Article pubs.acs.org/est
Predicting Redox Conditions in Groundwater at a Regional Scale Anthony J. Tesoriero,*,† Silvia Terziotti,‡ and Daniel B. Abrams§ †
USGS, 2130 SW Fifth Avenue, Portland, Oregon 97201, United States USGS, 3916 Sunset Ridge Road, Raleigh, North Carolina 27607, United States § Illinois State Water Survey, 2204 Griffith Drive, M/C 674, Champaign, Illinois 61820, United States ‡
Downloaded by UNIV OF NEBRASKA-LINCOLN on September 1, 2015 | http://pubs.acs.org Publication Date (Web): July 31, 2015 | doi: 10.1021/acs.est.5b01869
S Supporting Information *
ABSTRACT: Defining the oxic-suboxic interface is often critical for determining pathways for nitrate transport in groundwater and to streams at the local scale. Defining this interface on a regional scale is complicated by the spatial variability of reaction rates. The probability of oxic groundwater in the Chesapeake Bay watershed was predicted by relating dissolved O2 concentrations in groundwater samples to indicators of residence time and/or electron donor availability using logistic regression. Variables that describe surficial geology, position in the flow system, and soil drainage were important predictors of oxic water. The probability of encountering oxic groundwater at a 30 m depth and the depth to the bottom of the oxic layer were predicted for the Chesapeake Bay watershed. The influence of depth to the bottom of the oxic layer on stream nitrate concentrations and time lags (i.e., time period between land application of nitrogen and its effect on streams) are illustrated using model simulations for hypothetical basins. Regional maps of the probability of oxic groundwater should prove useful as indicators of groundwater susceptibility and stream susceptibility to contaminant sources derived from groundwater.
■
INTRODUCTION
Concentrations of products and reactants in key redox reactions have been used for redox zonation and to assess contaminant transport in both groundwater and streams.5,10−12 A comparison of the percentages of wells in networks across the United States that are in each redox category (e.g., oxic, nitratereducing, etc.) provided one of the first assessments of how redox conditions vary at a national scale.13 Although redox conditions have been mapped at the local scale using electrodes14 and at the small catchment scale (101 km2) using modeling,15 the spatial variability of redox reaction rates has made the determination of redox conditions at regional scales (e.g., >100 000 km2) problematic. As a result, studies that have estimated the depth at which redox boundaries are likely to occur at the regional scale are rare. To assess redox conditions at a regional scale, it is necessary to characterize the spatial variability of redox conditions. To accomplish this, dissolved oxygen data from samples collected from thousands of wells in the Chesapeake Bay watershed were related to geographic information on soils, geology, and other explanatory variables. Specifically, the probability of oxic groundwater was predicted on a regional scale by relating O2 concentrations in groundwater to variables describing landscape characteristics that are likely to affect redox conditions (e.g., soil drainage characteristics, organic carbon content in soils) and
The fate of many contaminants in groundwater depends to a large extent on the redox processes that occur along flow paths.1 Redox processes influence groundwater transport and potential toxicity either by directly transforming contaminants to other species (e.g., chromium reduction;2 denitrification3) or by causing the precipitation or dissolution of compounds that contain or sorb contaminants (e.g., phosphate sorption on iron oxides4). Redox reactions in groundwater have been shown to markedly affect the transport of nutrients,5 volatile organic compounds,6 and trace metals.7 The effect that redox reactions have on the transport of many contaminants underscores the importance of defining redox reaction zones in groundwater (e.g., zones of dominant terminal electron acceptors). Microbial metabolism depends on the oxidation of organic or inorganic (e.g., FeS2) species to generate energy for growth and maintenance, with the metabolic reaction that yields the most energy typically dominating over competing reactions.8 Microbes will first oxidize organic carbon (or reduced minerals) using O2 as an electron acceptor through aerobic respiration because it is the most energetically favorable reaction. When O2 is depleted, facultative anaerobes begin to use nitrate (NO3−) as an electron acceptor during denitrification. Subsequent reactions include the reduction of Mn(IV), Fe(III), and sulfate and methanogenesis. Terminal electron-accepting processes may often be deduced by examining the concentrations of the reactants and products of these reactions.9 © 2015 American Chemical Society
Received: Revised: Accepted: Published: 9657
April 16, 2015 June 26, 2015 July 14, 2015 July 31, 2015 DOI: 10.1021/acs.est.5b01869 Environ. Sci. Technol. 2015, 49, 9657−9664
Article
Environmental Science & Technology sample position in the flow system. The objectives of this work were to (1) develop a method to evaluate redox conditions on a regional scale, (2) apply this method to the Chesapeake Bay watershed, and (3) demonstrate the effect that depth to the bottom of the oxic layer has on the transport of nitrate to streams.
odds of an event occurring (oxic groundwater in this case). Odds ratios are computed by dividing the odds of a sample being classified as an event with a unit change in the variable being examined by the odds of an event occurring without a unit change in this variable. Standardized coefficients22 were calculated to compare the rank order of the magnitude of the influences of the predictor variables on the presence or absence of oxic conditions. Logistic regression analyses were performed using Statistical Analysis Software23 and R, a free software environment for statistical computing and graphics.24 Several metrics were used to evaluate LR model predictions of oxic water for both the training and independent data sets: percent of samples that were correctly classified (overall accuracy rate), the area under the receiver operating characteristic curve (AUC), and the Hosmer−Lemeshow statistic. These metrics are suitable for evaluating the accuracy of predicted classes and the probabilities of an event.16,25 The percent of samples that were correctly classified was determined by classifying predictions as oxic or suboxic on the basis of a threshold probability of 50% (i.e., when predicted probabilities were ≥50%, samples were classified as oxic, suboxic when predicted probabilities were
