Geology Reports⌕ Search

SEARCH · Geology Reports

Results for “Data Series”

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 253 records · Page 14Linked to original sources
USGS2012

Estimating rate uncertainty with maximum likelihood: differences between power-law and flicker–random-walk models

Recent studies have documented that global positioning system (GPS) time series of position estimates have temporal correlations which have been modeled as a combination of power-law and white noise processes. When estimating quantities such as a constant rate from GPS time series data, the estimated uncertainties on these quantities are more realistic when using a noise model that includes temporal correlations than simply assuming temporally uncorrelated noise. However, the choice of the specific representation of correlated noise can affect the estimate of uncertainty. For many GPS time series, the background noise can be represented by either: (1) a sum of flicker and random-walk noise or, (2) as a power-law noise model that represents an average of the flicker and random-walk noise. For instance, if the underlying noise model is a combination of flicker and random-walk noise, then incorrectly choosing the power-law model could underestimate the rate uncertainty by a factor of two. Distinguishing between the two alternate noise models is difficult since the flicker component can dominate the assessment of the noise properties because it is spread over a significant portion of the measurable frequency band. But, although not necessarily detectable, the random-walk component can be a major constituent of the estimated rate uncertainty. None the less, it is possible to determine the upper bound on the random-walk noise.

Journal of Geodesy↗
USGS2017

Evaluation and use of U.S. Environmental Protection Agency Clean Watersheds Needs Survey data to quantify nutrient loads to surface water, 1978–2012

Changes in municipal and industrial point-source discharges over time have been an important factor affecting nutrient trends in many of the Nation’s streams and rivers. This report documents how three U.S. Environmental Protection Agency (EPA) national datasets—the Permit Compliance System, the Integrated Compliance Information System, and the Clean Watersheds Needs Survey—were evaluated for use in the U.S. Geological Survey National Water-Quality Assessment project to assess the causes of nutrient trends. This report also describes how a database of total nitrogen load and total phosphorous load was generated for select wastewater treatment facilities in the United States based on information reported in the EPA Clean Watersheds Needs Survey. Nutrient loads were calculated for the years 1978, 1980, 1982, 1984, 1986, 1988, 1990, 1992, 1996, 2000, 2004, 2008, and 2012 based on average nitrogen and phosphorous concentrations for reported treatment levels and on annual reported flow values. The EPA Permit Compliance System (PCS) and Integrated Compliance Information System (ICIS), which monitor point-source facility discharges, together are the Nation’s most spatially comprehensive dataset for nutrients released to surface waters. However, datasets for many individual facilities are incomplete, the PCS/ICIS historical data date back only to 1989, and historical data are available for only a limited number of facilities. Additionally, inconsistencies in facility reporting make it difficult to track or identify changes in nutrient discharges over time. Previous efforts made by the U.S. Geological Survey to “fill in” gaps in the PCS/ICIS data were based on statistical methods—missing data were filled in through the use of a statistical model based on the Standard Industrial Classification code, size, and flow class of the facility and on seasonal nutrient discharges of similar facilities. This approach was used to estimate point-source loads for a single point in time; it was not evaluated for use in generating a consistent data series over time. Another national EPA dataset that is available is the Clean Watersheds Needs Survey (CWNS), conducted every 4 years beginning 1973. The CWNS is an assessment of the capital needs of wastewater facilities to meet the water-quality goals set in the Clean Water Act. Data collected about these facilities include location and contact information for the facilities; population served; flow and treatment level of the facility; estimated capital needs to upgrade, repair, or improve facilities for water quality; and nonpoint-source best management practices. Total nitrogen and total phosphorous load calculations for each of the CWNS years were based on treatment level information and average annual outflow (in million gallons per day) from each of the facilities that had reported it. Treatment levels categories (such as Primary, Secondary, or Advanced) were substituted with average total nitrogen and total phosphorous concentrations for each treatment level based on those reported in literature. The CWNS dataset, like the PCS/ICIS dataset, has years where facilities did not report either a treatment level or an annual average outflow, or both. To fill in the data gaps, simple linear assumptions were made based on each facility’s responses to the survey in years bracketing the data gap or immediately before or after the data gap if open ended. Treatment level and flow data unique to each facility were used to complete the CWNS dataset for that facility.

Scientific Investigations Report↗
USGS2022

A water quality barometer for Chesapeake Bay: Assessing spatial and temporal patterns using long-term monitoring data

This paper develops a barometer that indexes water quality in the Chesapeake Bay and summarizes quality over spatial regions and temporal periods. The barometer has a basis in risk assessment and hydrology, and is a function of three different metrics of water quality relative to numerical criteria: relative frequency of criterion attainment; magnitude of deviation from a numerical criterion; and duration of criterion attainment. Metrics associated with these features are calculated at the station level, allowing flexibility for simultaneously evaluating multiple stressors, different designated uses, and physical characteristics of the water. The barometer score is then created as a geometric mean of the three metrics. The water quality barometer (WQB) station scores may be spatially aggregated to report habitat scores across a spectrum of spatial resolutions (e.g., management segment, tidal subsystem, or the whole tidal bay). Dissolved oxygen measurements in the Chesapeake Bay collected during summer seasons of 1985 to 2020 are used to evaluate water quality. The WQB score and its bootstrapped confidence interval are reported at the station, segment, tidal subsystem and whole tidal bay levels. Notably, water quality interpreted through application of the WQB with dissolved oxygen concentration data and averaged over the 29-year period of record is good (i.e. protects aquatic living resources) in tributaries such as the James River, Rappahannock River and others; but is not as good in areas such as the Upper Tributaries and the York River. Recent summaries indicate that while the water quality is improving in much of the bay and its tidal tributaries, however, there is an indication of decline in quality in the period 20182020, especially in the upper regions of the Bay. The barometer is designed around using the time series data produced by the Chesapeake Bay Programs annual monitoring strategy; the approach has application to other large water bodies with large scale monitoring programs with extended time series or for integrating information from environmental sensor systems.

Maryland, Virginia↗
USGS2012

TracerLPM (Version 1): An Excel® workbook for interpreting groundwater age distributions from environmental tracer data

TracerLPM is an interactive Excel® (2007 or later) workbook program for evaluating groundwater age distributions from environmental tracer data by using lumped parameter models (LPMs). Lumped parameter models are mathematical models of transport based on simplified aquifer geometry and flow configurations that account for effects of hydrodynamic dispersion or mixing within the aquifer, well bore, or discharge area. Five primary LPMs are included in the workbook: piston-flow model (PFM), exponential mixing model (EMM), exponential piston-flow model (EPM), partial exponential model (PEM), and dispersion model (DM). Binary mixing models (BMM) can be created by combining primary LPMs in various combinations. Travel time through the unsaturated zone can be included as an additional parameter. TracerLPM also allows users to enter age distributions determined from other methods, such as particle tracking results from numerical groundwater-flow models or from other LPMs not included in this program. Tracers of both young groundwater (anthropogenic atmospheric gases and isotopic substances indicating post-1940s recharge) and much older groundwater (carbon-14 and helium-4) can be interpreted simultaneously so that estimates of the groundwater age distribution for samples with a wide range of ages can be constrained. TracerLPM is organized to permit a comprehensive interpretive approach consisting of hydrogeologic conceptualization, visual examination of data and models, and best-fit parameter estimation. Groundwater age distributions can be evaluated by comparing measured and modeled tracer concentrations in two ways: (1) multiple tracers analyzed simultaneously can be evaluated against each other for concordance with modeled concentrations (tracer-tracer application) or (2) tracer time-series data can be evaluated for concordance with modeled trends (tracer-time application). Groundwater-age estimates can also be obtained for samples with a single tracer measurement at one point in time; however, prior knowledge of an appropriate LPM is required because the mean age is often non-unique. LPM output concentrations depend on model parameters and sample date. All of the LPMs have a parameter for mean age. The EPM, PEM, and DM have an additional parameter that characterizes the degree of age mixing in the sample. BMMs have a parameter for the fraction of the first component in the mixture. An LPM, together with its parameter values, provides a description of the age distribution or the fractional contribution of water for every age of recharge contained within a sample. For the PFM, the age distribution is a unit pulse at one distinct age. For the other LPMs, the age distribution can be much broader and span decades, centuries, millennia, or more. For a sample with a mixture of groundwater ages, the reported interpretation of tracer data includes the LPM name, the mean age, and the values of any other independent model parameters. TracerLPM also can be used for simulating the responses of wells, springs, streams, or other groundwater discharge receptors to nonpoint-source contaminants that are introduced in recharge, such as nitrate. This is done by combining an LPM or user-defined age distribution with information on contaminant loading at the water table. Information on historic contaminant loading can be used to help evaluate a model's ability to match real world conditions and understand observed contaminant trends, while information on future contaminant loading scenarios can be used to forecast potential contaminant trends.

Techniques and Methods↗
USGS2025

Artificial neural network multilayer perceptron models to classify California’s crops using Harmonized Landsat Sentinel (HLS) data

Advances in remote sensing and machine learning are enhancing cropland classification, vital for global food and water security. We used multispectral Harmonized Landsat 8 Sentinel-2 (HLS) 30-m data in an artificial neural network (ANN) multi-layer perceptron (MLP) model to classify five crop classes (cotton, alfalfa, tree crops, grapes, and others) in California's Central Valley. The ANN MLP model, trained on 2021 data from the United States Department of Agriculture's Cropland Data Layer, was validated by classifying crops for an independent year, 2022. Across the five crop classes, the overall accuracy was 74%. Producer's and user's accuracies ranged from 65% to 87%, with cotton achieving the highest accuracies. The study highlights the potential of using deep learning with HLS time series data for accurate global crop classification.

California↗
USGS1985

ANNIE - INTERACTIVE PROCESSING OF DATA BASES FOR HYDROLOGIC MODELS.

ANNIE is a data storage and retrieval system that was developed to reduce the time and effort required to calibrate, verify, and apply watershed models that continuously simulate water quantity and quality. Watershed models have three categories of input: parameters to describe segments of a drainage area, linkage of the segments, and time-series data. Additional goals for ANNIE include the development of software that is easily implemented on minicomputers and some microcomputers and software that has no special requirements for interactive display terminals. Another goal is for the user interaction to be based on the experience of the user so that ANNIE is helpful to the inexperienced user and yet efficient and brief for the experienced user. Finally, the code should be designed so that additional hydrologic models can easily be added to ANNIE.

Conference Paper↗
USGS2015

U.S. Geological Survey groundwater toolbox, a graphical and mapping interface for analysis of hydrologic data (version 1.0): user guide for estimation of base flow, runoff, and groundwater recharge from streamflow data

This report is a user guide for the streamflow-hydrograph analysis methods provided with version 1.0 of the U.S. Geological Survey (USGS) Groundwater Toolbox computer program. These include six hydrograph-separation methods to determine the groundwater-discharge (base-flow) and surface-runoff components of streamflow—the Base-Flow Index (BFI; Standard and Modified), HYSEP (Fixed Interval, Sliding Interval, and Local Minimum), and PART methods—and the RORA recession-curve displacement method and associated RECESS program to estimate groundwater recharge from streamflow data. The Groundwater Toolbox is a customized interface built on the nonproprietary, open source MapWindow geographic information system software. The program provides graphing, mapping, and analysis capabilities in a Microsoft Windows computing environment. In addition to the four hydrograph-analysis methods, the Groundwater Toolbox allows for the retrieval of hydrologic time-series data (streamflow, groundwater levels, and precipitation) from the USGS National Water Information System, downloading of a suite of preprocessed geographic information system coverages and meteorological data from the National Oceanic and Atmospheric Administration National Climatic Data Center, and analysis of data with several preprocessing and postprocessing utilities. With its data retrieval and analysis tools, the Groundwater Toolbox provides methods to estimate many of the components of the water budget for a hydrologic basin, including precipitation; streamflow; base flow; runoff; groundwater recharge; and total, groundwater, and near-surface evapotranspiration.

Techniques and Methods↗
USGS2009

Using a coupled groundwater/surfacewater model to predict climate-change impacts to lakes in the Trout Lake watershed, Northern Wisconsin

A major focus of the U.S. Geological Survey’s Trout Lake Water, Energy, and Biogeochemical Budgets (WEBB) project is the development of a watershed model to allow predictions of hydrologic response to future conditions including land-use and climate change. The coupled groundwater/surface-water model GSFLOW was chosen for this purpose because it could easily incorporate an existing groundwater flow model and it provides for simulation of surface-water processes. The Trout Lake watershed in northern Wisconsin is underlain by a highly conductive outwash sand aquifer. In this area, streamflow is dominated by groundwater contributions; however, surface runoff occurs during intense rainfall periods and spring snowmelt. Surface runoff also occurs locally near stream/lake areas where the unsaturated zone is thin. A diverse data set, collected from 1992 to 2007 for the Trout Lake WEBB project and the co-located and NSF-funded North Temperate Lakes LTER project, includes snowpack, solar radiation, potential evapotranspiration, lake levels, groundwater levels, and streamflow. The timeseries processing software TSPROC (Doherty 2003) was used to distill the large time series data set to a smaller set of observations and summary statistics that captured the salient hydrologic information. The timeseries processing reduced hundreds of thousands of observations to less than 5,000. Model calibration included specific predictions for several lakes in the study area using the PEST parameter estimation suite of software (Doherty 2007). The calibrated model was used to simulate the hydrologic response in the study lakes to a variety of climate change scenarios culled from the IPCC Fourth Assessment Report of the Intergovernmental Panel on Climate Change (Solomon et al. 2007). Results from the simulations indicate climate change could result in substantial changes to the lake levels and components of the hydrologic budget of a seepage lake in the flow system. For a drainage lake lower in the flow system, the impacts of climate change are diminished.

Conference Paper↗
USGS2009

Remote monitoring of tamarisk defoliation and evapotranspiration following saltcedar leaf beetle attack

Tamarisk (Tamarix spp.) has invaded riparian ecosystems throughout the Western United States, including significant portions of riparian ecosystems within U.S. National Parks and Monuments. Recently, the saltcedar leaf beetle (Diorhabda elongata) was released as a tamarisk biocontrol agent. Although initial releases have been monitored, no comprehensive program is currently in place to monitor the rapid spread of Diorhabda that has resulted from numerous subsequent releases by county and state agencies. Long term monitoring of tamarisk defoliation and its impacts on habitat and water resources is needed. This study examines the potential for using higher spatial resolution Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) data and lower spatial resolution Moderate Resolution Imaging Spectroradiometer (MODIS) data for monitoring defoliation caused by Diorhabda and subsequent changes in evapotranspiration (ET). Widespread tamarisk defoliation was observed in an eastern Utah study area during summer 2007. ASTER normalized difference vegetation index (NDVI) showed only minor changes between 2005 and 2006, but a significant drop in NDVI was found within riparian areas between 2006 and 2007. The decrease in NDVI caused by defoliation was apparent despite partial refoliation within the study area. MODIS time series data revealed that absolute decline in EVI varied by site, but that the timing of EVI decline during summer 2007 was early with respect to phenological patterns from 2001 through 2006. Defoliation caused decreases in ET values estimated from both ASTER and MODIS data. MODIS estimated ET declined earlier than in previous years, although annual ET was not significantly different than ET in previous years due to high year-to-year variability. Challenges to detection and monitoring of tamarisk defoliation include spectral mixing of tamarisk and other cover types at subpixel spatial resolution, spatial coregistration of time series images, the timing of image acquisition, and changes unrelated to defoliation in non-tamarisk land cover over time. Continued development of the techniques presented in this paper may allow monitoring the spread of Diorhabda and assessment of potential water salvage resulting from biocontrol of tamarisk. ?? 2009 Elsevier Inc.

Remote Sensing of Environment↗
USGS2006

Estimating crustal heterogeneity from double-difference tomography

Seismic velocity parameters in limited, but heterogeneous volumes can be inferred using a double-difference tomographic algorithm, but to obtain meaningful results accuracy must be maintained at every step of the computation. MONTEILLER et al. (2005) have devised a double-difference tomographic algorithm that takes full advantage of the accuracy of cross-spectral time-delays of large correlated event sets. This algorithm performs an accurate computation of theoretical travel-time delays in heterogeneous media and applies a suitable inversion scheme based on optimization theory. When applied to Kilauea Volcano, in Hawaii, the double-difference tomography approach shows significant and coherent changes to the velocity model in the well-resolved volumes beneath the Kilauea caldera and the upper east rift. In this paper, we first compare the results obtained using MONTEILLER et al.'s algorithm with those obtained using the classic travel-time tomographic approach. Then, we evaluated the effect of using data series of different accuracies, such as handpicked arrival-time differences ("picking differences"), on the results produced by double-difference tomographic algorithms. We show that picking differences have a non-Gaussian probability density function (pdf). Using a hyperbolic secant pdf instead of a Gaussian pdf allows improvement of the double-difference tomographic result when using picking difference data. We completed our study by investigating the use of spatially discontinuous time-delay data. ?? Birkha??user Verlag, Basel, 2006.

Pure and Applied Geophysics↗
USGS2003

Processing large remote sensing image data sets on Beowulf clusters

High-performance computing is often concerned with the speed at which floating- point calculations can be performed. The architectures of many parallel computers and/or their network topologies are based on these investigations. Often, benchmarks resulting from these investigations are compiled with little regard to how a large dataset would move about in these systems. This part of the Beowulf study addresses that concern by looking at specific applications software and system-level modifications. Applications include an implementation of a smoothing filter for time-series data, a parallel implementation of the decision tree algorithm used in the Landcover Characterization project, a parallel Kriging algorithm used to fit point data collected in the field on invasive species to a regular grid, and modifications to the Beowulf project's resampling algorithm to handle larger, higher resolution datasets at a national scale. Systems-level investigations include a feasibility study on Flat Neighborhood Networks and modifications of that concept with Parallel File Systems.

Open-File Report↗
USGS2022

From data to interpretable models: Machine learning for soil moisture forecasting

Soil moisture is critical to agricultural business, ecosystem health, and certain hydrologically driven natural disasters. Monitoring data, though, is prone to instrumental noise, wide ranging extrema, and nonstationary response to rainfall where ground conditions change. Furthermore, existing soil moisture models generally forecast poorly for time periods greater than a few hours. To improve such forecasts, we introduce two data-driven models, the Naive Accumulative Representation (NAR) and the Additive Exponential Accumulative Representation (AEAR). Both of these models are rooted in deterministic, physically based hydrology, and we study their capabilities in forecasting soilmoisture over time periods longer than a fewhours. Learned model parameters represent the physically based unsaturated hydrological redistribution processes of gravity and suction. We validate our models using soil moisture and rainfall time series data collected from a steep gradient, post-wildfire site in southern California. Data analysis is complicated by rapid landscape change observed in steep, burned hillslopes in response to even small to moderate rain events. The proposed NAR and AEAR models are, in forecasting experiments, shown to be competitive with several established and state-of-the-art baselines. The AEAR model fits the data well for three distinct soil textures at variable depths below the ground surface (5, 15, and 30 cm). Similar robust results are demonstrated in controlled, laboratory-based experiments. Our AEAR model includes readily interpretable hydrologic parameters and provides more accurate forecasts than existing models for time horizons of 10–24 h. Such extended periods of warning for natural disasters, such as floods and landslides, provide actionable knowledge to reduce loss of life and property.

International Journal of Data Science and Analytic↗
USGS2003

National Geochronological Database

The National Geochronological Data Base (NGDB) was established by the United States Geological Survey (USGS) to collect and organize published isotopic (also known as radiometric) ages of rocks in the United States. The NGDB (originally known as the Radioactive Age Data Base, RADB) was started in 1974. A committee appointed by the Director of the USGS was given the mission to investigate the feasibility of compiling the published radiometric ages for the United States into a computerized data bank for ready access by the user community. A successful pilot program, which was conducted in 1975 and 1976 for the State of Wyoming, led to a decision to proceed with the compilation of the entire United States. For each dated rock sample reported in published literature, a record containing information on sample location, rock description, analytical data, age, interpretation, and literature citation was constructed and included in the NGDB. The NGDB was originally constructed and maintained on a mainframe computer, and later converted to a Helix Express relational database maintained on an Apple Macintosh desktop computer. The NGDB and a program to search the data files were published and distributed on Compact Disc-Read Only Memory (CD-ROM) in standard ISO 9660 format as USGS Digital Data Series DDS-14 (Zartman and others, 1995). As of May 1994, the NGDB consisted of more than 18,000 records containing over 30,000 individual ages, which is believed to represent approximately one-half the number of ages published for the United States through 1991. Because the organizational unit responsible for maintaining the database was abolished in 1996, and because we wanted to provide the data in more usable formats, we have reformatted the data, checked and edited the information in some records, and provided this online version of the NGDB. This report describes the changes made to the data and formats, and provides instructions for the use of the database in geographic information system (GIS) applications. The data are provided in .mdb (Microsoft Access), .xls (Microsoft Excel), and .txt (tab-separated value) formats. We also provide a single non-relational file that contains a subset of the data for ease of use.

Open-File Report↗
USGS2008

Seismic monitoring to assess performance of structures In near‐real time: Recent progress

Earlier papers have described how observed data from classical accelerometers deployed in structures or from differential GPS with high sampling ratios deployed at roofs of tall buildings can be configured to establish seismic health monitoring of structures. In these configurations, drift ratios 1 are the main parametric indicator of damage condition of a structure or component of a structure. Real‐time measurement of displacements are acquired either by double integration of accelerometer time‐series data, or by directly using GPS. Recorded sensor data is then related to the performance level of a building. Performance‐based design method stipulates that for a building the amplitude of relative displacement of the roof of a building (with respect to its base) indicates its performance. Usually, drift ratio is computed using relative displacement between two consecutive floors. When accelerometers are used, a specific software is used to compute displacements and drift ratios in realtime by double integration of accelerometer data from several floors. However, GPS‐measured relative displacements are limited to being acquired only at the roof with respect to its reference base. Thus, computed drift ratio is the average drift ratio for the whole building. Until recently, the validity of measurements using GPS was limited to long‐period structures ( T >1  s ) "> ( T >1 s ) because GPS systems readily available were limited to 10–20 samples per seconds (sps) capability. However, presently, up to 50 sps differential GPS systems are available on the market and have been successfully used to monitor drift ratios [1,2]—thus enabling future usefulness of GPS to all types of structures. Several levels of threshold drift ratios can be postulated in order to make decisions for inspections and/or occupancy. Experience with data acquired from both accelerometers and GPS deployments indicates that they are reliable and provide pragmatic alternatives to alert the owners and other authorized parties to make informed decisions and select choices for pre‐defined actions following significant events.

Conference Paper↗
USGS2009

Seismic monitoring to assess performance of structures in near-real time: Recent progress

Earlier papers have described how observed data from classical accelerometers deployed in structures or from differential GPS with high sampling ratios deployed at roofs of tall buildings can be configured to establish seismic health monitoring of structures. In these configurations, drift ratios are the main parametric indicator of damage condition of a structure or component of a structure. Real-time measurement of displacements are acquired either by double integration of accelerometer time-series data, or by directly using GPS. Recorded sensor data is then related to the performance level of a building. Performance-based design method stipulates that for a building the amplitude of relative displacement of the roof of a building (with respect to its base) indicates its performance. Usually, drift ratio is computed using relative displacement between two consecutive floors. When accelerometers are used, determination of displacement is possible by strategically deploying them at a select number of pairs of consecutive floors. For these determinations, software is used to compute displacements and drift ratios in real-time by double integration of accelerometer data. However, GPS-measured relative displacements are limited to being acquired only at the roof with respect to its reference base. Thus, computed drift ratio is the average drift ratio for the whole building. Until recently, the validity of measurements using GPS was limited to long-period structures (T>1 s) because GPS systems readily available were limited to 10–20 samples per seconds (sps) capability. However, presently, up to 50 sps differential GPS systems are available on the market and have been successfully used to monitor drift ratios [1, Panagitou et al. (Seismic Response of ReinForced Concrete Wall Buildings, 2006)], (Restrepo, pers. comm. 2007) – thus enabling future usefulness of GPS to all types of structures. Several levels of threshold drift ratios can be postulated in order to make decisions for inspections and/or occupancy. Experience with data acquired from both accelerometers and GPS deployments indicates that they are reliable and provide pragmatic alternatives to alert the owners and other authorized parties to make informed decisions and select choices for pre-defined actions following significant events. Furthermore, recent adoption of such methods by financial and industrial enterprises is testimony to their viability.

Book chapter↗
USGS2010

Fundamental changes in the activity of the natrocarbonatite volcano Oldoinyo Lengai, Tanzania

On September 4, 2007, after 25 years of effusive natrocarbonatite eruptions, the eruptive activity of Oldoinyo Lengai (OL), N Tanzania, changed abruptly to episodic explosive eruptions. This transition was preceded by a voluminous lava eruption in March 2006, a year of quiescence, resumption of natrocarbonatite eruptions in June 2007, and a volcano-tectonic earthquake swarm in July 2007. Despite the lack of ground-based monitoring, the evolution in OL eruption dynamics is documented based on the available field observations, ASTER and MODIS satellite images, and almost-daily photos provided by local pilots. Satellite data enabled identification of a phase of voluminous lava effusion in the 2 weeks prior to the onset of explosive eruptions. After the onset, the activity varied from 100 m high ash jets to 2–15 km high violent, steady or unsteady, eruption columns dispersing ash to 100 km distance. The explosive eruptions built up a ∼400 m wide, ∼75 m high intra-crater pyroclastic cone. Time series data for eruption column height show distinct peaks at the end of September 2007 and February 2008, the latter being associated with the first pyroclastic flows to be documented at OL. Chemical analyses of the erupted products, presented in a companion paper (Keller et al. 2010 ), show that the 2007–2008 explosive eruptions are associated with an undersaturated carbonated silicate melt. This new phase of explosive eruptions provides constraints on the factors causing the transition from natrocarbonatite effusive eruptions to explosive eruptions of carbonated nephelinite magma, observed repetitively in the last 100 years at OL.

Oldoinyo Lengai↗
USGS2014

Practical limitations on the use of diurnal temperature signals to quantify groundwater upwelling

Groundwater upwelling to streams creates unique habitat by influencing stream water quality and temperature; upwelling zones also serve as vectors for contamination when groundwater is degraded. Temperature time series data acquired along vertical profiles in the streambed have been applied to simple analytical models to determine rates of vertical fluid flux. These models are based on the downward propagation characteristics (amplitude attenuation and phase-lag) of the surface diurnal signal. Despite the popularity of these models, there are few published characterizations of moderate-to-strong upwelling. We attribute this limitation to the thermodynamics of upwelling, under which the downward conductive signal transport from the streambed interface occurs opposite the upward advective fluid flux. Governing equations describing the advection–diffusion of heat within the streambed predict that under upwelling conditions, signal amplitude attenuation will increase, but, counterintuitively, phase-lag will decrease. Therefore the extinction (measurable) depth of the diurnal signal is very shallow, but phase lag is also short, yielding low signal to noise ratio and poor model sensitivity. Conversely, amplitude attenuation over similar sensor spacing is strong, yielding greater potential model sensitivity. Here we present streambed thermal time series over a range of moderate to strong upwelling sites in the Quashnet River, Cape Cod, Massachusetts. The predicted inverse relationship between phase-lag and rate of upwelling was observed in the field data over a range of conditions, but the observed phase-lags were consistently shorter than predicted. Analytical solutions for fluid flux based on signal amplitude attenuation return results consistent with numerical models and physical seepage meters, but the phase-lag analytical model results are generally unreasonable. Through numerical modeling we explore reasons why phase-lag may have been over-predicted by the analytical models, and develop guiding relations of diurnal temperature signal extinction depth based on stream diurnal signal amplitude, upwelling magnitude, and streambed thermal properties that will be useful in designing future experiments.

Journal of Hydrology↗
USGS2022

Regularizing priors for Bayesian VAR applications to large ecological datasets

Using multi-species time series data has long been of interest for estimating inter-specific interactions with vector autoregressive models (VAR) and state space VAR models (VARSS); these methods are also described in the ecological literature as multivariate autoregressive models (MAR, MARSS). To date, most studies have used these approaches on relatively small food webs where the total number of interactions to be estimated is relatively small. However, as the number of species or functional groups increases, the length of the time series must also increase to provide enough degrees of freedom with which to estimate the pairwise interactions. To address this issue, we use Bayesian methods to explore the potential benefits of using regularized priors, such as Laplace and regularized horseshoe, on estimating interspecific interactions with VAR and VARSS models. We first perform a large-scale simulation study, examining the performance of alternative priors across various levels of observation error. Results from these simulations show that for sparse matrices, the regularized horseshoe prior minimizes the bias and variance across all inter-specific interactions. We then apply the Bayesian VAR model with regularized priors to a output from a large marine food web model (37 species) from the west coast of the USA. Results from this analysis indicate that regularization improves predictive performance of the VAR model, while still identifying important inter-specific interactions.

PeerJ↗