Geology Reports⌕ Search

SEARCH · Geology Reports

Results for “Journal of the Nevada Water Resources Association”

Search indexed USGS publications on groundwater, aquifers, geologic maps, mineral resources and earthquakes. Explore source records by subject and place.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

1,691 records · Page 7Linked to original sources
USGS2026

Keeping pace: Migrating sea lampreys adjust swim speed in response to shifting water velocity to maintain a constant ground speed

Upstream migrating fish must balance energetic costs with environmental and ecological constraints. Theoretical models predict that fish can conserve energy by adjusting swim speed in response to changing water velocity, but detailed field tests of these predictions remain rare. We used fine-scale acoustic telemetry and hydrodynamic modeling to track adult sea lamprey (Petromyzon marinus) migrating through a river and evaluated how movement speeds varied with environmental conditions and individual traits. Swim speed increased with faster water velocities, while ground speed remained constant, supporting the hypothesis that lampreys regulate swim effort to maintain a fixed migration pace. The average ground speed aligned with a predicted optimum of ∼1 body length per second. Both swim and ground speeds declined slightly in deeper water, suggesting a trade-off between movement efficiency and predator avoidance. These findings provide field-based support for a theoretical model of energy optimization and underscore how migratory fish adaptively adjust behavior to local hydrodynamic conditions. Our results highlight the value of using high-resolution tracking and environmental modeling to reveal behavioral tactics that underlie efficient migration strategies.

Michigan↗
USGS2024

Taking heat (downstream): Simulating groundwater and thermal equilibrium controls on annual paired air–water temperature signal transport in headwater streams

Headwater stream temperature often exhibits spatial variation at the kilometer-scale, but the relative importance of the underlying hydrogeological processes and riverine perturbations remains poorly understood. In this study, we investigated the relative importance of groundwater (GW) and other processes on downstream annual stream temperature signal characteristics using deterministic heat budget model (HFLUX) scenarios within an idealized stream reach representative of mountainous forested conditions. We summarized annual stream thermal regimes from the relationship of paired sinusoidal air and water temperature signals (amplitude ratio, phase lag, and mean ratio). Results showed that downstream changes in annual temperature depended on the thermal gradient between water and the hypothetical equilibrium temperature (where all heat fluxes sum to zero). GW inflow, riparian shading, and the boundary input signal were the most significant factors affecting downstream annual water temperature signals, while flow volume and channel dimensions impacted how quickly annual temperature signals changed. Effects of GW were dominated by advective rather than conductive heat exchange processes, but conduction played a larger role when GW input was more spatially diffuse. Our results indicated several mechanisms by which local processes may affect stream thermal resilience to disturbances and can help guide management of wildfire and climate change.

Journal of Hydrology↗
USGS2026

High-resolution transboundary vegetation community maps of the Sonoran and Mojave Desert ecoregion to support critical landscape conservation planning and habitat management needs

We produced a 30-m resolution binational land cover map of Bird Conservation Region 33 (BCR 33) for the U.S. North American Bird Conservation Initiative. The region covers large portions of the Sonoran and Mojave Deserts. The map can support the U.S. Fish and Wildlife Service (FWS) Migratory Bird Program’s recovery planning efforts and constitutes the first known binational land cover dataset spanning sections of the United States–Mexico border and using a consistent classification system for both countries. The mapped region includes 152 distinct land cover classes, covering a total area of 38,421,453 ha (148,345 mi 2 ), of which 13,148,345 ha (52,706 mi 2 ) are located in Mexico and 24,770,640 ha (95,639 mi 2 ) in the United States. We primarily used Landsat 8 (OLI) imagery, supplemented by limited ground surveys from two field campaigns, drone-based aerial data, and existing vegetation classification frameworks from both countries. The classification applied a data-fusion approach integrating 30-m Landsat 8 imagery, decadal phenology metrics from vegetation indices, and a random forest model trained mainly with datasets from a comprehensive national mapping project from the U.S. Geological Survey (USGS) GAP Analysis Project (GAP) and federal wildland fire agencies’ Landscape Fire and Resource Management Planning Tools (LANDFIRE) (GAP/LANDFIRE) [United States side] and the National Institute of Statistics and Geography (INEGI) [Mexico side] as well as land cover maps and opportunistic open-access and field observations. Mapping of the full BCR 33 region was carried out in two phases: 1) Phase I, the prototype map, covered a smaller portion of the transboundary area and identified 31 land cover classes, and 2) Phase II, the full BCR 33 map (refer to Figure 1), which resulted in 152 land cover classes. Using a Random Forest classifier, we achieved an overall prediction accuracy of 92% for the Phase I map and 87% for the Phase II full region map. This slight decrease can be attributed to working on a larger, more complex area with a greater number of land cover classes. No formal validation was conducted, aside from using a subset of the collected field observations and training data to assess model performance during and after training. The training sites were further verified using Google Earth (Google, 2026) imagery. Two undergraduate students who worked for over a year visually inspected imagery and open access public images to confirm each training site during model training using in-house developed, online, visual tools. A portion of this field training data was reserved for model validation, and the corresponding results are to be presented in later sections. The project developed an end-to-end, medium- and fine-resolution remote sensing–based data fusion mapping approach. This effort produced a map (Nagler et al., 2025) and the online tools to support a dynamic, live, online map for visualizing the transboundary vegetation communities in BCR 33. The toolset is currently hosted by the University of Arizona (UofA) Vegetation Index and Phenology (VIP) Lab to support FWS partners (https://vip.arizona.edu/viplab_data_explorer?LCM_BCR33). The online map is designed to allow rapid updates using new training, validation, or correction data, making it dynamic and maintainable. The approach we took established a framework for rapid updating and correction of land cover maps, as the model can be quickly retrained with new field observations, updated training data, or other sources. This enables dynamic mapping and change detection of the region’s vegetation. This framework is an advance in data fusion and crowdsourced mapping of complex, vulnerable regions, providing support to regional stakeholders and the wider user community. This transboundary map can inform the protection, conservation, and restoration of vegetation, habitat, and ecosystems, particularly for threatened and endangered species across the two nations using consistent and harmonized binational mapping systems. Beyond supporting land management decisions and stakeholders in the transboundary desert ecoregions, this BCR 33 mapping effort establishes a foundation for future rapid, low-cost, cross-border land cover mapping that can benefit and advance ecosystem management.

Arizona, Baja California, California, Nevada, Sina↗
USGS2025

Controls on water quality below a reclaimed surface coal mine, southeastern Montana

Coal mining and reclamation can have a profound influence on hydrogeologic systems, with clear consequences for groundwater quality, yet their long-term influence on downgradient water quality over time following reclamation is less well documented. Geochemical trends were evaluated in water quality downgradient of a fully reclaimed landscape at the former Big Sky Mine in the Rosebud Creek watershed (southeastern Montana, USA), over a 3-year period (2020–2022), including bond release in 2022. Within 6 km downgradient from the reclaimed area, sulfate concentrations decreased from approximately 3500 to 1800 mg l −1 within the Miller Coulee alluvial aquifer. Major ions, δ 34 S SO4 values, and residence time tracers suggest that the observed decreases in sulfate concentration result from a combination of dilution by mixed-age inflows and incomplete transit of the high salinity plume from the mine boundary. Both bedrock and alluvial aquifers of the Rosebud Creek corridor contained contributions of millennia-old regional groundwater, which may serve to mitigate mine-derived high salinity waters. Rosebud Creek, which traverses the outflow zone of Miller Coulee in the study area, exhibited high sulfate concentrations during low flows and consistent downgradient increases in sulfate concentration. The possibility of plume dynamics in Miller Coulee suggests that the greatest water quality impacts may not yet have reached Rosebud Creek.

Montana↗
USGS2025

Water resources related to breccia pipe uranium mining in the Grand Canyon region

Introduction In the arid Grand Canyon region, water resources are limited to primarily the Colorado River and associated tributaries and to groundwater in the form of seeps and springs. Groundwater resources in the region supply water for human use and support diverse and rich ecosystems in the locations immediately surrounding the seeps and springs. Throughout the region, uranium resources occur and may interact with water resources in both mined and unmined uranium deposits. There is a need to better understand groundwater in the region and the effects from uranium mining in order to better manage the limited water resources in the area. This Fact Sheet summarizes results from U.S. Geological Survey studies that were conducted on this topic from 2012 to 2023.

Arizona↗
USGS2026

Thermal detectability of subsurface water ice on Mars: A comparative analysis for the Subsurface Water Ice Mapping (SWIM) Project

We have developed a new global map of shallow ground ice distribution, SWIM23, based on Mars Global Surveyor Thermal Emission Spectrometer data and made systematic comparisons between this new map and two similar, previously developed data products. We have explored the origin of differences between the three ice maps by detailing technical and procedural differences in their development, by making global pixel-by-pixel comparisons, and by carrying out a series of one-dimensional thermal simulations to explore fundamental physical limitations of thermal ice-detection techniques. These efforts and the production of a composite thermal ice-consistency map supported integration of multiple geophysical data products relevant to ice detection in the upper meter of the Martian regolith into the larger Mars Subsurface Water Ice Mapping project. Our work also highlights fundamental physical limitations to thermal ice detection as a technique, particularly the rapid fall-off in ice detection sensitivity at depths >30 cm, which produces maximum uncertainty in the presence and depth of ice within regions preferred for potential human landing sites. A future Mars orbiter mission designed to detect ice and support crewed landing site selection in the midlatitude region should give payload priority to an instrument capable of probing the 1–5 m depth range (i.e., a high-frequency radar), over a next-generation thermal spectrometer, which is unlikely to offer clarity on ice table depths or lateral continuity of the ice table in the locations of highest interest.

Planetary Science Journal↗
USGS2025

Reaction kinetics and accelerant effects of sulfides in early mature hydrocarbon generation using hydrous pyrolysis

Hydrocarbon generation in organic-rich sediments is influenced by the molecular organic composition and relative abundance of associated minerals. Certain mineral-derived elements act as catalysts and reaction intermediaries, facilitating early-stage hydrocarbon formation in potential source rocks. This study investigated the role of sulfur contributed from pyrite as an accelerant in thermal reaction, focusing on its effects on early maturation and consequent hydrocarbon generation from gilsonite (low-sulfur solid petroleum). Hydrous pyrolysis (HP) experiments were conducted on mixtures of gilsonite and pyrite in varying ratios (1:0.1, 1:0.5, 1:1, 1:2, and 1:10 w/w gilsonite:pyrite) at 320, 350, and 370 °C for 72 h. Untreated and thermally altered residues were analyzed using solid bitumen reflectance (BR o , %), total organic carbon (TOC) content, programmed temperature pyrolysis, scanning electron microscopy with energy-dispersive spectroscopy (SEM-EDS), and X-ray diffraction (XRD) to evaluate the potential accelerant role of pyritic sulfur in hydrocarbon formation. The results show HP residues at 320 and 350 °C with greater pyrite concentrations had higher BR o , while reflectance values were similar in the 370 °C residues, regardless of pyrite concentration, suggesting enhanced reaction at lower thermal conditions. Increasing pyrite content systematically decreased hydrogen index (HI) values while increasing the transformation ratio (TR) and production index (PI), indicating enhanced conversion of organic matter to hydrocarbons with increasing pyrite concentrations. Gas yields increased with pyrite addition, particularly at 350 °C, confirming secondary cracking effects. However, gas production stabilized or declined at higher pyrite loadings (1:10), suggesting alternative reaction pathways such as coke formation. Our data indicate the presence of pyrite lowers the activation energy for thermal cracking, shifting peak experimental hydrocarbon generation temperatures downward by 20–30 °C, with the most pronounced accelerant effects observed at moderate pyrite concentrations (1:0.5 and 1:1). The thermodynamic framework reveals that pyrite stability is influenced by experimental conditions, with pyrrhotite formation favored in the presence of gilsonite due to reduced oxygen fugacity. Pyrite transformation to pyrrhotite, as observed through XRD, SEM-EDS, and predicted by thermodynamic data, further supports the accelerant role of S, as pyrrhotite exhibits a higher hydrogen transfer potential, promoting early oil generation. These findings highlight the importance of pyrite in modulating hydrocarbon generation pathways in organic-rich systems.

Journal of Analytical and Applied Pyrolysis↗
USGS2026

A process-based model for forecasting wave runup along the coast of Georgia

Wave runup is an important nearshore process that impacts total water level, sediment transport, and coastal design. Current methods for forecasting wave runup implement an empirical model that considers offshore wave height, wave period, and generalized beach slope. In this study, the authors generated wave runup forecasts from offshore wave conditions and a system of polynomial equations derived from numerical simulations at three different still water datums for each beach profile. They developed a process-based methodology that incorporated site-specific cross-shore topobathy into the phase-resolving numerical model. A comparison between the system of equations, deterministic hydrodynamic simulations, and observed high-water marks was made using Hurricanes Matthew (2016) and Irma (2017) for 12 cases, and it showed that the polynomials were capable of being consistent with the results from full simulation runs, while not requiring hours of runtime when a forecast was needed—the differences between the polynomial and the observed high water marks ranged from 3 to 32 cm for the Irma hindcast and 9–70 cm for Matthew. Then, using forcings from Hurricanes Ian and Nicole (2022), the model predicted the occurrence of dune collision, overwash, and inundation for the coast of Georgia and suggested that wave runup was impacted by the still water level and local topobathy.

Georgia↗
USGS2025

Using hydrogeologic context and water budgets to evaluate the potential for groundwater contributions to contaminants in Lake Superior

This study presents a synthesis of the hydrogeology in the U.S. Lake Superior watershed and the contribution of groundwater to the water budget of the U.S. Lake Superior basin. Much of the shoreline of Lake Superior in Minnesota and Michigan is composed of hydrogeologic units contributing very little direct groundwater discharge to the lake. Groundwater in watersheds adjacent to the lake typically flows in short, local flow systems characterized by thin glacial sediments with active groundwater flow in fractured bedrock within the top 60–90 m below land surface. The complex groundwater system in Wisconsin’s Bayfield Peninsula has the largest groundwater reservoir near the lake, characterized by thick sand and gravel glacial deposits and underlying sandstone aquifer. Although these thick sandy glacial deposits are not in direct contact with the lake at the shoreline, groundwater discharge may still be significant via subsurface exposures beyond the shoreline or flow through the underlying sandstone aquifer. Overall, most groundwater in the watershed is contributed as indirect base flow in streams around the lakeshore and comprises about 60 percent of the upland water budget. Direct groundwater flow to the shoreline contributes 2 to 9 percent of the inflow. Identifying possible contamination sources through direct sampling of groundwater would be an inefficient way to detect problems if sources are unknown, particularly for some chemicals of concern such as PFAS, pesticides, PCBs, chloride, and nutrients. Evaluating the chemical characteristics of contamination is also important to consider in evaluating how groundwater may contribute to pollution in Lake Superior.

Lake Superior↗
USGS2026

Assessing future hydrologic extremes using an integrated hydrology and river operations model in the Russian River watershed

Study region The Russian River watershed, situated in coastal, northern California, experiences hydrologic extremes, including periodic droughts and flooding. Water managers are working to maintain sustainable water supplies and environmental flows, while mitigating flood risks. Study focus This paper introduces an integrated hydrology and river operations model for the Russian River watershed. This model is distinct from models in previous studies because it represents surface-groundwater interactions and uses climate forcings to estimate dynamic water use demands that are superimposed onto both reservoir operations and water supply constraints. The model was used to examine three historical (1990–2015) and eight future (2016–2099) water use and climate change scenarios. New hydrological insights for the region The direct connection between streams and aquifers facilitated both annual aquifer replenishment by high winter streamflows and streamflow depletion by groundwater wells (19 % of pumped groundwater in alluvial aquifers from stream leakage) during critical low flow periods. Simulated streamflow changes included 59 % longer and 54 % more severe streamflow droughts, 26 % lower seasonal low streamflows, and up to 125 % higher peak streamflows, averaged over future climate and water use scenarios, suggesting increased future flood and water availability risks. Results showed the importance of reservoir operations for mitigating the impacts of increased hydroclimatic volatility, despite a decrease in reservoir reliability at Lake Mendocino, suggesting that reservoir management may be used to decrease future risks.

California↗
USGS2025

Prospectivity modeling of the NASA VIPER landing site at Mons Mouton near the Lunar South Pole

We use a high-resolution digital elevation model and a numerical thermal model to produce a variety of inputs for a water-ice prospectivity model for the Volatiles Investigating Polar Exploration Rover (VIPER) landing site. These input data are maps of topography, surface slope, surface aspect, surface curvature, maximum temperature, depth to ice stability, permanently shadowed regions (PSRs), distance to PSRs, and PSR density. This model predicts where water ice is most likely within the top meter of regolith, assuming plausible relationships between ice concentration and the various inputs. The model is designed to be adjusted in near-real time as data are collected during the VIPER mission. As such, it is a tool for both analyzing data from the mission as well as planning operations. Since the current model, at this point, relies only on orbital remote sensing, the final version will also be a tool to extrapolate the VIPER mission results across the lunar poles.

Planetary Science Journal↗
USGS2025

Assessing earthquake risks to lifeline infrastructure systems in the United States

The security and economic stability of the United States rely heavily on robust lifeline infrastructure systems and yet the risks to such systems are seldom quantified at the national scale. For example, while earthquake risks to buildings in the United States have been investigated at the national scale regularly, such risks to gas pipelines have rarely been investigated nationally. In this paper, we use examples from two critical infrastructure sectors to demonstrate (1) the nature of earthquake risks to lifeline infrastructure systems, (2) complexities involved in regional seismic risk assessments, and (3) how such risks change with time. We found that bridge risks can be underestimated by at least 64 % when viewed from repair costs instead of traffic demands and that regional risks can be underestimated by 19 % when spatial correlations of ground motion are ignored. Further, exceedance of traffic demand can be 50 times more likely to occur when viewed at the regional scale than when viewed at an individual bridge. Similarly, exceedance of repairs can be 180 times more likely to occur when viewed at the pipeline network level than at a segment-specific level. Finally, sensitivity analyses with the 2018 and 2023 USGS National Seismic Hazard Models indicate an increase in bridge risk of at least 24 % and an increase in exposed gas pipeline mileage of 43 %. The evolution of risks, complexities involved in assessments, and limited resources jointly underscore the need for more routine updates to nationwide seismic risk assessments of lifeline systems in the United States.

International Journal of Critical Infrastructure P↗
USGS2026

PFAS remediation in a bioelectrochemical system inoculated with the west branch consortium (WBC-2)

Groundwater contamination by per - and polyfluoroalkyl substances (PFAS) poses a persistent environmental and public health concern. This study evaluates a two-chambered bioelectrochemical system (BES) inoculated with the West Branch Consortium (WBC-2) for PFAS remediation. Under an applied cathodic potential of −450 mV (versus Ag/AgCl), the BES with active WBC-2 achieved >99.0% perfluorooctanesulfonic acid (PFOS) removal within 21 days in deionized water with culture medium and > 98.9% removal of PFOS, perfluorooctanoic acid (PFOA), perfluorohexanoic acid (PFHxA), and perfluorohexanesulfonic acid (PFHxS) in contaminated groundwater after 102 days. Intermediate formation (e.g., PFOA, 6:2 fluorotelomer sulfonate (6:2 FTS), perfluoropropionic acid (PFPrA), perfluorobutanoic acid (PFBA)) and background-corrected fluoride release were consistent with PFOS transformation under anaerobic reducing conditions potentially involving defluorination. Following repeated PFOS spikes (100 μg/L on Days 0, 50, and 399), PFOA, PFPrA, and PFBA accumulated over 664 days. Despite being the dominant accumulated compound, PFOA accounted for <1.8% of the total spiked PFOS mass. Minimal PFOS transformation occurred in controls without active WBC-2, highlighting the importance of microbial metabolism. Biofilm analysis revealed dense colonization of rod-shaped bacteria on carbon fiber brushes. Enrichment of Bacillus , Agrobacterium , and other low-abundance taxa suggests selective adaptation to BES and PFAS conditions. These findings highlight BES driven by electrochemically stimulated microbial activity as a promising strategy for PFAS remediation.

Journal of Water Process Engineering↗
USGS2025

A spatiotemporal deep learning approach for predicting daily air-water temperature signal coupling and identification of key watershed physical parameters in a montane watershed

Seasonal shifts from runoff to groundwater dominance influence daily headwater stream temperatures, especially where local groundwater input is strong. This input buffers temperature during hot periods, supporting cold-water habitats. Recent studies use air–water temperature signal metrics to identify zones of strong stream–groundwater connectivity. While Previous studies used air–water signal ratios as proxies for groundwater influence but were limited to specific sites and periods, without dynamic forecasting. This study is the first to forecast daily A r as a spatiotemporal signal using a Graph Convolutional Network–Long Short-Term Memory (GCN-LSTM) model. The model was trained using hydroclimate data (air temperature, precipitation, shortwave radiation, streamflow) and watershed physical features (e.g., sand content, slope). Results showed high predictive skill, achieving R 2 (NSE, RMSE) of 0.86 (0.73, 0.0004) for one-day-ahead to 0.52 (0.50, 0.0009) for seven-days ahead forecasts. Prior studies often have not explicitly incorporated spatial hydrogeologic drivers, but this model explicitly incorporates them to assess their impact on A r forecasting and stream-groundwater connectivity. Feature analysis identified mean sand, elevation, slope, clay, and TWI as key predictors of A r . Stronger groundwater signals appeared in hillslopes, elevations, and tributaries, highlighting watershed influence on streamflow. However, limitations include reliance on historical air–water temperature patterns for training and limited representation of extreme climate conditions. Despite these limitations, unlike previous studies relying on measured in-situ stream and air temperature, this study forecasts A r directly from climate and physiographic features after training, avoiding in-situ data requirements. Findings aiding predictions of stream ecosystem resilience.

New York↗
USGS2026

Top-down targeted network analysis of critical mineral commodities applied to international geochemistry database

The global demand for critical mineral commodities is rapidly increasing, making domestic production an important factor in supporting the economy and national security. Large scale, publicly available geochemical databases allow for the application of data informatics methods to interrogate critical mineral commodities data for correlations in deposit formation and distribution, particularly for identifying enrichment of multiple critical mineral commodities at the same deposit. In this study, we applied network analysis to the Critical Minerals Mapping Initiative (CMMI) ore geochemistry (Critical Minerals in Ores, CMiO) database to identify the high concentration (defined as 10× bulk crustal abundance) co-occurrence of different critical mineral commodities across a mineral system hierarchy from deposit environments to individual deposits. Identifying patterns or unique outliers in enrichment in network communities will allow for the location of secondary critical mineral commodity resources from under-utilized deposits. We find trends in the enrichment of critical mineral commodities in network-communities between the elements praseodymium (Pr), neodymium (Nd), terbium (Tb), and dysprosium (Dy) across multiple CMiO database deposit environments and groups down to specific deposit types and sites. A separate trend in network community deposition is observed as well between iridium (Ir) and platinum (Pt) in deposit environments, groups, types, and sites. Network analysis focused on critical minerals in magmatic-hydrothermal deposits identified multiple deposit sites from different deposit types within the CMiO database with concentrations of Dy, Nd, Tb, Pr, Ir, and Pt that are at least ten times greater than the crustal average. This approach can be applied to any target element(s) or deposit(s) of interest, allowing broad investigation of co-enriched critical mineral commodities.

Journal of Geochemical Exploration↗
USGS2024

Exploring the uncertainty of machine learning models and geostatistical mapping of rare earth element potential in Indiana coals, USA

Rare earth elements and yttrium (REEs) have a wide range of applications in high- and low-carbon technologies. The strategic significance of REEs has grown due to their expanding applications in manufacturing industries and the constrained availability of these essential resources. This research explores the applicability of machine learning models and their uncertainty for assessing the REE potential in coal beds using various coal parameters as inputs. The work focuses on developing a predictive model based on geological variables, excluding considerations related to potential shifts in the commodities market. The Indiana Coal Quality Database was used as the data source. The promising and unpromising indicators derived from the outlook coefficient of samples from the database were used as the REE potential indicator for machine learning classification models. The filter-based approach with bootstrap was used to evaluate the importance of the coal parameters and their prediction uncertainties. Four machine learning methods (linear discriminant analysis (LDA), random forest (RF), support vector machine (SVM), and artificial neural networks (ANN), a data balancing and augmentation approach (Synthetic Minority Over-sampling Technique), and bootstrap resampling techniques were used for building the models and evaluating their prediction capabilities under uncertainty. It was determined that the SVM bootstrap model with ten-times balanced and augmented data provided superior results compared with other models. Finally, stochastic spatial maps of the REE potential within the coal basin were generated using sequential indicator simulation. The spatial maps of the REE potential showed that a 29% area of the Indiana section of the Illinois coal basin has economic potential of REEs, with 90% confidence.

Indiana↗
USGS2000

Biodegradation of chlorinated ethenes at a karst site in middle Tennessee

This report presents results of field and laboratory investigations examining the biodegradation of chlorinated ethenes in a karst aquifer contaminated with trichloroethylene (TCE). The study site, located in Middle Tennessee, was selected because of the presence of TCE degradation byproducts in the karst aquifer and available site hydrologic and chlorinated-ethene information. Additional chemical, biological, and hydrologic data were gathered to evaluate whether the occurrence of TCE degradation byproducts in the karst aquifer was the result of biodegradation within the aquifer or simply transport into the aquifer. Geochemical analysis established that sulfate-reducing conditions, essential for reductive dechlorination of chlorinated solvents, existed in parts of the contaminated karst aquifer. Other areas of the aquifer fluctuated between anaerobic and aerobic conditions and contained compounds associated with cometabolism, such as ethane, methane, ammonia, and dissolved oxygen. A large, diverse bacteria population inhabits the contaminated aquifer. Bacteria known to biodegrade TCE and other chlorinated solvents, such as sulfate-reducers, methanotrophs, and ammonia-oxidizers, were identified from karst-aquifer water using the RNA-hybridization technique. Results from microcosms using raw karst-aquifer water found that aerobic cometabolism and anaerobic reductive-dechlorination degradation processes were possible when appropriate conditions were established in the microcosms. These chemical and biological results provide circumstantial evidence that several biodegradation processes are active in the aquifer. Additional site hydrologic information was developed to determine if appropriate conditions persist long enough in the karst aquifer for these biodegradation processes to be significant. Continuous monitoring devices placed in four wells during the spring of 1998 indicated that pH, specific conductance, dissolved oxygen, and oxidation-reduction potentials changed very little in areas isolated from active ground-water flow paths. These stable areas in the karst aquifer had geochemical conditions and bacteria conducive to reductive dechlorination of chlorinated ethenes. Other areas of the karst aquifer were associated with active ground-water flow paths and fluctuated between anaerobic and aerobic conditions in response to rain events. Associated with this dynamic environment were bacteria and geochemical conditions conducive to cometabolism. In summary, multiple lines of evidence developed from chemical, biological, and hydrologic data demonstrate that a variety of biodegradation processes are active in this karst aquifer.

Tennessee↗
USGS2023

Don’t Let Negatives Hold You Back: Accounting for Underlying Physics and Natural Distributions of Hydrothermal Systems When Selecting Negative Training Sites Leads to Better Machine Learning Predictions

Selecting negative training sites is an important challenge to resolve when utilizing machine learning (ML) for predicting hydrothermal resource favorability because ideal models would discriminate between hydrothermal systems (positives) and all types of locations without hydrothermal systems (negatives). The Nevada Machine Learning project (NVML) fit an artificial neural network to identify areas favorable for hydrothermal systems by selecting 62 negative sites where the research team had confidence that no hydrothermal resource exists. Herein, we compare the implications of the expert selection of negatives (i.e., the NVML strategy) with a random sample strategy, where it is assumed that areas outside the favorable structural ellipses defined by NVML are negative. Because hydrothermal systems are sparse, it is highly probable that, in the absence of a favorable geological structure, hydrothermal favorability is low. We compare three training strategies: 1) the positive and negative labeled examples from NVML; 2) the positive examples from NVML with randomly selected negatives in equal frequency as NVML; and 3) the positive examples from NVML with randomly selected negatives reflecting the expected natural distribution of hydrothermal systems relative to the total area. We apply these training strategies to the NVML feature data (input data) using two ML algorithms (XGBoost and logistic regression) to create six favorability maps for hydrothermal resources. When accounting for the expected natural distribution of hydrothermal systems, we find that XGBoost performs better than the NVML neural network and its negatives. Model validation was less reliable using F1 scores, a common performance metric, than comparing probability estimates at known positives, likely because of the extreme natural class imbalance and the lack of negatively labeled sites. This work demonstrates that expert selection of negatives for training in NVML likely imparted modeling bias. Accounting for the sparsity of hydrothermal systems and all the types of locations without hydrothermal systems allows us to create better models for predicting hydrothermal resource favorability.

Geothermal Resources Council Transactions↗