Geology Reports⌕ Search

SEARCH · Geology Reports

Results for “Applied Sciences”

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.

At least 721 records · Page 40Linked to original sources
USGS2024

Evolutionary perspectives on thiamine supplementation of managed Pacific salmonid populations

Thiamine deficiency complex (TDC) has been identified in an ever-expanding list of species and populations. In many documented occurrences of TDC in fishes, juvenile mortality can be high—up to 90% at the population level. Such sweeping demographic losses and concomitant decreases in genetic diversity due to TDC can be prevented by treating pre-spawn females or fertilized eggs with supplemental thiamine. However, some fisheries managers are hesitant to widely apply thiamine treatments due to the potential for unforeseen evolutionary consequences. With these concerns in mind, we first review the existing data regarding genetic adaptation to low-thiamine conditions and provide perspectives on evolution-informed treatment strategies with specific population examples. We also provide practical treatment information, consider the potential logistical constraints of thiamine supplementation, and explore the consequences of deciding against supplementation. Until new evidence bolsters or refutes the genetic adaptation hypothesis, we suggest that TDC mitigation strategies should be designed to support maximum population genetic diversity through thiamine supplementation.

Canadian Journal of Fisheries and Aquatic Sciences↗
USGS2016

Bivalve grazing can shape phytoplankton communities

The ability of bivalve filter feeders to limit phytoplankton biomass in shallow waters is well-documented, but the role of bivalves in shaping phytoplankton communities is not. The coupled effect of bivalve grazing at the sediment-water interface and sinking of phytoplankton cells to that bottom filtration zone could influence the relative biomass of sinking (diatoms) and non-sinking phytoplankton. Simulations with a pseudo-2D numerical model showed that benthic filter feeding can interact with sinking to alter diatom:non-diatom ratios. Cases with the smallest proportion of diatom biomass were those with the fastest sinking speeds and strongest bivalve grazing rates. Hydrodynamics modulated the coupled sinking-grazing influence on phytoplankton communities. For example, in simulations with persistent stratification, the non-sinking forms accumulated in the surface layer away from bottom grazers while the sinking forms dropped out of the surface layer toward bottom grazers. Tidal-scale stratification also influenced vertical gradients of the two groups in opposite ways. The model was applied to Suisun Bay, a low-salinity habitat of the San Francisco Bay system that was transformed by the introduction of the exotic clam Potamocorbula amurensis . Simulation results for this Bay were similar to (but more muted than) those for generic habitats, indicating that P. amurensis grazing could have caused a disproportionate loss of diatoms after its introduction. Our model simulations suggest bivalve grazing affects both phytoplankton biomass and community composition in shallow waters. We view these results as hypotheses to be tested with experiments and more complex modeling approaches.

California↗
USGS1990

Overview of the effects and influence of the activity of Mount St. Helens in the 1980s

The cataclysmic eruption of Mount St. Helens on May 18, 1980, made an enormous impact on the science of volcanology. The eruption was in daylight in clear weather, which provided an unprecedented opportunity to investigate relations among observations, products, and effects of a large explosive eruption. The May 18 events and subsequent activity stimulated perhaps the most intensive studies ever made at an active composite volcano, leading to greatly enhanced insights into both geologic and hydrologic processes operative in explosive volcanism. The eruption also disrupted much of the social and economic fabric of the Pacific Northwest. Volcanologists were called upon to explain the activity, in layman's terms, to government and corporate officials, the news media, schools, and the public at large. People eventually learned to live with the volcano and its uncertainties, and volcanologists better learned their role in helping society deal with a major natural disaster. Difficulties encountered at volcanic crises elsewhere in the world in the 1980s demonstrate that these are hard lessons. In future years, a paramount challenge for scientists will be to help society apply what has been learned at Mount St. Helens to crises both nearby and far away.

Washington↗
USGS2008

Applying the land use portfolio model to estimate natural-hazard loss and risk — A hypothetical demonstration for Ventura County, California

With costs of natural disasters skyrocketing and populations increasingly settling in areas vulnerable to natural hazards, society is challenged to better allocate its limited risk-reduction resources. In 2000, Congress passed the Disaster Mitigation Act, amending the Robert T. Stafford Disaster Relief and Emergency Assistance Act (Robert T. Stafford Disaster Relief and Emergency Assistance Act, Pub. L. 93-288, 1988; Federal Emergency Management Agency, 2002, 2008b; Disaster Mitigation Act, 2000), mandating that State, local, and tribal communities prepare natural-hazard mitigation plans to qualify for pre-disaster mitigation grants and post-disaster aid. The Federal Emergency Management Agency (FEMA) was assigned to coordinate and implement hazard-mitigation programs, and it published information about specific mitigation-plan requirements and the mechanisms (through the Hazard Mitigation Grant Program-HMGP) for distributing funds (Federal Emergency Management Agency, 2002). FEMA requires that each community develop a mitigation strategy outlining long-term goals to reduce natural-hazard vulnerability, mitigation objectives and specific actions to reduce the impacts of natural hazards, and an implementation plan for those actions. The implementation plan should explain methods for prioritizing, implementing, and administering the actions, along with a 'cost-benefit review' justifying the prioritization. FEMA, along with the National Institute of Building Sciences (NIBS), supported the development of HAZUS ('Hazards U.S.'), a geospatial natural-hazards loss-estimation tool, to help communities quantify potential losses and to aid in the selection and prioritization of mitigation actions. HAZUS was expanded to a multiple-hazard version, HAZUS-MH, that combines population, building, and natural-hazard science and economic data and models to estimate physical damages, replacement costs, and business interruption for specific natural-hazard scenarios. HAZUS-MH currently performs analyses for earthquakes, floods, and hurricane wind. HAZUS-MH loss estimates, however, do not account for some uncertainties associated with the specific natural-hazard scenarios, such as the likelihood of occurrence within a particular time horizon or the effectiveness of alternative risk-reduction options. Because of the uncertainties involved, it is challenging to make informative decisions about how to cost-effectively reduce risk from natural-hazard events. Risk analysis is one approach that decision-makers can use to evaluate alternative risk-reduction choices when outcomes are unknown. The Land Use Portfolio Model (LUPM), developed by the U.S. Geological Survey (USGS), is a geospatial scenario-based tool that incorporates hazard-event uncertainties to support risk analysis. The LUPM offers an approach to estimate and compare risks and returns from investments in risk-reduction measures. This paper describes and demonstrates a hypothetical application of the LUPM for Ventura County, California, and examines the challenges involved in developing decision tools that provide quantitative methods to estimate losses and analyze risk from natural hazards.

California↗
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↗
USGS2015

An evaluation of unsupervised and supervised learning algorithms for clustering landscape types in the United States

Knowledge of landscape type can inform cartographic generalization of hydrographic features, because landscape characteristics provide an important geographic context that affects variation in channel geometry, flow pattern, and network configuration. Landscape types are characterized by expansive spatial gradients, lacking abrupt changes between adjacent classes; and as having a limited number of outliers that might confound classification. The US Geological Survey (USGS) is exploring methods to automate generalization of features in the National Hydrography Data set (NHD), to associate specific sequences of processing operations and parameters with specific landscape characteristics, thus obviating manual selection of a unique processing strategy for every NHD watershed unit. A chronology of methods to delineate physiographic regions for the United States is described, including a recent maximum likelihood classification based on seven input variables. This research compares unsupervised and supervised algorithms applied to these seven input variables, to evaluate and possibly refine the recent classification. Evaluation metrics for unsupervised methods include the Davies–Bouldin index, the Silhouette index, and the Dunn index as well as quantization and topographic error metrics. Cross validation and misclassification rate analysis are used to evaluate supervised classification methods. The paper reports the comparative analysis and its impact on the selection of landscape regions. The compared solutions show problems in areas of high landscape diversity. There is some indication that additional input variables, additional classes, or more sophisticated methods can refine the existing classification.

Cartography and Geographic Information Science↗
USGS2012

Semi-discrete biomass dynamic modeling: an improved approach for assessing fish stock responses to pulsed harvest events

Continuous harvest over an annual period is a common assumption of continuous biomass dynamics models (CBDMs); however, fish are frequently harvested in a discrete manner. We developed semidiscrete biomass dynamics models (SDBDMs) that allow discrete harvest events and evaluated differences between CBDMs and SDBDMs using an equilibrium yield analysis with varying levels of fishing mortality ( F ). Equilibrium fishery yields for CBDMs and SDBDMS were similar at low fishing mortalities and diverged as F approached and exceeded maximum sustained yield ( F MSY ). Discrete harvest resulted in lower equilibrium yields at high levels of F relative to continuous harvest. The effect of applying harvest continuously when it was in fact discrete was evaluated by fitting CBDMs and SDBDMs to time series data generated from a hypothetical fish stock undergoing discrete harvest and evaluating parameter estimates bias. Violating the assumption of continuous harvest resulted in biased parameter estimates for CBDM while SDBDM parameter estimates were unbiased. Biased parameter estimates resulted in biased biological reference points derived from CBDMs. Semidiscrete BDMs outperformed continuous BDMs and should be used when harvest is discrete, when the time and magnitude of harvest are known, and when F is greater than F MSY .

Canadian Journal of Fisheries and Aquatic Sciences↗
USGS1995

Removal of algae by the zebra mussel ( Dreissena polymorpha ) population in western Lake Erie: a bioenergetics approach

A bioenergetics model for growth of a zebra mussel ( Dreissena polymorpha ) individual was verified with observations on zebra mussel growth in western Lake Erie. The bioenergetics model was then applied to the zebra mussel population in the western basin of Lake Erie to estimate the removal of phytoplankton by mussels. According to the modeling results, the zebra mussel population consumed 5.0 million tonnes of phytoplankton, while 1.4 million tonnes of phytoplankton was deposited in pseudofeces from the mussels. Thus, a total of 6.4 ± 2.4 million tonnes of phytoplankton was removed from the water column by zebra mussel in western Lake Erie during 1990. Primary production was estimated to be 24.8 million tonnes; therefore, zebra mussel removed the equivalent of 26 ± 10% of the primary production for western Lake Erie.

Canadian Journal of Fisheries and Aquatic Sciences↗
USGS2020

Science plan for improving three-dimensional seismic velocity models in the San Francisco Bay region, 2019–24

This five-year science plan outlines short-term and long-term goals for improving three-dimensional seismic velocity models in the greater San Francisco Bay region as well as how to foster a community effort in reaching those goals. The short-term goals focus on improving the current U.S. Geological Survey San Francisco Bay region geologic and seismic velocity model using existing data. The long-term goals focus on acquiring new data and leveraging better analytic tools to improve the model and characterize the uncertainty. The plan describes opportunities for contributions by members of the community to develop these seismic velocity models, provides current and potential users with general information on where efforts will likely be focused to improve these models and how new versions of the models will be released, and outlines funding needs and obstacles for improving and maintaining such models. Several aspects of this plan, including how to foster a community effort, are independent of the geographic region and apply to other similar efforts.

California↗
USGS2006

Scaling local species-habitat relations to the larger landscape with a hierarchical spatial count model

Much of what is known about avian species-habitat relations has been derived from studies of birds at local scales. It is entirely unclear whether the relations observed at these scales translate to the larger landscape in a predictable linear fashion. We derived habitat models and mapped predicted abundances for three forest bird species of eastern North America using bird counts, environmental variables, and hierarchical models applied at three spatial scales. Our purpose was to understand habitat associations at multiple spatial scales and create predictive abundance maps for purposes of conservation planning at a landscape scale given the constraint that the variables used in this exercise were derived from local-level studies. Our models indicated a substantial influence of landscape context for all species, many of which were counter to reported associations at finer spatial extents. We found land cover composition provided the greatest contribution to the relative explained variance in counts for all three species; spatial structure was second in importance. No single spatial scale dominated any model, indicating that these species are responding to factors at multiple spatial scales. For purposes of conservation planning, areas of predicted high abundance should be investigated to evaluate the conservation potential of the landscape in their general vicinity. In addition, the models and spatial patterns of abundance among species suggest locations where conservation actions may benefit more than one species. ?? 2006 Springer Science+Business Media B.V.

Landscape Ecology↗
USGS2012

Quantitative estimation of climatic parameters from vegetation data in North America by the mutual climatic range technique

The mutual climatic range (MCR) technique is perhaps the most widely used method for estimating past climatic parameters from fossil assemblages, largely because it can be conducted on a simple list of the taxa present in an assemblage. When applied to plant macrofossil data, this unweighted approach (MCR un ) will frequently identify a large range for a given climatic parameter where the species in an assemblage can theoretically live together. To narrow this range, we devised a new weighted approach (MCR wt ) that employs information from the modern relations between climatic parameters and plant distributions to lessen the influence of the "tails" of the distributions of the climatic data associated with the taxa in an assemblage. To assess the performance of the MCR approaches, we applied them to a set of modern climatic data and plant distributions on a 25-km grid for North America, and compared observed and estimated climatic values for each grid point. In general, MCR wt was superior to MCR un in providing smaller anomalies, less bias, and better correlations between observed and estimated values. However, by the same measures, the results of Modern Analog Technique (MAT) approaches were superior to MCR wt . Although this might be reason to favor MAT approaches, they are based on assumptions that may not be valid for paleoclimatic reconstructions, including that: 1) the absence of a taxon from a fossil sample is meaningful, 2) plant associations were largely unaffected by past changes in either levels of atmospheric carbon dioxide or in the seasonal distributions of solar radiation, and 3) plant associations of the past are adequately represented on the modern landscape. To illustrate the application of these MCR and MAT approaches to paleoclimatic reconstructions, we applied them to a Pleistocene paleobotanical assemblage from the western United States. From our examinations of the estimates of modern and past climates from vegetation assemblages, we conclude that the MCR un technique provides reliable and unbiased estimates of the ranges of possible climatic conditions that can reasonably be associated with these assemblages. The application of MCR wt and MAT approaches can further constrain these estimates and may provide a systematic way to assess uncertainty. The data sets required for MCR analyses in North America are provided in a parallel publication.

Quaternary Science Reviews↗
USGS2000

Occurrence and load of selected herbicides and metabolites in the lower Mississippi River

Analyses of water samples collected from the Mississippi River at Baton Rouge, Louisiana, during 1991–1997 indicate that hundreds of metric tons of herbicides and herbicide metabolites are being discharged annually to the Gulf of Mexico. Atrazine, metolachlor, and the ethane-sulfonic acid metabolite of alachlor (alachlor ESA) were the most frequently detected herbicides and, in general, were present in the largest concentrations. Almost 80% of the annual herbicide load to the Gulf of Mexico occurred during the growing season from May to August. The concentrations and loads of alachlor in the Mississippi River decreased dramatically after 1993 in response to decreased use in the basin. In contrast, the concentrations and loads of acetochlor increased after 1994, reflecting its role as a replacement for alachlor. The peak annual herbicide load occurred in 1993, when approximately 640 metric tons (t) of atrazine, 320 t of cyanazine, 215 t of metolachlor, 53 t of simazine, and 50 t of alachlor were discharged to the Gulf of Mexico. The annual loads of atrazine and cyanazine were generally 1–2% of the amount annually applied in the Mississippi River drainage basin; the annual loads of acetochlor, alachlor, and metolachlor were generally less than 1%. Despite a reduction in atrazine use, historical data do not indicate a long-term downward trend in the atrazine load to the Gulf of Mexico. Although a relation ( r 2 =0.62) exists between the atrazine load and stream discharge during May to August, variations in herbicide use and rainfall patterns within subbasins can have a large effect on herbicide loads in the Mississippi River Basin and probably explain a large part of the annual variation in atrazine load to the Gulf of Mexico.

Louisana, Mississippi↗
USGS2019

The dream and the reality: Meeting decision-making time frames while incorporating ecosystem and economic models into management strategy evaluation

Atlantic herring ( Clupea harengus ) in the Northwest Atlantic have been managed with interim harvest control rules (HCRs). A stakeholder-driven management strategy evaluation (MSE) was conducted that incorporated a broad range of objectives. The MSE process was completed within 1 year. Constant catch, conditional constant catch, and a biomass-based (BB) HCR with a 15% restriction on the interannual change in the quota could achieve more stable yields than BB HCRs without such restrictions, but could not attain as high of yields and resulted in more negative outcomes for terns ( Sterna hirundo ; a predator of herring). A similar range of performance could be achieved by applying a BB HCR annually every 3 years or every 5 years. Predators (i.e., dogfish ( Squalus acanthias ), bluefin tuna ( Thunnus thynnus ), and terns) were generally insensitive to the range of HCRs. While median net revenues were sensitive to some HCRs, time series analysis suggests that most HCRs produced a stable equilibrium of net revenue. To meet management needs, some aspects of the simulations were less than might be considered scientifically ideal, but using “models of intermediate complexity” were informative for managers and formed a foundation for future improvements.

Northwest Atlantic↗
USGS1999

Geologic map of the Galaxias quadrangle (MTM 35217) of Mars

The Galaxias region (MTM 35217) is one of a series of 1:500,000-scale science study areas on Mars sponsored by NASA's Planetary Geology and Geophysics Program. Situated near the northern limit of lava flows associated with Elysium Mons, this region includes a mixture of volcanic and nonvolcanic terrains. The region is also of interest for the fluvial systems that originate along the distal margins of the Elysium lava flows. Resolution of Viking Orbiter images used to prepare the base map ranges from 40 to 160 m/pixel. High-resolution frames (40 to 80 m/pixel) are found in the southeastern part of the map area and along the north edge of the quadrangle, but over half the quadrangle is included in medium-resolution frames (150 m/pixel). Two 8 m/pixel, very high resolution scenes are available (see fig. 1). Interpretation is complicated by variable resolution and sun angles that vary from east to west illumination on different images. Mapping methods and principles are adapted from those developed for lunar photogeologic mapping by Shoemaker and Hackman (1962), refined by Wilhelms (1972), and successfully applied by many workers to a variety of planetary surfaces. Mapping units are distinguished by topography and texture and are ranked by relative age on the basis of superposition and transection relations. Material units are assigned to time-stratigraphic systems defined by Scott and Carr (1978) and Tanaka (1986). This area is included within earlier maps that used Mariner 9 images at 1:5,000,000 scale (Elston, 1979) and globally at 1:25,000,000 scale (Scott and Carr, 1978). Regional maps based on the much higher resolutions of Viking Orbiter allowed more detailed discrimination of materials by Greeley and Guest (1987) at 1:15,000,000 scale and Tanaka and others (1992) at 1:5,000,000 scale. Some map units on this 1:500,000-scale map correspond to, or are partially equivalent to, units on the larger scale maps of Greeley and Guest (1987) and Tanaka and others (1992). Established terminology is used where feasible, but the scale of this map requires that some new units be introduced and that some previous terminology be redefined. Photogeologic methods are limited; therefore, more than one geologic explanation is given for some material units that do not readily lend themselves to an unequivocal interpretation.

IMAP↗
USGS2010

Volcanic hazards and their mitigation: progress and problems

At the beginning of the twentieth century, volcanology began to emerge as a modern science as a result of increased interest in eruptive phenomena following some of the worst volcanic disasters in recorded history: Krakatau (Indonesia) in 1883 and Mont Pelée (Martinique), Soufrière (St. Vincent), and Santa María (Guatemala) in 1902. Volcanology is again experiencing a period of heightened public awareness and scientific growth in the 1980s, the worst period since 1902 in terms of volcanic disasters and crises. A review of hazards mitigation approaches and techniques indicates that significant advances have been made in hazards assessment, volcano monitoring, and eruption forecasting. For example, the remarkable accuracy of the predictions of dome-building events at Mount St. Helens since June 1980 is unprecedented. Yet a predictive capability for more voluminous and explosive eruptions still has not been achieved. Studies of magma-induced seismicity and ground deformation continue to provide the most systematic and reliable data for early detection of precursors to eruptions and shallow intrusions. In addition, some other geophysical monitoring techniques and geochemical methods have been refined and are being more widely applied and tested. Comparison of the four major volcanic disasters of the 1980s (Mount St. Helens, U.S.A. (1980), El Chichón, Mexico (1982); Galunggung, Indonesia (1982); and Nevado del Ruíz, Colombia (1985) illustrates the importance of predisaster geoscience studies, volcanic hazards assessments, volcano monitoring, contingency planning, and effective communications between scientists and authorities. The death toll (>22,000) from the Ruíz catastrophe probably could have been greatly reduced; the reasons for the tragically ineffective implementation of evacuation measures are still unclear and puzzling in view of the fact that sufficient warnings were given. The most pressing problem in the mitigation of volcanic and associated hazards on a global scale is that most of the world's dangerous volcanoes are in densely populated countries that lack the economic and scientific resources or the political will to adequately study and monitor them. This problem afflicts both developed and developing countries, but it is especially acute for the latter. The greatest advances in volcanic hazards mitigation in the near future are most likely to be achieved by wider application of existing technology to poorly understood and studied volcanoes, rather than by refinements or new discoveries in technology alone.

Reviews of Geophysics↗
USGS2010

Global positioning system and associated technologies in animal behaviour and ecological research

Biologists can equip animals with global positioning system (GPS) technology to obtain accurate (less than or equal to 30 m) locations that can be combined with sensor data to study animal behaviour and ecology. We provide the background of GPS techniques that have been used to gather data for wildlife studies. We review how GPS has been integrated into functional systems with data storage, data transfer, power supplies, packaging and sensor technologies to collect temperature, activity, proximity and mortality data from terrestrial species and birds. GPS 'rapid fixing' technologies combined with sensors provide location, dive frequency and duration profiles, and underwater acoustic information for the study of marine species. We examine how these rapid fixing technologies may be applied to terrestrial and avian applications. We discuss positional data quality and the capability for high-frequency sampling associated with GPS locations. We present alternatives for storing and retrieving data by using dataloggers (biologging), radio-frequency download systems (e.g. very high frequency, spread spectrum), integration of GPS with other satellite systems (e.g. Argos, Globalstar) and potential new data recovery technologies (e.g. network nodes). GPS is one component among many rapidly evolving technologies. Therefore, we recommend that users and suppliers interact to ensure the availability of appropriate equipment to meet animal research objectives.

Philosophical Transactions of the Royal Society B:↗
USGS2003

Photodegradation of roxarsone in poultry litter leachates

Arsenic compounds have been used extensively in agriculture in the US for applications ranging from cotton herbicides to animal feed supplements. Roxarsone (3-nitro-4-hydroxyphenylarsonic acid), in particular, is used widely in poultry production to control coccidial intestinal parasites. It is excreted unchanged in the manure and introduced into the environment when litter is applied to farmland as fertilizer. Although the toxicity of roxarsone is less than that of inorganic arsenic, roxarsone can degrade, biotically and abiotically, to produce more toxic inorganic forms of arsenic, such as arsenite and arsenate. Experiments were conducted on aqueous litter leachates to test the stability of roxarsone under different conditions. Laboratory experiments have shown that arsenite can be cleaved photolytically from the roxarsone moiety at pH 4-8 and that the degradation rate increases with increasing pH. Furthermore, the rate of photodegradation increases with nitrate and natural organic matter concentration, reactants that are commonly found in poultry-litter-water leachates. Additional photochemical reactions rapidly oxidize the cleaved arsenite to arsenate. The formation of arsenate is not entirely undesirable, because it is less mobile in soil systems and less toxic than arsenite. A possible mechanism for the degradation of roxarsone in poultry litter leachates is proposed. The results suggest that poultry litter storage and field application practices could affect the degradation of roxarsone and subsequent mobilization of inorganic arsenic species. ?? 2002 Elsevier Science B.V. All rights reserved.

Science of the Total Environment↗
USGS2026

Evaluating machine learning approaches to identify and predict oil and gas produced water lithium concentrations

Recently, the demand for battery-grade lithium has substantially increased, largely due to electrification of the transportation sector. The search for new lithium sources has turned to produced waters (frequently brines), a large-volume wastewater by-product of oil and gas extraction. Geochemical analysis indicates the presence of varying concentrations of lithium from produced water samples collected across the United States and represented in the U.S. Geological Survey’s National Produced Water Geochemical Database, as well as mixtures of Marcellus Shale produced water included in the Pennsylvania Department of Environmental Protection’s Oil and Gas Well Waste Reports. We first examined whether the geochemical signature of the lithium-bearing produced waters is sufficiently distinct so that machine learning (ML) can be used to correctly classify samples to the formation of origin. The produced water sample data used to assess classification accuracy were from the Marcellus Shale, Utica Shale and Point Pleasant Formation (Utica), and Smackover Formation oil and gas wells. Further, we evaluated the potential for ML to accurately classify Marcellus Shale produced water spatially (i.e., northeast versus southwest Pennsylvania). We then investigated whether ML algorithms applied to a suite of geochemical concentration data (i.e. Ba, Br, Cl, K, Mg, Sr) may be used to predict the lithium concentration of an unknown sample. Finally, we applied an estimated economic lithium grade cutoff of 150 milligrams per liter (mg/l) and assessed the utility of ML to predict whether a produced water sample would fall above or below the grade cutoff based on the suite of geochemical parameters. Four machine learning algorithms—Random Forest (RF), Gradient Boosting Trees (GBT), Extreme Boosting (XGBoost), and Deep Neural Networks (DNN) were assessed. This study successfully demonstrates that all four machine learning methods can precisely and accurately estimate lithium concentrations and geologic formation classification. The products of this study contribute to the growing body of knowledge aimed at expanding the lithium resource base within the United States.

Alabama, Arkansas, Florida, Georgia, Louisiana, Mi↗