An integrated Markov chain Monte Carlo algorithm for upscaling hydrological and geochemical parameters from column to field scale.

Predicting and controlling the concentrations of redox-sensitive elements are primary concerns for environmental remediation of contaminated sites. These predictions are complicated by dynamic flow processes as hydrologic variability is a governing control on conservative and reactive chemical concentrations. Subsurface heterogeneity in the form of layers and lenses further complicates the flow dynamics of the system impacting chemical concentrations including redox-sensitive elements. In response to these complexities, this study investigates the role of heterogeneity and hydrologic processes in an effective parameter upscaling scheme from the column to the landfill scale. We used a Markov chain Monte Carlo (MCMC) algorithm to derive upscaling coefficients for hydrological and geochemical parameters, which were tested for variations across heterogeneous systems (layers and lenses) and interaction of flow processes based on the output uncertainty of dominant biogeochemical concentrations at the Norman Landfill site, a closed municipal landfill with prevalent organic and trace metal contamination. The results from MCMC analysis indicated that geochemical upscaling coefficients based on effective concentration ratios incorporating local heterogeneity across layered and lensed systems produced better estimates of redox-sensitive biogeochemistry at the field scale. MCMC analysis also suggested that inclusion of hydrological parameters in the upscaling scheme reduced the output uncertainty of effective mean geochemical concentrations by orders of magnitude at the Norman Landfill site. This was further confirmed by posterior density plots of the scaling coefficients that revealed unimodal characteristics when only geochemical processes were involved, but produced multimodal distributions when hydrological parameters were included. The multimodality again suggests the effect of heterogeneity and lithologic variability on the distribution of redox-sensitive elements at the Norman Landfill site.

Prenatal nitrate intake from drinking water and selected birth defects in offspring of participants in the national birth defects prevention study.

BACKGROUND: Previous studies of prenatal exposure to drinking-water nitrate and birth defects in offspring have not accounted for water consumption patterns or potential interaction with nitrosatable drugs. OBJECTIVES: We examined the relation between prenatal exposure to drinking-water nitrate and selected birth defects, accounting for maternal water consumption patterns and nitrosatable drug exposure. METHODS: With data from the National Birth Defects Prevention Study, we linked addresses of 3,300 case mothers and 1,121 control mothers from the Iowa and Texas sites to public water supplies and respective nitrate measurements. We assigned nitrate levels for bottled water from collection of representative samples and standard laboratory testing. Daily nitrate consumption was estimated from self-reported water consumption at home and work. RESULTS: With the lowest tertile of nitrate intake around conception as the referent group, mothers of babies with spina bifida were 2.0 times more likely (95% CI: 1.3, 3.2) to ingest ≥ 5 mg nitrate daily from drinking water (vs. < 0.91 mg) than control mothers. During 1 month preconception through the first trimester, mothers of limb deficiency, cleft palate, and cleft lip cases were, respectively, 1.8 (95% CI: 1.1, 3.1), 1.9 (95% CI: 1.2, 3.1), and 1.8 (95% CI: 1.1, 3.1) times more likely than control mothers to ingest ≥ 5.42 mg of nitrate daily (vs. < 1.0 mg). Higher water nitrate intake did not increase associations between prenatal nitrosatable drug use and birth defects. CONCLUSIONS: Higher water nitrate intake was associated with several birth defects in offspring, but did not strengthen associations between nitrosatable drugs and birth defects.

Estimating Escherichia coli loads in streams based on various physical, chemical, and biological factors.

Microbes have been identified as a major contaminant of water resources. Escherichia coli (E. coli) is a commonly used indicator organism. It is well recognized that the fate of E. coli in surface water systems is governed by multiple physical, chemical, and biological factors. The aim of this work is to provide insight into the physical, chemical, and biological factors along with their interactions that are critical in the estimation of E. coli loads in surface streams. There are various models to predict E. coli loads in streams, but they tend to be system or site specific or overly complex without enhancing our understanding of these factors. Hence, based on available data, a Bayesian Neural Network (BNN) is presented for estimating E. coli loads based on physical, chemical, and biological factors in streams. The BNN has the dual advantage of overcoming the absence of quality data (with regards to consistency in data) and determination of mechanistic model parameters by employing a probabilistic framework. This study evaluates whether the BNN model can be an effective alternative tool to mechanistic models for E. coli loads estimation in streams. For this purpose, a comparison with a traditional model (LOADEST, USGS) is conducted. The models are compared for estimated E. coli loads based on available water quality data in Plum Creek, Texas. All the model efficiency measures suggest that overall E. coli loads estimations by the BNN model are better than the E. coli loads estimations by the LOADEST model on all the three occasions (three-fold cross validation). Thirteen factors were used for estimating E. coli loads with the exhaustive feature selection technique, which indicated that six of thirteen factors are important for estimating E. coli loads. Physical factors included temperature and dissolved oxygen; chemical factors include phosphate and ammonia; biological factors include suspended solids and chlorophyll. The results highlight that the LOADEST model estimates E. coli loads better in the smaller ranges, whereas the BNN model estimates E. coli loads better in the higher ranges. Hence, the BNN model can be used to design targeted monitoring programs and implement regulatory standards through TMDL programs.

Enhanced biogeochemical cycling and subsequent reduction of hydraulic conductivity associated with soil-layer interfaces in the vadose zone.

Biogeochemical dynamics in the vadose zone are poorly understood due to the transient nature of chemical and hydrologic conditions but are nonetheless critical to understanding chemical fate and transport. This study explored the effects of a soil layer on linked geochemical, hydrological, and microbiological processes. Three laboratory soil columns were constructed: a homogenized medium-grained sand, a homogenized organic-rich loam, and a sand-over-loam layered column. Upward and downward infiltration of water was evaluated during experiments to simulate rising water table and rainfall events, respectively. In situ collocated probes measured soil water content, matric potential, and Eh. Water samples collected from the same locations were analyzed for Br, Cl, NO, SO, NH, Fe, and total sulfide. Compared with homogeneous columns, the presence of a soil layer altered the biogeochemistry and water flow of the system considerably. Enhanced biogeochemical cycling was observed in the layered column over the texturally homogeneous soil columns. Enumerations of iron- and sulfate-reducing bacteria showed 1 to 2 orders of magnitude greater community numbers in the layered column. Mineral and soil aggregate composites were most abundant near the soil-layer interface, the presence of which likely contributed to an observed order-of-magnitude decrease in hydraulic conductivity. These findings show that quantifying coupled hydrologic-biogeochemical processes occurring at small-scale soil interfaces is critical to accurately describing and predicting chemical changes at the larger system scale. These findings also provide justification for considering soil layering in contaminant fate and transport models because of its potential to increase biodegradation or to slow the rate of transport of contaminants.

Inverse estimation of parameters for multidomain flow models in soil columns with different macropore densities.

Soil and crop management practices have been found to modify soil structure and alter macropore densities. An ability to accurately determine soil hydraulic parameters and their variation with changes in macropore density is crucial for assessing potential contamination from agricultural chemicals. This study investigates the consequences of using consistent matrix and macropore parameters in simulating preferential flow and bromide transport in soil columns with different macropore densities (no macropore, single macropore, and multiple macropores). As used herein, the term"macropore density" is intended to refer to the number of macropores per unit area. A comparison between continuum-scale models including single-porosity model (SPM), mobile-immobile model (MIM), and dual-permeability model (DPM) that employed these parameters is also conducted. Domain-specific parameters are obtained from inverse modeling of homogeneous (no macropore) and central macropore columns in a deterministic framework and are validated using forward modeling of both low-density (3 macropores) and high-density (19 macropores) multiple-macropore columns. Results indicate that these inversely modeled parameters are successful in describing preferential flow but not tracer transport in both multiple-macropore columns. We believe that lateral exchange between matrix and macropore domains needs better accounting to efficiently simulate preferential transport in the case of dense, closely spaced macropores. Increasing model complexity from SPM to MIM to DPM also improved predictions of preferential flow in the multiple-macropore columns but not in the single-macropore column. This suggests that the use of a more complex model with resolved domain-specific parameters is recommended with an increase in macropore density to generate forecasts with higher accuracy.

Polycyclic aromatic hydrocarbons in soil of the Canadian river floodplain in Oklahoma.

The accumulation of polycyclic aromatic hydrocarbons (PAH) in soil, plants, and water may impart negative effects on ecosystem and human health. We quantified the concentration and distribution of 41 PAH (n = 32), organic C, total N, and S (n = 140) and investigated PAH sources using a chronosequence of floodplain soils under a natural vegetation succession. Soil samples were collected between 0- and 260-cm depth in bare land (the control), wetland, forest, and grassland areas near a closed municipal landfill and an active asphalt plant (the contaminant sources) in the north bank of the Canadian River near Norman, OK. Principal component, cluster, and correlation analyses were used to investigate the spatial distribution of PAH, in combination with diagnostic ratios to distinguish pyrogenic vs. petrogenic PAH suites. Total PAH concentration (SigmaPAH) had a mean of 1300 ng g(-1), minimum of 16 ng g(-1), and maximum of 12,000 ng g(-1). At 0- to 20-cm depth, SigmaPAH was 3500 +/- 1600 ng g(-1) (mean +/- 1 SE) near the contaminant sources. The most common compounds were nonalkylated, high molecular weight PAH of pyrogenic origin, i.e., fluoranthene (17%), pyrene (14%), phenanthrene (9%), benzo(b)fluoranthene (7%), chrysene (6%), and benzo(a)anthracene (5%). SigmaPAH in the control (130 +/- 23 ng g(-1)) was comparable to reported concentrations for the rural Great Plains. Perylene had a unique distribution pattern suggesting biological inputs. The main PAH contamination mechanisms were likely atmospheric deposition due to asphalt production at the 0- to 20-cm depth and past landfill operations at deeper depths.