reconstruction error over time serves as an anomaly score to identify moments of active SP variation, whereas spatial distributions of errors within these moments are analyzed to image and localize regions associated with anomalous subsurface fluid movement. The results demonstrate that ConvLSTM- and ViT-AE are most capable for the localization task with contrasting error distributions and consistent delineation of anomalies. Applying the method to both SP arrays parallel and perpendicular to the stream produced consistent anomaly zones near a fault or karst feature, validating the robustness and generalization of the approach. These results demonstrate the potential of the proposed framework as a scalable and interpretable tool for spatiotemporal analysis of subsurface flow dynamics in complex hydrogeological systems.

Tennessee↗

USGS2025

Imaging hyporheic exchange by integrating deep learning and physics-informed inversion of time-lapse self-potential data

Self-potential (SP) monitoring is increasingly used for subsurface flow characterization due to its sensitivity to hydrogeological and geochemical processes. However, SP inversion remains challenging due to its ill-posed nature, sparse data coverage, and strong transient noise. This study proposes a hybrid framework to image hyporheic exchange using a time-lapse SP data set monitored from a streamflow site in Oak Ridge, Tennessee. Dipole moment tomography grids generated from the physics-informed numerical inversion is first used to train a Vision Transformer (ViT) model that maps surface SP sequences to 2D source distributions. While the numerical method is more responsive to transient signals, the ViT model better captures persistent spatial structures. Their complementary outputs are jointly analyzed in the spatiotemporal domain to isolate dynamic hyporheic exchange zones and distinguish transient from steady state subsurface flow features. This approach integrates physical inversion and deep learning to enhance interpretability, generalization, and temporal awareness in SP analysis.

Geophysical Research Letters↗

USGS2024

Self-potential tomography preconditioned by particle swarm optimization— Application to monitoring hyporheic exchange in a bedrock river

A self-potential (SP) data-inversion algorithm was developed and tested on an analytical model of electrical-potential profile data attributed to single and multiple polarized electrical sources. The developed algorithm was then validated by an application to SP-monitoring field data measured on the floodplain of East Fork Poplar Creek, Oak Ridge, Tennessee, to image electrical sources in areas conducive to preferential flow into the flood plain from the bedrock-lined riverbed. The algorithm combined stochastic source-localization by particle-swarm-optimization (PSO) of electrical sources characterized by simplified geometries with source tomography by regularized weighted least-squares minimization of a quadratic objective function. Prior information was incorporated by preconditioning the tomography algorithm by PSO results. Variable percentages of random noise were added to analytical-model data to evaluate the algorithm performance. Results indicated that true parameters of single-source models were inverted and approximated with small residual error, whereas inversion of analytical-model data representing multiple electrical sources accurately approximated the locations of the sources but miscalculated some parameters because of the non-uniqueness of the inverse-model solution. Source tomography applied to analytical model data during testing produced a spatially continuous parameter field that identified the locations of point-scale synthetic dipole sources of electrical current flow with varying degrees of accuracy depending on the prior information incorporated into the tomography. When applied to SP-monitoring field data, the algorithm imaged electrical sources within a known fault that intersects the bedrock riverbed and flood plain of East Fork Poplar Creek and depicted dynamic electrical conditions attributed to hyporheic exchange.

Tennessee↗

USGS2023

Geoelectric monitoring of the electric potential field of the Lower Rio Grande before, during, and after Intermittent Streamflow, May–October, 2022

Understanding the intermittent hydraulic connectivity between ephemeral streams and alluvial aquifers is a key challenge for managing water resources in arid environments. The lower Rio Grande flows for short, discontinuous periods during the irrigation season through the Mesilla Basin in southeastern New Mexico and southwestern Texas. Hydraulic connections between the Rio Grande and the Rio Grande alluvial aquifer in the Mesilla Basin vary spatially and temporally and are not well understood. Self-potential (SP) monitoring and time-lapse electric resistivity tomography (ERT) were therefore performed along linear cross-sections spanning the riverbed and flood plain for more than 4 months to monitor the transient hydraulic connection between the river and the alluvial aquifer by measuring time-lapse changes in the electric potential field in the riverbed and flood plain. The monitoring period began on 21 May 2022, when the riverbed was completely dry, continued through the irrigation season while streamflow was provided by reservoir releases from upstream dams, and ended on 4 October 2022, when the riverbed was again dry. SP monitoring data show (1) a background condition in the dry riverbed consisting of (a) a positive electric potential anomaly with a maximum amplitude of about +100 mV attributed predominantly to a subsurface vertical salt concentration gradient and (b) diurnal electric potential fluctuations with amplitudes of 40,000–90,000 mV attributed to near-surface heat conduction driven by weather variability, in addition to (2) a streaming potential anomaly during the irrigation season with a maximum amplitude of about −3500 mV whose transient behavior clearly exhibited a change from the background anomaly to depict exclusively losing streamflow conditions that persisted through the irrigation season. Time-lapse ERT monitoring results depict rapid infiltration of streamflow into the subsurface and imply the river and Rio Grande alluvial aquifer established a full hydraulic connection within a few hours after streamflow arrival at the monitoring site. SP monitoring data show an apparent transition from hydraulic connection to disconnection at the end of the irrigation season and indicate that the transitional phase between connection and disconnection may last substantially longer than the transition from disconnection to connection. The combination of SP and ERT monitoring demonstrated herein shows the potential for broader applications of time-lapse monitoring of hydraulic intermittency and near-surface heat fluxes in different rivers.

New Mexico, Texas↗

USGS2022

Mesilla / Conejos-Médanos Basin: U.S.-Mexico transboundary water resources and research needs

Synthesizing binational data to characterize shared water resources is critical to informing binational management. This work uses binational hydrogeology and water resource data in the Mesilla/Conejos-Médanos Basin (Basin) to describe the hydrologic conceptual model and identify potential research that could help inform sustainable management. The Basin aquifer is primarily composed of continuous basin-fill Santa Fe Group sediments, allowing for transboundary throughflow. Groundwater flow, however, may be partially or fully restricted by intrabasin uplifts and limited recharge. The shallow groundwater in the Rio Grande alluvium receives recharge from the Rio Grande and responds to changes in water supply and demand. About 11% of Rio Grande alluvial groundwater volume is recharged annually, an amount that is less than recent withdrawals. Potentially recoverable fresh to slightly brackish groundwater was estimated at 82,600 cubic hectometers in the U.S. portion of the Basin and 69,100 cubic hectometers in the Mexican portion. Alluvial groundwater geochemistry is governed by the evaporative concentration of the Rio Grande and agricultural diversions, whereas deeper groundwater geochemistry is governed by mixing and geochemical processes. Continued refinements to storage estimates, the water budget, and deep groundwater extent and geochemistry can improve estimates of sustainable use and inform alternative water sources.

Mesilla/Conejos-Médanos Basin↗

USGS2021

An integrated geochemical approach for defining sources of groundwater salinity in the southern Rio Grande Valley of the Mesilla Basin, New Mexico and west Texas, USA

A significantly elevated groundwater salinity zone was identified in the southern part of the Mesilla Valley. This investigation characterized the occurrence, spatial extent, and source of the plume of elevated groundwater salinity using a wide range of geochemical and geophysical data and methods.

New Mexico, Texas↗

USGS2016

Geochemistry of formation waters from the Wolfcamp and “Cline” shales: Insights into brine origin, reservoir connectivity, and fluid flow in the Permian Basin, USA

Despite being one of the most important oil producing provinces in the United States, information on basinal hydrogeology and fluid flow in the Permian Basin of Texas and New Mexico is lacking. The source and geochemistry of brines from the basin were investigated (Ordovician- to Guadalupian-age reservoirs) by combining previously published data from conventional reservoirs with geochemical results for 39 new produced water samples, with a focus on those from shales. Salinity of the Ca–Cl-type brines in the basin generally increases with depth reaching a maximum in Devonian (median = 154 g/L) reservoirs, followed by decreases in salinity in the Silurian (median = 77 g/L) and Ordovician (median = 70 g/L) reservoirs. Isotopic data for B, O, H, and Sr and ion chemistry indicate three major types of water. Lower salinity fluids (<70 g/L) of meteoric origin in the middle and upper Permian hydrocarbon reservoirs (1.2–2.5 km depth; Guadalupian and Leonardian age) likely represent meteoric waters that infiltrated through and dissolved halite and anhydrite in the overlying evaporite layer. Saline (>100 g/L), isotopically heavy (O and H) water in Leonardian [Permian] to Pennsylvanian reservoirs (2–3.2 km depth) is evaporated, Late Permian seawater. Water from the Permian Wolfcamp and Pennsylvanian “Cline” shales, which are isotopically similar but lower in salinity and enriched in alkalis, appear to have developed their composition due to post-illitization diffusion into the shales. Samples from the “Cline” shale are further enriched with NH 4 , Br, I and isotopically light B, sourced from the breakdown of marine kerogen in the unit. Lower salinity waters (<100 g/L) in Devonian and deeper reservoirs (>3 km depth), which plot near the modern local meteoric water line, are distinct from the water in overlying reservoirs. We propose that these deep meteoric waters are part of a newly identified hydrogeologic unit: the Deep Basin Meteoric Aquifer System. Chemical, isotopic, and pressure data suggest that despite over-pressuring in the Wolfcamp shale, there is little potential for vertical fluid migration to the surface environment via natural conduits.

New Mexico, Texas↗

USGS2016

Volatile-organic molecular characterization of shale-oil produced water from the Permian Basin

Growth in unconventional oil and gas has spurred concerns on environmental impact and interest in beneficial uses of produced water (PW), especially in arid regions such as the Permian Basin, the largest U.S. tight-oil producer. To evaluate environmental impact, treatment, and reuse potential, there is a need to characterize the compositional variability of PW. Although hydraulic fracturing has caused a significant increase in shale-oil production, there are no high-resolution organic composition data for the shale-oil PW from the Permian Basin or other shale-oil plays (Eagle Ford, Bakken, etc.). PW was collected from shale-oil wells in the Midland sub-basin of the Permian Basin. Molecular characterization was conducted using high-resolution solid phase micro extraction gas chromatography time-of-flight mass spectrometry. Approximately 1400 compounds were identified, and 327 compounds had a >70% library match. PW contained alkane, cyclohexane, cyclopentane, BTEX (benzene, toluene, ethylbenzene, and xylene), alkyl benzenes, propyl-benzene, and naphthalene. PW also contained heteroatomic compounds containing nitrogen, oxygen, and sulfur. 3D van Krevelen and double bond equivalence versus carbon number analyses were used to evaluate molecular variability. Source composition, as well as solubility, controlled the distribution of volatile compounds found in shale-oil PW. The salinity also increased with depth, ranging from 105 to 162 g/L total dissolved solids. These data fill a gap for shale-oil PW composition, the associated petroleomics plots provide a fingerprinting framework, and the results for the Permian shale-oil PW suggest that partial treatment of suspended solids and organics would support some beneficial uses such as onsite reuse and bio-energy production.

New Mexico, Texas↗

USGS2008

Scaling sap flux measurements of grazed and ungrazed shrub communities with fine and coarse-resolution remote sensing

We measured transpiration by black greasewood ( Sarcobatus vermiculatus ) (SAVE) and fourwing saltbush ( Atriplex canescens ) (ATCA) over a nitrate-contaminated aquifer in Monument Valley, Arizona, on the Colorado Plateau. Heat balance sap flow sensors were used to measure transpiration by shrubs in 2006 and 2007 and results were scaled to larger landscape units and longer time scales using leaf area index (LAI), fractional vegetation cover, meteorological data, and the enhanced vegetation index (EVI) from MODIS sensors on the Terra satellite. Transpiration was high depending on leaf area (2·95–6·72 kg m −2 d −1 ) and was controlled by vapour pressure deficit ( D ) in the atmosphere. SAVE tended to have higher transpiration rates than ATCA and had a steeper response to D , but both exhibited midday depression of leaf conductance. Over most of the site, fractional vegetation cover ( f c ) and area-wide LAI were low (0·10 and 0·37, respectively) due to heavy grazing by cattle and sheep. However, a portion of the plume that had been protected from grazing for 10 years had f c = 0·75, LAI = 2·88. Transpiration rates on a ground-area basis varied with LAI, with midsummer daily values ranging from 1·44 mm d −1 (LAI = 0·36) to 13·1 mm d −1 (LAI = 2·88 mm) over the site, corresponding to projected annual values of 159–1447 mm year −1 . Controlling grazing could, theoretically, slow or halt the movement of the contamination plume by allowing the shrub community to extract more water than is recharged in the aquifer.

Utah↗