Geology topics

William H. Asquith

Publications and source records attributed to William H. Asquith.

At least 19 recordsLinked to original sources

USGS2026

Altitude of the potentiometric surface and depth to water in the Mississippi River Valley alluvial aquifer, spring 2024

Potentiometric-surface and depth-to-water maps for spring 2024 were created for the Mississippi River Valley alluvial aquifer using groundwater-altitude data from 1,151 wells completed in the Mississippi River Valley alluvial aquifer and from the altitude of the top of the water surface in area rivers from 160 U.S. Geological Survey and U.S. Army Corps of Engineers streamgages. The potentiometric-surface and depth-to-water maps for 2024 were created to support investigations to characterize the Mississippi River Valley alluvial aquifer as part of the U.S. Geological Survey Water Availability and Use Science Program. Sufficient data were available to map the potentiometric surface and depth to water of the Mississippi River Valley alluvial aquifer for spring 2024 for about 83 percent of the aquifer area. The potentiometric contours ranged from 0 to 330 feet (ft) above the North American Vertical Datum of 1988. The regional direction of the groundwater gradient was generally to the south-southwest, except in areas of groundwater-altitude depressions, where the groundwater gradient direction was into the depression, and near rivers, where the groundwater gradient direction was either from the aquifer to the river or from the river into the aquifer. Groundwater depressions in the potentiometric-surface map are in the lower one-half of the Cache region and in most of the Grand Prairie and Delta regions. Depth to water by well in the Mississippi River Valley alluvial aquifer in spring 2024 ranged from 0.27 ft above land surface to 145.65 ft below land surface.

Arkansas, Illinois, Kentucky, Louisiana, Mississip↗

USGS2026

Assessment of long-term trends in streamflow statistics within and near the Mobile Bay and Perdido Bay watersheds, United States, 1950–2022

The U.S. Geological Survey, in cooperation with the Gulf Coast Ecosystem Restoration Council, assessed monotonic trends for a variety of streamflow statistics for 69 long-term U.S. Geological Survey streamgages within either the Mobile Bay or Perdido Bay watersheds that were active through at least at the end of calendar year 2019. Long-term data were defined for this investigation as having at least 50 years of cumulative record within the period since January 1, 1950, with a requirement for a complete record of streamflow during the 2010s (2010–19). The 69 streamgages have at least 54 years and as many as 73 years of daily mean streamflow data; the median period of record is 72 years; and 15 of the streamgages are identified as “major nodes” on the basis of the criteria described. The occurrence of statistically monotonic significant trends for the 69 streamgages at the 0.05 significance level is spatially shown for six statistics. For the major node streamgages, the study depicts (1) time-series graphics of annual mean, annual harmonic mean, decadal 10th, 50th, and 90th-percentile streamflows, and (2) a variation on Quantile-Kendall plots of Kendall’s tau and streamflow nonexceedance probabilities for each of the 365 days of a year. Trend assessment synthesis shows that, except for a few streamgages with relatively greater counts of statistically significant trends than others, the majority (about 93 percent) of individual trend tests indicate no trend in the streamflow and ecological metrics considered.

Alabama, Florida, Georgia, Louisiana, Mississippi,↗

USGS2026

Multiple machine-learning estimation of groundwater levels and trends for the regional Mississippi River Valley alluvial aquifer

The Mississippi River Valley alluvial aquifer provides irrigation, public, and domestic water supplies across the south-central United States. Declining groundwater levels require improved characterization of changing conditions. Traditional potentiometric-surface mapping does not use all available water-level data or quantify uncertainty. To address these limitations, we developed a data-driven multiple machine-learning (MML) framework delivered through two open-source R packages. The covMRVAgen1 software assembles covariates to 155,960 monthly groundwater levels from 57,695 wells; the mmlMRVAgen1 software trains Cubist and Random Forest models, blends them, and makes 1-kilometer gridded predictions of monthly potentiometric surfaces for the period January 1980–December 2022. The MML approach provides a methodological foundation for region-scale spatiotemporal groundwater prediction and uncertainty quantification, generating 90-percent prediction limits with appropriate empirical coverage. Model performance is acceptable, with a root-mean-square error of about 4.2 feet, standard deviation of 24.82 feet, and a normalized Nash–Sutcliffe efficiency of 0.973.

Arkansas, Illinois, Louisiana, Mississippi, Missou↗

USGS2025

Altitude of the potentiometric surface and depth to water in the Mississippi River Valley alluvial aquifer, spring 2022

Potentiometric-surface and depth-to-water maps for spring 2022 were created for the Mississippi River Valley alluvial aquifer (MRVA) using groundwater-altitude data from 1,136 wells completed in the MRVA and from the altitude of the top of the water surface in area rivers from 160 streamgages. The potentiometric-surface and depth-to-water maps for 2022 were created to support investigations to characterize the MRVA as part of the U.S. Geological Survey Water Availability and Use Science Program. Sufficient data were available to map the potentiometric surface and depth to water of the MRVA for spring 2022 for about 83 percent of the aquifer area. The potentiometric contours ranged from 0 to 340 feet (ft) above the North American Vertical Datum of 1988. The regional direction of groundwater gradient was generally to the south-southwest, except in areas of groundwater-altitude depressions, where the groundwater gradient was into the depression, and near rivers, where the groundwater gradient can be from aquifer to the river or from the river into the aquifer. There are large depressions in the potentiometric-surface map in the lower one-half of the Cache region and in much of the Grand Prairie and Delta regions. Depth to water in the MRVA, spring 2022, by well ranged from 5.00 ft above land surface to 145.66 ft below land surface.

Mississippi River Valley alluvial aquifer↗

USGS2024

Use of Doppler velocity radars to monitor and predict debris and flood wave velocities and travel times in post-wildfire basins

The magnitude and timing of extreme events such as debris and floodflows (collectively referred to as floodflows) in post-wildfire basins are difficult to measure and are even more difficult to predict. To address this challenge, a sensor ensemble consisting of noncontact, ground-based (near-field), Doppler velocity (velocity) and pulsed (stage or gage height) radars, rain gages, and a redundant radio communication network was leveraged to monitor flood wave velocities, to validate travel times, and to compliment observations from NEXRAD weather radar. The sensor ensemble (DEbris and Floodflow Early warNing System, DEFENS) was deployed in Waldo Canyon, Pike National Forest, Colorado, USA, which was burned entirely (100 percent burned) by the Waldo Canyon fire during the summer of 2012 ( MTBS, 2020 ). Surface velocity, stage, and precipitation time series collected during the DEFENS deployment on 10 August 2015 were used to monitor and predict flood wave velocities and travel times as a function of stream discharge (discharge; streamflow). The 10 August 2015 event exhibited spatial and temporal variations in rainfall intensity and duration that resulted in a discharge equal to 5.01 cubic meters per second (m 3 /s). Discharge was estimated post-event using a slope-conveyance indirect discharge method and was verified using velocity radars and the probability concept algorithm. Mean flood wave velocities – represented by the kinematic celerity ( c k = 2.619 m e t e r s p e r s e c o n d , m / s ± 0.556 p e r c e n t "> 𝑐 𝑘 = 2.619 𝑚𝑒𝑡𝑒𝑟𝑠 𝑝𝑒𝑟 𝑠𝑒𝑐𝑜𝑛𝑑, m/s ± 0.556 𝑝𝑒𝑟𝑐𝑒𝑛𝑡) and dynamic celerity ( c d = 3.533 m / s ± 0.181 p e r c e n t a n d t h e i r u n c e r t a i n t i e s "> 𝑐 𝑑 = 3.533 m/s ± 0.181 𝑝𝑒𝑟𝑐𝑒𝑛𝑡) and their uncertainties were computed. L-moments were computed to establish probability density functions (PDFs) and associated statistics for each of the at-a-section hydraulic parameters to serve as a workflow for implementing alert networks in hydrologically similar basins that lack data. Measured flood wave velocities and travel times agreed well with predicted values. Absolute percent differences between predicted and measured flood wave velocities ranged from 1.6 percent to 49 percent and varied with water slope, hydraulic radius, and depth. The kinematic celerity was a better predictor for steep slopes and wide flood plains associated with the Upper Waldo and Middle Waldo radar streamgages; whereas, the dynamic celerity was a better surrogate for shallow slopes and incised channels such as the Lower Waldo radar streamgage. The method demonstrates the potential extensibility of a post-wildfire warning system by (1) leveraging multiple systems (i.e., weather radar, near-field velocity and stage radars, and rain gages) for accurate and timely warnings of debris and floodflows, (2) establishing an order of operations to site, install, and operate near-field radars and conventional rain gages to record floodflows, forecast travel times, and document geomorphic change in this basin and hydrologically similar basins that lack data, and (3) communicating data operationally with the Colorado Department of Transportation engineering staff, National Weather Service forecasters, and emergency managers.

Colorado↗

USGS2024

Application of a workflow to determine the feasibility of using simulated streamflow for estimation of streamflow frequency statistics

Streamflow records from hydrologic models are attractive for use in operational hydrology, such as a streamflow frequency analysis. The amount of bias inherent to simulated streamflow from hydrologic models is often unknown, but it is likely present in derivative products. Therefore, a workflow may help determine where streamflow frequency analysis is credibly feasible from simulated streamflow and allow for a systematic way to assess and correct for bias. The proposed workflow consists of hydrologically matching model output locations with streamflow-gauging station (stream gauge) locations, computing the desired statistic from the simulated and observed streamflow record, computing the differences between the simulated and observed statistic (i.e., the bias), and constructing generalized additive models (GAMs) from the differences to determine bias corrections. The US Geological Survey, in cooperation with the Gulf Coast Ecosystem Restoration Council and the US Environmental Protection Agency, is testing the proposed workflow on a low-streamflow frequency (LFF) analysis. Simulated streamflows for the LFF analysis were sourced from a machine-learning model that estimated daily streamflow at Level-12 hydrologic unit code (HUC12) pour points (outlets) in the southern and southeastern US for 1950–2010. The comparison data set consists of 497 stream gauges that are coincident with a HUC12 outlet. The simulated LFF statistics were being overestimated on average; thus, there are limits to using simulated streamflow for frequency analysis. The magnitude of the overprediction generally increases where no-flow conditions are common. Bias corrections determined from the GAMs decreased the magnitude of bias observed in the simulated LFF statistics on average, suggesting it is feasible to expand the operational use of simulated streamflows to frequency analyses with the proposed workflow. The proposed workflow could be advantageous to practitioners interested in leveraging existing and future simulated streamflow data sets with regional and or global coverage.

Journal of Hydrologic Engineering↗

USGS2023

Simulation of monthly mean and monthly base flow of streamflow using random forests for the Mississippi River Alluvial Plain, 1901 to 2018

Improved simulations of streamflow and base flow for selected sites within and adjacent to the Mississippi River Alluvial Plain area are important for modeling groundwater flow because surface-water flows have a substantial effect on groundwater levels. One method for simulating streamflow and base flow, random forest (RF) models, was developed from the data at gaged sites and, in turn, was used to make monthly mean streamflow and base-flow predictions at 162 ungaged sites in the study area. Daily streamflow observations and computed base flow from 247 streamgages were used as the basis for the development of these RF models. RF models were constructed from basin and climatic characteristics and related to observed monthly mean streamflow values; models were used to compute monthly base-flow estimates from selected streamgages in and adjacent to the Mississippi River Alluvial Plain extent, which includes streamflows from parts of Alabama, Arkansas, Colorado, Florida, Illinois, Indiana, Kansas, Kentucky, Louisiana, Mississippi, Missouri, New Mexico, Tennessee, and Texas. The explanatory variables for the models were selected to represent physical characteristics and climatic time series for the contributing drainage basins to the streamgages and ungaged locations of interest. The Nash-Sutcliffe efficiency between observed and simulated monthly mean streamflow was greater than 0.80 for 155 of the 247 streamgages, with a median Nash-Sutcliffe efficiency value of 0.83. The streamflow and base-flow simulations can be used to improve inflow values and to verify the Mississippi River Alluvial Plain groundwater flow model. The statistical model, input data, and response data (simulated monthly mean streamflows) are available as a U.S. Geological Survey software release and a U.S. Geological Survey data release.

Alabama, Arkansas, Illinois, Kentucky, Louisiana, ↗

USGS2023

Generalized additive model estimation of no-flow fractions and L-moments to support flow-duration curve quantile estimation using selected probability distributions for bay and estuary restoration in the Gulf States

Censored and uncensored generalized additive models (GAMs) were developed using streamflow data from 941 U.S. Geological Survey streamflow-gaging stations (streamgages) to predict decadal statistics of daily streamflow for streams draining to the Gulf of Mexico. The modeled decadal statistics comprise no-flow fractions and L-moments of logarithms of nonzero streamflow for six decades (1950–2009). These statistics represent metrics of decadal flow-duration curves (dFDCs) derived from about 10 million daily mean streamflows. The L-moments comprise the mean, coefficient of L-variation, and the third through fifth L-moment ratios. The GAMs were fit to the statistics from 941 streamgages and 2,750 streamgage-decades by using watershed properties such as basin area and slope, decadal precipitation and temperature, and decadal values of flood storage and urban development percentages. The GAMs then estimated decadal statistics for 9,220 prediction locations (stream reaches) coincident with outlets of level-12 hydrologic unit codes. Both entire dataset (whole model) and leave-one-watershed-out model results are reported. No-flow fractions are censored data, and Tobit extensions to GAMs were used to model ephemeral streamflow conditions. Conversely, uncensored GAMs were used for estimation of the L-moments. The GAMs are shown, by coverage probabilities, to construct reliable 95-percent prediction limits. An example shows how no-flow fractions and L-moments may be used to approximate dFDCs by using selected probability distributions (mathematical formulas) including the asymmetric exponential power, generalized normal, and kappa distributions.

Alabama, Arkansas, Florida, Georgia, Kentucky, Lou↗

USGS2023

System-scale airborne electromagnetic surveys in the lower Mississippi River Valley support multidisciplinary applications

The lower Mississippi River Valley spans over 200,000 square kilometres in parts of seven states, encompassing areas of critical groundwater supplies, natural hazards, infrastructure, and low-lying coastal regions. From 2018 - 2022, the U.S. Geological Survey acquired over 82,000 line-kilometres of airborne electromagnetic, radiometric, and magnetic data over this region to provide comprehensive and systematic information about subsurface geologic and hydrologic properties that support multiple scientific and societal interests. Most of the data were acquired on a regional grid of west-east flight lines separated by 3 - 6 kilometres; however, several high-resolution inset grids with line spacing as close as 200 m were acquired in targeted areas of interest. Approximately 8,000 line-kilometres were acquired along streams and rivers to characterise the potential for surface water-groundwater connection, and another 6,000 line-kilometres were acquired along the Mississippi and Arkansas River levees to characterise this critical infrastructure. Here, we present a summary of the data along with several examples of how they are being used to inform regional groundwater model development, inferences of groundwater salinity, identification of faults in the New Madrid seismic zone, and levee infrastructure.

lower Mississippi River Valley↗

USGS2021

Altitude of the potentiometric surface in the Mississippi River Valley alluvial aquifer, spring 2020

The purpose of this report is to present a potentiometric-surface map for the Mississippi River Valley alluvial aquifer (MRVA). The source data for the map were groundwater-altitude data from wells measured manually or continuously generally in spring 2020 and from the altitude of the top of the water surface measured generally on April 9, 2020, in rivers in the area.

Arkansas, Illinois, Kentucky, Louisiana, Mississip↗

USGS2021

Technique to estimate generalized skew coefficients of annual peak streamflow for natural watershed conditions in Texas, Oklahoma, and eastern New Mexico

Reliable information about the frequency of annual peak streamflow is needed for floodplain management, objective assessment of flood risk, and cost-effective design of dams, levees, other flood-control structures, and roads, bridges, and culverts. Generalized skew coefficients are among the data needed for log-Pearson type III peak-streamflow frequency analyses of annual peak streamflows. A technique is presented to estimate generalized skew coefficients used for log-Pearson type III peak-streamflow frequency analyses of annual peak streamflow from natural watersheds (minimal regulation and minimal impervious cover). The estimation of generalized skew coefficients was based on annual and historical peak streamflow data from an initial set of 444 selected USGS streamgaging stations (streamgages) with at least 30 years of recorded annual peak streamflows from natural watersheds in Texas, Oklahoma, and the part of New Mexico east of the Great Continental Divide. The primary focus was to obtain information that could be used to update previously published generalized skew coefficients in Texas. Of the 444 candidate streamgages, 341 were used in the final construction of statistical models. Two generalized additive models (GAMs) were used to predict generalized skew based on a 2-dimensional smooth on projected Albers equal area coordinates of either (1) the locations of the centroids of the gaged watersheds or (2) the streamgage locations. To create maps of generalized skew coefficients, predictions were made on a 1-kilometer grid and contour lines were superimposed. The centroid-location map, with a mean-squared error (MSE) of 0.216, is preferred. Generalized skew coefficients from the centroid-location map, along with the MSE, are useful for computing weighted-skew values when conducting frequency analyses of annual peak streamflow following the guidelines set forth in Bulletin 17C. Based on the results of the study, text revision of the TxDOT Hydraulic Design Manual could be made.

New Mexico, Oklahoma, Texas↗

USGS2020

Regional regression equations for estimation of four hydraulic properties of streams at approximate bankfull conditions for different ecoregions in Texas

The U.S. Geological Survey, in cooperation with the U.S. Army Corps of Engineers, assessed statistical relations between hydraulic properties of streams at approximate bankfull conditions for different ecological regions (ecoregions) in Texas. Data from more than 103,000 records of measured discharge and ancillary hydraulic properties were assembled from summaries of discharge measurements for 424 U.S. Geological Survey streamgages in Texas. The data were subsequently subsetted at each streamgage for a streamgage-specific discharge interval centered on the estimated median annual peak discharge (0.5 annual exceedance probability) obtained from previously published regional regression equations in Texas in conjunction with the streamgage-specific sample median annual peak discharge for the period of record for each streamgage. Discharge measurements at gaged locations representing bankfull conditions (approximated from a discharge interval centered on the estimated median annual peak discharge at a given site) and associated watershed properties were subjected to rigorous statistical analysis. For most discharge measurements (where discharge is symbolically represented as Q ), the following hydraulic properties are available: cross-section area ( A ), water-surface top width ( B ), and reported mean velocity ( V ). Statewide summary statistics were computed by using these four hydraulic properties ( Q , A , B , and V ) and the following five watershed properties: (1) watershed area (contributing drainage area), (2) a multiple of main-channel slope (1,000 times main-channel slope), (3) mean annual precipitation, (4) drainage density, and (5) sinuosity ratio. From the initial set of 424 streamgages, summary statistics were computed for 372 selected streamgages in Texas and constitute the subsetted measurements dataset described in this report. Eight of the 10 ecoregions in Texas are represented in the statewide summary statistics. The resulting statistical relations, expressed as regression equations, can be used to estimate cross-section area, water-surface top width, discharge, and mean velocity of streams in different Texas ecoregions, at approximate bankfull conditions. In the regression equations, watershed properties were the independent variables for applicable watersheds, and predictions from the equations might be useful for estimating the four hydraulic properties at ungaged or unmonitored locations from selected characteristics measured at both the ungaged locations and gaged locations. Four regression equations to estimate the four hydraulic properties were identified as the preferred equations from this study. The four preferred equations use watershed area, mean annual precipitation, and aggregated ecoregion (treated as a categorical variable) to estimate the hydraulic properties, and justification is provided for this preference. For the four equations, the proportions of variance explained by the regression equations as measured by Nash-Sutcliffe efficiency are about 71 percent for cross-section area, 36 percent for top width, 76 percent for discharge, and 25 percent for mean velocity. Residual standard error (RSEs) of the regression equations are 0.252 log10 square feet for cross-section area, 0.319 log10 feet for top width, 0.247 log10 cubic feet per second for discharge, and 0.190 log10 feet per second for mean velocity, and the corresponding standard deviations of response are 0.465 log10 square feet, 0.397 log10 feet, 0.507 log10 cubic feet per second, and 0.220 log10 feet per second, respectively. The residual standard errors are less than the standard deviations as anticipated but show that the uncertainty reduction (percent change) for cross-section area is about −46 percent, about −20 percent for top width, about −51 percent for discharge, and about −14 percent for mean velocity.

Texas↗

USGS2020

Altitude of the potentiometric surface in the Mississippi River Valley alluvial aquifer, spring 2018

A potentiometric-surface map for spring 2018 was created for the Mississippi River Valley alluvial (MRVA) aquifer using available groundwater-altitude data from 1,126 wells completed in the MRVA aquifer and from the altitude of the top of the water surface in area rivers from 66 streamgages. Personnel from Arkansas Natural Resources Commission, Arkansas Department of Health, Arkansas Geological Survey, Illinois Department of Agriculture, Illinois State Water Survey, Louisiana Department of Natural Resources, Louisiana Department of Transportation and Development, Mississippi Department of Environmental Quality, Missouri Department of Natural Resources, Yazoo Mississippi Delta Joint Water Management District, U.S. Department of Agriculture-Natural Resources Conservation Service, and U.S. Geological Survey (USGS) routinely collect groundwater-level data from wells screened in the MRVA aquifer. The USGS and the U.S. Army Corps of Engineers routinely collect data on river stage and streamflow for the rivers overlying the MRVA aquifer area. The potentiometric-surface map for 2018 was created utilizing existing groundwater and surface-water altitudes to support investigations to characterize the MRVA aquifer as part of the USGS Water Availability and Use Science Program. Sufficient data were available to map the potentiometric surface of the MRVA aquifer for spring 2018 for about 87 percent of the aquifer area. The potentiometric contours ranged from 10 to 340 feet above North American Vertical Datum of 1988. The regional direction of groundwater flow was generally to the south-southwest, except in areas of groundwater-altitude depressions, where groundwater flowed into the depression, and near rivers, where flow can be from aquifer to the river or from the river into the aquifer. There are large depressions in the potentiometric-surface map in the lower one-half of the Cache region and in most of the Grand Prairie and Delta regions.

Mississippi River Valley alluvial aquifer↗

USGS2020

Methods to quality assure, plot, summarize, interpolate, and extend groundwater-level information—Examples for the Mississippi River Valley alluvial aquifer

Large-scale computational investigations of groundwater levels are proposed to accelerate science delivery through a workflow spanning database assembly, statistics, and information synthesis and packaging. A water-availability study of the Mississippi River alluvial plain, and particularly the Mississippi River Valley alluvial aquifer (MRVA), is ongoing. Software (visGWDBmrva) has been released as part of the study that demonstrates groundwater informatics for the aquifer. Considerable water-level data collected by multiple agencies over a seven-state area exist (18,903 wells; 287,272 measurements [April 22, 2019]). Data and metadata quality assurance methods, basic statistics, hydrograph visualization, outlier identification, hypothesis testing, and time-series modeling are described. Two approaches (generalized additive models [GAMs] and support vector machines [SVMs]) are used for data interpolation and extension to monthly water-level estimates. Numerical congruence between GAM and SVM estimates will be useful to limit inclusion of monthly estimates from subsequent science activities.

Mississippi River valley↗

USGS2020

The use of support vectors from support vector machines for hydrometeorologic monitoring network analyses

Hydrometeorologic monitoring networks are ubiquitous in contemporary earth-system science. Network stakeholders often inquire about the importance of sites and their locations when discussing funding and monitoring design. Support vector machines (SVMs) can be useful by their assigning each monitoring site as either a support or nonsupport vector. A potentiometric surface was created from synthetic data and 800 random observation locations (sites) as an analog to a groundwater-level network. Using generalized additive models for potentiometric surface prediction, simulations show that a subsample of support vectors from the 800 sites will out perform random samples of sample size equaling the support vector count. Support vector percentages from simulation quantify the recurrence that SVMs assign each site as a support vector, and these percentages in turn measure site importance. An example application of support vector percentages identifies important monitoring sites needed to regionalize the 0.1 annual exceedance probability peak streamflow. The results indicate that 152 of 283 streamgages with support vector percentages equalling 100 percent have not operated since about 2000 and generally have much smaller drainage areas than the greater streamgage network in Texas. The drainage area disparity is an indication of historical imbalance in peak streamflow data acquisition from various stream sizes in Texas.

Texas↗

USGS2019

Copula theory as a generalized framework for flow-duration curve-based streamflow estimates in ungaged and partially gaged catchments

Flow‐duration curve (FDC) based streamflow estimation methods involve estimating an FDC at an ungaged or partially gaged location and using the time series of nonexceedance probabilities estimated from donor streamgage sites to generate estimates of streamflow. We develop a mathematical framework to illustrate the connection between copulas and prior FDC‐based approaches. The performance of copula methods is compared to several other streamflow estimation methods using a decade of daily streamflow data from 74 sites located within two river basins in the southeast United States with different climate characteristics and physiographic properties. We show that copula approaches: (1) outperform other methods in the limiting case of perfect information with regard to the rank‐based correlation structure and FDCs across the gaging network; (2) provide a hedge against poor performance when donor information becomes sparser and less informative; (3) outperform other methods when used for partially gaged sites with several years of available data; and (4) remain a competitive albeit nondominating method for ungaged sites and partially gaged sites with limited data when realistic error is introduced in the estimation of FDCs and correlations across the gaging network.

Water Resources Research↗

USGS2019

Potentiometric surface of the Mississippi River Valley alluvial aquifer, spring 2016

A potentiometric surface map for spring 2016 was created for the Mississippi River Valley alluvial (MRVA) aquifer using selected available groundwater-altitude data from wells and surface-water-altitude data from streamgages. Most of the wells were measured annually or one time after installation, but some wells were measured more than one time or continually; streamgages are typically operated continuously. Personnel from the Arkansas Natural Resources Commission, Arkansas Department of Health, Arkansas Geological Survey, Illinois Department of Agriculture, Illinois State Water Survey, Louisiana Department of Natural Resources, Louisiana Department of Transportation and Development, Mississippi Department of Environmental Quality, Yazoo Mississippi Delta Joint Water Management District, U.S. Department of Agriculture–Natural Resources Conservation Service, and the U.S. Geological Survey (USGS) routinely collect groundwater data from wells screened in the MRVA aquifer. The USGS and the U.S. Army Corps of Engineers routinely collect data on river stage and discharge for the rivers overlying the MRVA aquifer. The potentiometric surface map for 2016 was created using existing data as part of the USGS Water Availability and Use Science Program to support investigations that characterize the MRVA aquifer. Sufficient groundwater-altitude data were available to create a potentiometric-surface map for spring 2016 for about 81 percent of the aquifer area. The potentiometric contours ranged from 10 to 340 feet. The regional direction of groundwater flow in the MRVA aquifer was generally towards the south-southwest, except in areas of groundwater-altitude depressions, where groundwater flows into the depressions, and near rivers, where groundwater flow generally parallels the flow in the rivers. There are large depressions in the potentiometric surface of the MRVA aquifer in the lower half of the Cache region and in most of the Grand Prairie and Delta regions.

Arkansas, Illinois, Kentucky, Louisiana, Mississip↗

USGS2019

Prediction and inference of flow-duration curves using multi-output neural networks

We develop multi-output neural network models (MNNs) to predict flow-duration curves (FDCs) in 9,203 ungaged locations in the Southeastern United States for six decades between 1950-2009. The model architecture contains multiple response variables in the output layer that correspond to individual quantiles along the FDC. During training, predictions are made for each quantile, and a combined loss function is used for back propagation and parameter updating. The loss function accounts for the covariance between the quantiles and generates physically consistent outputs (i.e., monotonically increasing quantiles with increasing nonexceedance probabilities). We use neural-network dropout to generate posterior-predictive distributions for FDCs, and test model performance under cross validation. Finally, we demonstrate how local surrgotate models, via the Local Interpretable Model-agnostic Explanations (LIME) method, can be used to infer the relation between basin characteristics and the predicted FDCs. Results suggest that MNNs can learn the monotonic relations between adjacent quantiles on an FDC, they result in better predictions than single output neural-network models that predict each quantile independently, and basin characteristics are most useful for predicting smaller quantiles, whereas bias terms from neighboring quantiles are most informative for predicting higher quantiles.

Water Resources Research↗