SEARCH · USGS Science

Results for “Journal of Applied Volcanology”

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

SourceResearch connectionFrom yearThrough year

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

1,659 records · Page 17Linked to original sources

USGS2026

Abandonment of the Upper Devonian Greenland Gap Group and Scherr Formation and revision of the Upper Devonian Brallier and Foreknobs Formations in the central Valley and Ridge Province

This study revises the lithostratigraphic framework of the Upper Devonian interval traditionally assigned to the Greenland Gap Group across the central Appalachian Valley and Ridge Province. The work aims to modernize and standardize lithostratigraphic nomenclature, establish a new reference section and demonstrate how the revised stratigraphy improves edge-matching of 1:24,000 scale geologic maps and supports compilation mapping at scales of 1:100,000 and larger. The revision eliminates the names Greenland Gap Group, Scherr Formation and the Minnehaha Springs Member of the Scherr Formation; reassigns all strata previously designated as Scherr Formation by Dennison (1970) to the upper Brallier Formation; and abandons the basal Mallow Member of the Foreknobs Formation, placing its strata within the upper Brallier as originally defined by Butts (1918). The contact between the Brallier and Foreknobs formations is placed at the base of the first mappable, ridge-forming package of fine- to coarse-grained, cross-bedded, sandstone beds, often containing rounded quartz pebbles with minor interbeds of shale and siltstone. This contact may be gradational in places but, even in absence of good exposure, can usually be distinguished topographically in recently produced lidar-derived imagery as having elevated relief due to the presence of more resistant, compositionally mature coarse-grained sandstone-rich strata. Applying this criterion for mapping the contact between the Brallier and Foreknobs formations has resulted in reconciliation of mismatches of geologic contacts along several 7.5-minute quadrangle boundaries in the states of Virginia, West Virginia, Maryland and Pennsylvania. A new reference section at Baker, West Virginia showcases the contacts between the Harrell Shale, Brallier Formation, Foreknobs Formation and Hampshire Formation. A digital outcrop model of the reference section is provided for future preservation.

Maryland, Pennsylvania, Virginia, West Virginia↗

USGS2024

ECCOE Landsat quarterly Calibration and Validation report—Quarter 1, 2024

Executive Summary The U.S. Geological Survey Earth Resources Observation and Science Calibration and Validation (Cal/Val) Center of Excellence (ECCOE) focuses on improving the accuracy, precision, calibration, and product quality of remote-sensing data, leveraging years of multiscale optical system geometric and radiometric calibration and characterization experience. The ECCOE Landsat Cal/Val Team continually monitors the geometric and radiometric performance of active Landsat missions and makes calibration adjustments, as needed, to maintain data quality at the highest level. This report provides observed geometric and radiometric analysis results for Landsats 8 and 9 for quarter 1 (January–March), 2024. All data used to compile the Cal/Val analysis results presented in this report are freely available from the U.S. Geological Survey EarthExplorer website: https://earthexplorer.usgs.gov . This quarterly report is the third to include analysis results for Landsat 9, which was launched in September 2021. The inclusion of Landsat 9 analysis results was dependent on two factors: a complete reprocessing of the Landsat 9 data archive and enough time elapsing to begin formulating lifetime trends. In April 2023, all Landsat 9 image data acquired since the satellite’s launch were reprocessed to take advantage of calibration updates identified by the ECCOE Landsat Cal/Val Team. Additional information about the Landsat 9 reprocessing effort is available at https://www.usgs.gov/landsat-missions/news/upcoming-reprocessing-all-landsat-9-data . Additional information about Landsat 9 prelaunch, commissioning, and early on-orbit imaging performance is available at https://www.mdpi.com/journal/remotesensing/special_issues/15B4V2K92K . This quarterly report is the first to not include analysis results for Landsat 7 because Enhanced Thematic Mapper Plus imaging was suspended on January 19, 2024, after the satellite transitioned into full sunlight. The satellite has been drifting since early 2022 after being lowered from the nominal orbit altitude, and the transition into full sunlight is a result of the satellite operating in its extended science mission. Additional information about the imaging suspension is available at https://www.usgs.gov/landsat-missions/news/landsat-7-imaging-suspended . Additional information about the Landsat 7 extended science mission is available at https://www.usgs.gov/landsat-missions/landsat-7-extended-science-mission .

Open-File Report↗

USGS2024

ECCOE Landsat quarterly Calibration and Validation report—Quarter 2, 2024

Executive Summary The U.S. Geological Survey Earth Resources Observation and Science Calibration and Validation (Cal/Val) Center of Excellence (ECCOE) focuses on improving the accuracy, precision, calibration, and product quality of remote-sensing data, leveraging years of multiscale optical system geometric and radiometric calibration and characterization experience. The ECCOE Landsat Cal/Val Team continually monitors the geometric and radiometric performance of active Landsat missions and makes calibration adjustments, as needed, to maintain data quality at the highest level. This report provides observed geometric and radiometric analysis results for Landsats 8 and 9 for quarter 2 (April–June) of 2024. All data used to compile the Cal/Val analysis results presented in this report are freely available from the U.S. Geological Survey EarthExplorer website: https://earthexplorer.usgs.gov . This is the fourth quarterly report to include analysis results for Landsat 9, which was launched in September 2021. The inclusion of Landsat 9 analysis results was dependent on two factors: a complete reprocessing of the Landsat 9 data archive and enough time elapsing to begin formulating lifetime trends. In April 2023, all Landsat 9 image data acquired since the satellite’s launch were reprocessed to take advantage of calibration updates identified by the ECCOE Landsat Cal/Val Team. Additional information about the Landsat 9 reprocessing effort is available at https://www.usgs.gov/landsat-missions/news/upcoming-reprocessing-all-landsat-9-data . Additional information about Landsat 9 prelaunch, commissioning, and early on-orbit imaging performance is available at https://www.mdpi.com/journal/remotesensing/special_issues/15B4V2K92K . This is the second quarterly report that does not include analysis results for Landsat 7 because Enhanced Thematic Mapper Plus imaging was suspended on January 19, 2024, after the satellite transitioned into full sunlight. The satellite has been drifting since early 2022 when it was lowered from the nominal orbit altitude, and the transition into full sunlight is a result of the satellite operating in its extended science mission. Additional information about the imaging suspension is available at https://www.usgs.gov/landsat-missions/news/landsat-7-imaging-suspended . Additional information about the Landsat 7 extended science mission is available at https://www.usgs.gov/landsat-missions/landsat-7-extended-science-mission .

Open-File Report↗

USGS2026

Incorporating location uncertainty improves inference with stop-level North American Breeding Bird Survey data

Ecological models should account for uncertainty to be most effective and useful. Yet, uncertainty from model covariates—unlike that from other sources, such as sampling error or process variability—is seldom explicitly incorporated. This can cause underestimates of uncertainty to cascade through model parameter estimates, predictions, and downstream uses. Burner et al. proposed a method for quantifying uncertainty in covariates and incorporating it into models using informative Bayesian priors. This method was applied to stop-level Breeding Bird Survey (BBS) analyses, where land cover uncertainty at each stop arises from substantial stop location uncertainty. A limited validation of model-estimated land cover, using stops with known locations, indicated the method’s potential effectiveness, but it was not rigorously evaluated. We conduct a robust simulation-based test, generating stop locations, extracting land cover, and simulating bird communities across 210 BBS routes in the upper Midwest. We compare 3 models: a “known” model with true land cover, a “naive” model assuming consistent 800-m stop spacing, and a “full” model using informative priors to estimate land cover. Species parameter estimates and predicted prevalence patterns across gradients in land cover from the full model approached those of the known model and were substantially closer to the true values used in simulations relative to those from the naive model. Naive model parameters were more biased relative to the other models, and credible intervals of predicted species prevalence rarely included the true simulated values. The full model also produced land cover covariate estimates closer to true simulation values relative to the mean informative priors. Our results show that, for the BBS, informative priors enable more accurate stop-level analyses despite location uncertainty. In contrast, naive models that ignore this uncertainty yield poor inferences. More broadly, we demonstrate empirically the utility of informative priors to account for covariate uncertainty in ecological models.

Michigan, Minnesota, Wisconson↗

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

SURF: An automated method for building nonplanar 3D fault models from earthquake hypocenters

Accurately characterizing 3D fault geometry is vital for improving our understanding of earthquake behavior and informing the development of seismic hazard models. Despite their importance, subsurface fault structures tend to be poorly constrained because of limitations in observational data. Improvements to the seismic networks and earthquake detection algorithms have increased the precision and volume of earthquake catalogs, which help illuminate detailed subsurface fault structure and provide the most direct information available about fault geometries at depth. We present a Python package to automate generating 3D fault geometries directly from hypocentral seismicity patterns. This method begins with clustering events based on their spatial density, identifying coherent patterns. Nearby clusters are then merged based on the similarity of their orientations. We fit nonplanar surfaces using support vector regression to balance surface accuracy with minimal deviations from planarity. The fault models are output as quadrilateral meshes at user‐defined resolution. In the process of generating the 3D fault surfaces, we compute the spatial density of seismicity around the surface and the planarity as quantitative metrics of the model outputs. As a proof of concept, we apply this approach to the San Andreas–Calaveras fault junction region and the 2019 Ridgecrest earthquake sequence, both in California, which contain complex subparallel faults well defined at the Earth’s surface and abundant microseismicity. These case studies demonstrate the method’s ability to model complex fault structures, including long continuous fault surfaces, crossing faults, variably dipping segments, and subparallel faults. We test the method on both standard network catalogs and double‐difference relocated catalogs. We find that our seismicity‐based fault model results align with published 3D models that incorporate additional constraints and interpretations ( Plesch et al. , 2020 ; Aagaard and Hirakawa, 2021 ). This workflow provides a low‐user‐input solution for estimating fault geometries at depth from earthquake catalogs.

California↗

USGS2026

Regional models for postfire debris-flow likelihood and rainfall thresholds across the western United States

The U.S. Geological Survey (USGS) uses an empirical model developed with logistic regression (the ‘M1’ model) to rapidly assess debris-flow likelihood and to identify quantitative rainfall thresholds for debris flows after wildfire in the western United States. The M1 model was calibrated to a debris-flow inventory from southern California (United States) and has been applied throughout the western United States. Limited spatial coverage in the calibration dataset has motivated evaluation of M1 model accuracy outside the calibration region (e.g., the Sierra Nevada or the eastern Cascade Range, United States). Previous test cases showed that M1 overpredicts debris-flow likelihood and underpredicts rainfall thresholds for some locations (e.g., Arizona, northern California, Colorado, New Mexico, United States). We sought to improve the regional applicability of a debris-flow likelihood model by expanding the debris-flow inventory used for calibration, testing multiple potential models and generating an updated model framework. The updated inventory includes 3788 observations from 67 burned areas paired with short duration rainfall ratios. The updated model framework consists of a modified model structure and sets of coefficients calibrated separately to the entire updated inventory and to subsets of the inventory that intersect three Environmental Protection Agency (EPA) Level 2 ecoregions (Mediterranean California, Upper Gila Mountains and Western Cordillera). Comparisons of predictions from the updated models with observed rainfall and debris-flow activity show that the updated models outperform the M1 model by ~15%–60% and improve the uniformity of predictive performance across the western United States. The updated models also reduce false positive rates relative to M1 and generate rainfall thresholds that are better aligned with relative differences in regional climatology and debris-flow activity.

Arizona, California, Colorado, Idaho, Montana, Nev↗

USGS2026

Using a temporary emigration model to estimate abundance of stream fishes from hybrid removal surveys with and without block nets

Monitoring programs are often faced with a decision to allocate resources into either robust spatiotemporal coverage to estimate a population index (e.g., not true abundance) or confirming closed sampling conditions (e.g., with block nets) for an unbiased population estimate at the cost of spatiotemporal coverage. However, making accurate and precise abundance estimates at robust spatiotemporal scales is possible when combining open and closed sampling designs with integrated modeling techniques. We used simulations and a case study of backpack electrofishing surveys in the Santa Ana River, California to test the efficacy of an integrated abundance model (temporary emigration model, TE) to estimate abundance of fishes using removal sampling methods with a hybrid sampling design (sampling with and without block nets during removal sampling). We found that the TE model performed well under most modeling scenarios (sample size, amount of closure violation, number of samples collected during closure), although at least a few samples with block nets were necessary for all parameters to be estimable. When applied to fish surveys in the Santa Ana River, we found that catch of the fishes fit to the TE model (Santa Ana Sucker, Arroyo Chub, Channel Catfish, Largemouth Bass, Yellow Bullhead) showed little evidence that the closure assumption was violated when block nets were not used. Additionally, we found that the abundance of non-native fishes negatively affected the abundance of the native Santa Ana Sucker, which was also found to adversely affect the native fish’s access to critical habitat consisting of gravel and cobble substrate. Our results indicate that the TE model presents a viable solution to common sampling problems that impact many monitoring programs, where precise and accurate population estimates can be made at large spatiotemporal scales even when most samples violate the closure assumption.

California↗

USGS2025

Deterministic physics-based earthquake sequence simulators match empirical ground-motion models and enable extrapolation to data poor regimes: Application to multifault multimechanism ruptures

We use the deterministic earthquake simulator RSQSim to generate complex sequences of ruptures on fault systems used for hazard assessment. We show that the source motions combined with a wave propagation code create surface ground motions that fall within the range of epistemic uncertainties for the Next Generation Attenuation‐West2 set of empirical models. We show the model is well calibrated where there are good data constraints, and has good correspondence in regions with fewer data constraints. We show magnitude, distance, and mechanism dependence all arising naturally from the same underlying friction. The deterministic physics‐based approach provides an opportunity for better understanding the physical origins of ground motions. For example, we find that reduced stress drops in shallow layers relative to constant stress drop with depth lead to peak ground velocities in the near field that better match empirical models. The simulators may also provide better extrapolations into regimes that are poorly empirically constrained by data because physics, rather than surface shaking data parameterizations, is underlying the extrapolations. Having shown the model is credible, we apply it to a problem where observations are lacking. We examine the case of crustal faults above a shallow subduction interface seen to break coseismically in simulations of the New Zealand fault system. These types of events were left out of consideration in the most recent New Zealand national seismic hazard model due to the modeling complexity and lack of observational data to constrain ground‐motion models (GMMs). Here, we show that in the model, by breaking up the coseismic crustal and interface rupturing fault motions into two separate subevents, and then recombining the resulting ground‐motion measures in a square‐root‐of‐sum‐of‐squares incoherent manner, we reproduce well the ground‐motion measures from the full event rupture. This provides a new method for extrapolating GMMs to more complex multifault ruptures.

Seismological Research Letters↗

USGS2026

A targeted approach for mapping groundwater discharge to surface water and fish thermal refuge in four Lake Ontario tributaries

The duration, magnitude, and frequency of heatwaves are predicted to increase in the coming decades, a combination that can reduce the survival of many fish species. Across the world, there is broad interest in identifying thermal refuge for heat-intolerant fish species and exploring opportunities to enhance or protect these areas. Because deeper groundwater maintains a relatively constant temperature, groundwater-influenced areas along streams can provide cool-water refuge for fish during periods of extreme heat. A targeted approach was developed for identifying existing cold-water zones and areas of substantial groundwater discharge in four high priority Lake Ontario tributaries. Our approach included: (1) predicting where groundwater discharge is most likely with a simple geospatial model and (2) using model predictions to select field sites for intensive high-resolution study, including ground-based mapping of groundwater features (springs, seeps, tributaries) as well as drone-based optical and thermal infrared surveys. Results from field sites were used to both verify model performance and map different types and aerial extents of thermal anomalies. Geospatial modelling successfully predicted regions of widespread groundwater upwelling, later verified and mapped by field and drone surveys. Comparison of model and field survey results further highlighted specific geospatial layers, such as soil/bedrock types and topographic wetness index, as being particularly useful for predicting groundwater influence on streams in the study area. In addition, a comparison of geospatial model results with a model of fish abundances along the studied streams showed significant positive correlations for many heat-intolerant fish species over a wide geographic area. The approach developed in this study can be applied to other watersheds to highlight areas of probable groundwater discharge and could be used by fishery and water resource managers to support cold-water fish habitat management decision-making and resource conservation.

New York↗

USGS2025

Assessing nonpoint-source uranium pollution in an irrigated stream-aquifer system

Uranium (U) in rocks and soils of arid and semi-arid environments can be mobilized by irrigation and fertilization, posing environmental and health risks. Elevated U, along with selenium (Se) and nitrate (NO 3 ) co-constituents, necessitates careful monitoring and management. We developed a distributed-parameter numerical model to assess U pollution in an irrigated stream-aquifer system, applying it to a 552 km 2 region in Colorado's Lower Arkansas River Valley (LARV) over 14 years. A MODFLOW model, describing groundwater and stream flow, was coupled with an RT3D-OTIS model to portray reactive U transport. Calibration using the PESTPP-iES iterative ensemble smoother (iES) software indicated good agreement with observed U concentrations. The model revealed substantial and variable U levels across the LARV, highlighting potential hotspots and possible contributing factors, such as geological composition of the bedrock and near-surface shale and aquifer sediments derived from them, irrigation practices, and riparian landscape. U levels exceed the chronic standard (85th percentile = 30 μg/L, set by the US Environmental Protection Agency), which is the permissible regulatory threshold, in groundwater across 44 % of the region and along the river by an average factor of 2.9. Simulated average U concentrations in the non-riparian aquifer and river are 124 μg/L and 60 μg/L, respectively, compared with 112 μg/L and 62 μg/L for measured values. The average 85th percentile U concentration is 222 μg/L in the aquifer and 82 μg/L in the river. Average simulated U mass loading to the river is 0.17 kg/day per km, compared to an estimated 0.23 kg/day per km. Findings provide a baseline for comparing future simulated outcomes of alternative best management practices (BMPs) for U pollution mitigation and offer a methodology applicable to other irrigated regions.

Colorado↗

USGS2025

Earthquake magnitude and source parameter estimation with a distributed acoustic sensing dataset in the Gorda subduction zone

Distributed acoustic sensing (DAS) systems offer a cost‐effective way to create large‐scale strainmeter arrays for seismological applications using fiber‐optic cables. DAS‐based strain measurements are known to be influenced by various factors, bringing into question their general reliability for accurate earthquake characterization. A 15‐km‐long DAS deployment in northern California was operational within 3 days of the 2022 M w 6.4 Ferndale earthquake and ran continuously throughout the aftershock sequence. We utilize these aftershock data to validate DAS‐based strain measurements in two ways. We first test the accuracy of DAS‐based magnitude estimates from peak dynamic strains by comparing them with magnitude and attenuation scaling relations derived independently from traditional borehole strainmeter (BSM) data. We demonstrate that DAS‐based magnitudes are comparable to BSM‐based magnitudes when corrections for variations in site response along the fiber‐optic cable are properly made. Magnitude errors are spatially correlated, potentially because of factors such as finite‐fault effects (e.g., stress drop) or more complex, unmodeled path attenuation or because of wave propagation effects in heterogeneous media. We then apply more advanced source characterization methodology to the DAS data using a time‐domain empirical Green’s function (EGF) deconvolution approach to measure details of the moment rate history. The EGF approach using DAS data depends on careful treatment of distorting factors such as anthropogenic sources of noise and optical phase wrapping but successfully isolates source spectra for moderate‐magnitude earthquakes: source spectral ratios obtained from DAS data, broadband seismometer data, and BSM data in the same region show consistent results, revealing differences in directivity and spectral shape among earthquakes. Although further research is needed to refine source‐time‐function estimation techniques for DAS data, particularly for larger magnitude events, these case studies demonstrate the clear potential of DAS for earthquake source characterization.

California↗

USGS2024

Conceptualization and simulation of groundwater flow and groundwater availability in the Boone and Roubidoux aquifers in northeastern Oklahoma, 1980–2017

Oklahoma Groundwater Law (Oklahoma Statute § 82-1020.5) requires that the Oklahoma Water Resources Board conduct hydrologic investigations to determine the maximum annual yield for the State’s groundwater basins. The Boone and Roubidoux aquifers (also known as the Springfield Plateau aquifer and Ozark aquifer, respectively) are bedrock aquifers that extend from northeastern Oklahoma into Kansas, Arkansas, and Missouri. At present (2024), the Oklahoma Water Resources Board has yet to legally issue orders for the final determination of maximum annual yields for the Boone and Roubidoux aquifers. To support determination of a maximum annual yield, the U.S. Geological Survey, in cooperation with the Oklahoma Water Resources Board, developed a hydrogeologic framework, a conceptual groundwater-flow model, and a calibrated numerical groundwater-flow model for the Boone and Roubidoux aquifers. Three types of groundwater-availability scenarios were simulated by using the calibrated numerical model. These scenarios were used to (1) estimate equal-proportionate-share groundwater withdrawal rates (groundwater withdrawal applied equally over the aquifer), (2) quantify the potential effects of projected groundwater withdrawals on groundwater storage over a 50-year period, and (3) simulate the potential effects of a hypothetical 10-year drought. For the Boone aquifer, equal-proportionate-share groundwater withdrawal rates were 1.10, 0.98, and 0.96 acre-feet per acre per year for the 20-, 40-, and 50-year scenarios, respectively. For the Roubidoux aquifer, equal-proportionate-share groundwater withdrawal rates were 1.76, 1.34, and 1.25 acre-feet per acre per year for the 20-, 40-, and 50-year simulations, respectively. For the 50-year scenarios, stream seepage was minimally affected. Over the 10-year drought scenario, groundwater storage in the Boone and Roubidoux aquifers decreased by 660,451 acre-feet (6.7 percent) and 508,472 acre-feet (1.0 percent), respectively.

Oklahoma↗

USGS2026

The 2025 Puerto Rico and Virgin Islands U.S. National Seismic Hazard Model Update: Ground motion model selection and comparison

We evaluate, select, and describe the ground-motion models (GMMs) used in the 2025 update of the U.S. National Seismic Hazard Model (NSHM) for Puerto Rico and the U.S. Virgin Islands (PRVI). We identify the most appropriate models that align with GMM selection criteria for use in the PRVI region to improve the accuracy of seismic hazard assessments. The update incorporates globally applicable GMMs suited for the active crustal and subduction earthquakes in the region. We include region-specific adjustments to these GMMs derived from local site response analyses derived from ground motion records. The unadjusted and regionally-corrected GMMs are combined to create a robust model for predicting median ground motion. The model integrates epistemic uncertainty through a median ground motion logic tree that accounts for variations in magnitude and distance. This study compares the GMMs selected for the 2025 PRVI NSHM, including both as-provided and regionally adjusted NGA-West2 and NGA-Subduction models, with those used in the 2003 PRVI NSHM. We evaluate how changes in model selection, weighting, aleatory variability, and epistemic uncertainty influence seismic hazard estimates. Trends with distance, magnitude, and spectral period are analyzed to evaluate how the scaling behavior of the newer GMMs differs from that of earlier models. Relative to the GMMs used in the 2003 NSHM for this region, the 2025 models generally predict lower ground motions. Comparisons with additional GMMs indicate that the adjustments applied for PRVI are consistent with regional-specific modifications developed elsewhere globally. The increase in aleatory variability and epistemic uncertainty in the 2025 update results in a notable increase in hazard levels from these wider uncertainty bounds. These changes can result in as much as a 10%–20% variation in probabilistic ground motion at the 2% in 50 years exceedance level for hazard maps computed across the region for representative site classes and periods.

Puerto Rico, Virgin Islands↗

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↗

USGS2025

Capturing the uncertainty of seismicity observations in earthquake rate estimates: Implications for probabilistic seismic hazard analysis and the USGS National Seismic Hazard Model

The rate of earthquakes in a region is a fundamental input to Probabilistic Seismic Hazard Analysis. We present a Monte Carlo method for computing that rate from seismicity catalogs while including a range of data and analysis uncertainties. This method is applied to regions for which the b value is assumed to be spatially invariant. Each region is broken down into epochs for which each epoch is estimated to have a uniform magnitude of completeness ( ⁠⁠ M c ). The distribution of earthquake rates for M ≥ M c is determined for each epoch by considering the Poisson likelihood of rates given the number of observed earthquakes with M ≥ M c ⁠ . We use a Monte Carlo process to include the uncertainty in b , ⁠ M c , and individual event magnitudes. The result for each epoch is the joint distribution of the Poisson rate of earthquakes with magnitudes larger than the minimum value used to calculate hazard ( ⁠⁠ M 1 ) and the Gutenberg–Richter b values, which control the extrapolation to other magnitudes. The rate for each region is either the duration‐weighted average over the epochs or, to better capture temporal variations, we also consider mixture models. The mixture models also provide an avenue to allow temporal variations in b values. To implement this joint distribution in a logic tree, we use the mean and 95% confidence branches, each of which is parameterized with an M ≥ M 1 rate and b value. We explore different ways of defining those branches, as well as non‐Gutenberg–Richter branches, and their impact on hazard estimates. The mean hazard, but not the fractiles, is robust with respect to these choices. To illustrate these new methods, we use synthetic data and catalogs from recent U.S. Geological Survey National Seismic Hazard Models for the Central and Eastern United States and for Puerto Rico and the U.S. Virgin Islands.

central and eastern United States, Puerto Rico, U.↗

USGS2024

Comparing subduction ground-motion models to observations for Cascadia

We evaluate Cascadia subduction ground-motion models (GMMs), considered for the 2023 US National Seismic Hazard Model (NSHM) update, by comparing observations to model predictions. The observations comprise regional recordings from intraslab earthquakes, including contributions from 2021 and 2022 events in southern Cascadia and global records from interface earthquakes. Since the 2018 NSHM update, new GMMs for Cascadia have been published by the Next Generation Attenuation (NGA)-Subduction Project that require independent evaluation. In the regional intraslab comparisons, we highlight a characteristic frequency dependence for Cascadia data, with short periods having lower ground motions and longer periods being comparable to other subduction zones. We evaluate differences in northern and southern Cascadia and find that the NGA-Subduction GMMs developed using southern Cascadia data perform better in this region than the model that did not consider these data. We compare ground-motion variability in Cascadia with the NGA-Subduction model predictions and find differences at short periods ( T = 0.1 s) due to the use of global versus regional data in the development of these models. Moreover, the within-event component of aleatory variability from the GMMs overpredicts the standard deviation of Cascadia recordings at very short periods ( T < 0.05 s). Using global interface earthquakes as a proxy to evaluate the Cascadia GMMs, we find long-period overprediction from a simulation-based GMM and some of the empirical GMMs. When comparing recent observations, we find a similar misfit to GMMs and the 2010 and 2022 Ferndale earthquakes. Finally, we observe different basin amplification factors arising in different subsets of the data, which indicate that differences in basin factors between empirical GMMs could arise from the data selection choices by the developers. As part of evaluating the regional basin terms, we apply basin amplification factors from the magnitude 9 Cascadia earthquake simulations to the empirical GMMs for interface earthquakes. The comparisons presented in this study indicate that the NGA-Subduction GMMs for Cascadia perform well relative to observations and older subduction GMMs.

British Columbia, California, Oregon, Washington↗

USGS2026

Navigating uncertainty and competing objectives: Spring chinook salmon recovery in the upper Willamette river

Globally, anadromous fish populations are threatened with extinction due to multiple factors, including river impoundments that block migration, widespread alteration of physical habitat, temperature, and flow regimes, commercial and recreational harvest, changes in biological communities, and long-term climate trends. In this synthesis, we describe how physical and biological factors, policy and law, and public resource allocation affect management of spring-run Chinook Salmon ( Oncorhynchus tshawytscha ) in the Upper Willamette River (UWR), Oregon, USA. Efforts to restore salmon populations often involve trade-offs, for example, water management decisions must account for flood control, agricultural and municipal water use, and management of other species. UWR Chinook Salmon recovery efforts are also governed by multiple laws, policies, and action agencies that in some cases have incongruent objectives. Competing and conflicting objectives, combined with uncertainty regarding the outcomes of proposed management actions, create significant challenges for decision makers. Resource limitations, regulatory constraints, and incongruent objectives have created challenges for managers in the UWR, but opportunities exist for improved adaptive management and collaboration, which could build trust and engagement among stakeholders and ultimately support the successful implementation of decision tools. A suite of decision support tools that have been used in the Willamette River basin and elsewhere across the range of Pacific salmon have the potential to enhance decision processes. These tools can be effective when decision makers and stakeholders commit to long-term, collaborative utilization of the tools for decision making, and when available resources and regulatory environments are conducive to implementation and adaptive refinement of both the decision-support tools and management plans. The general biological and institutional principles discussed for UWR Chinook Salmon in this work apply more generally to populations of anadromous fishes across their range, especially to populations in highly regulated river systems.

Oregon↗