Description: ABSTRACT Using epidemiologic time-series analysis, we examine associations between three hydroclimatic variables (temperature, precipitation, and streamflow) and waterborne acute gastro-intestinal illness (AGI) in two communities in the province of British Columbia (BC), Canada. The communities were selected to represent the major hydroclimatic regimes that characterize BC: rainfall-dominated and snowmelt-dominated. Our results show that the number of monthly cases of AGI increased with increasing temperature, precipitation, and streamflow in the same month in the context of a rainfall-dominated regime, and with increasing streamflow in the previous month in the context of a snowfall-dominated regime. These results suggest that hydroclimatology plays a role in driving the occurrence and variability of AGI in these settings. Further, this study highlights that the nature and magnitude of the effects of hydroclimatic variability on AGI are different in the context of a snowfall-dominated regime versus a rainfall-dominated regimes. We conclude by proposing that the watershed may be an appropriate context for enhancing our understanding of the complex linkages between hydroclimatic variability and waterborne illness in the context of a changing climate. This article is protected by copyright. All rights reserved.

Description: ABSTRACT The spatial heterogeneity of hydraulic conductivity (K) exerts a major control on groundwater flow and solute transport. The heterogeneous spatial distribution of K can be imaged using indirect geophysical data as long as reliable relations exist to link geophysical data to K. This paper presents a non-parametric learning machine approach to predict aquifer K from cone penetrometer tests (CPT) coupled with a soil moisture and resistivity probe (SMR) using relevance vector machines (RVMs). The learning machine approach is demonstrated with an application to a heterogeneous unconsolidated littoral aquifer in a 12-km 2 sub-watershed, where relations between K and multi-parameters CPT/SMR soundings appear complex. Our approach involved fuzzy clustering to define hydrofacies (HF) on the basis of CPT/SMR and K data prior to the training of RVMs for HFs recognition and K prediction on the basis of CPT/SMR data alone. The learning machine was built from a colocated training dataset representative of the study area that includes K data from slug tests and CPT/SMR data up-scaled at a common vertical resolution of 15 cm with K data. After training, the predictive capabilities of the learning machine were assessed through cross-validation with data withheld from the training dataset and with K data from flowmeter tests not used during the training process. Results show that HF and K predictions from the learning machine are consistent with hydraulic tests. The combined use of CPT/SMR data and RVM-based learning machine proved to be powerful and efficient for the characterization of high-resolution K heterogeneity for unconsolidated aquifers. This article is protected by copyright. All rights reserved.

Description: An empirical framework for assisting with water quality management is proposed that relies on open-source hydrologic data. Such data are measured periodically at fixed water stations and commonly available in time-series form. To fully exploit the data, we suggest that observations from multiple stations should be combined into a single long-panel data set, and an econometric model developed to estimate upstream management effects on downstream water quality. Selection of the model's functional form and explanatory variables would be informed by rating curves, and idiosyncrasies across and within stations handled in an error term by testing contemporary correlation, serial correlation, and heteroskedasticity. Our proposed approach is illustrated with an application to the Nakdong River Basin in South Korea. Three alternative policies to achieve downstream BOD level targets are evaluated: upstream water treatment, greater dam discharge, and development of a new water source. Upstream water treatment directly cuts off incoming pollutants, thereby presenting the smallest variation in its downstream effects on BOD levels. Treatment is advantageous when reliability of water quality is a primary concern. Dam discharge is a flexible tool, and may be used strategically during a low-flow season. We consider development of a new water corridor from an extant dam as our third policy option. This turns out to be the most cost-effective way for securing lower BOD levels in the downstream target city. Even though we consider a relatively simple watershed to illustrate the usefulness of our approach, it can be adapted easily to analyze more complex upstream-downstream issues. This article is protected by copyright. All rights reserved.

Description: Dense Non-Aqueous Phase Liquids (DNAPLs) induce variation in electromagnetic characteristics of the ground e.g. electric permittivity and resistivity. The most used indirect methods in the mapping of these physical characteristics are electrical resistivity and ground penetrating radar. To better understand the effect of DNAPL release on electrical permittivity and resistivity in a water saturated medium, we carried out a controlled laboratory experiment where the host material was simulated by glass beads and the DNAPL by HFE-7100 (hydrofluoroether). The experiment measured the electric resistivity and permittivity of each fluid, the multi-phase fluid system, and the host material, along with time-lapse electrical resistivity and GPR measurements in a controlled cell. We found that the different phases of DNAPL within a saturated medium (free, dissolved and gaseous phase) affect the physical characteristics differently. The reflection pull-up behind contaminated sediments, which is normally detected by GPR, was mainly inferred from the HFE free phase. The dissolved phase causes small variations in electric permittivity not usually readily detected by GPR measurements. Both the dissolved and free HFE phases induce variation in resistivity. This article is protected by copyright. All rights reserved.

Description: Characterizing the topology of three-dimensional steady-state flow fields is useful to describe the physical processes controlling the deformation of solute plumes and, consequently, obtain helpful information on mixing processes without solving the transport equation. In this work, we study the topology of flow in three-dimensional non-stationary anisotropic heterogeneous porous media. In particular, we apply a topological metric, i.e., the helicity density, and two complementary kinematic descriptors of mixing, i.e., stretching and folding, to investigate: i) the flow field resulting from applying a uniform-in-the-average hydraulic gradient within a fully-resolved heterogeneous three-dimensional porous medium with a non-stationary anisotropic covariance function of the locally isotropic hydraulic log-conductivity; ii) the flow field obtained by averaging a set of Monte Carlo realizations of the former field; iii) the flow field obtained considering the blockwise uniform anisotropic effective conductivity tensor computed for the fully-resolved case. While in the fully-resolved case the local helicity density is zero as a consequence of the local isotropy of hydraulic conductivity, it differs from zero in the other two cases. We show, therefore, that this topological metric is scale dependent and should be computed at the appropriate scale to be informative about the leading patterns of plume deformation. Indeed, streamlines are helical in all three cases at scales larger than the characteristic scale of spatial variability. We apply stretching and folding metrics to investigate the scales at which plume deformation is more influenced by helical motion than by the effect of small scale spatial heterogeneity in the hydraulic conductivity field. Under steady-state flow conditions, stretching, which quantifies the increasing length of an interface, dominates at short distances from a given starting plane, while folding, which describes how this interface is bent to fill a finite volume of space, dominates further downstream and can be correlated with the appearance of large scale secondary motion. We conclude that three-dimensional flows in porous media may show a complex topology whose analysis is relevant for the description of plume deformation. These results have important implications for the understanding of mixing processes, as shown in detail in the companion paper focusing on solute transport [ Cirpka et al. , submitted]. This article is protected by copyright. All rights reserved.

Description: ABSTRACT A reactive transport modeling framework is presented that allows simultaneous assessment of groundwater flow, water quality evolution including δ 13 C, and 14 C activity or “age”. Through application of this framework, simulated 14 C activities can be directly compared with measured 14 C activities. This bypasses the need for interpretation of a 14 C age prior to flow simulation through factoring out processes other than radioactive decay, which typically involves simplifying assumptions regarding spatial and temporal variability in reactions, flow, and mixing. The utility of the approach is demonstrated for an aquifer system with spatially variable carbonate mineral distribution, multiple organic carbon sources, and transient boundary conditions for 14 C activity in the recharge water. In this case the simulated 14 C age was shown to be relatively insensitive to isotopic fractionation during DOC oxidation and variations in assumed DOC degradation behaviour. We demonstrate that the model allows quantitative testing of hypotheses regarding controls on groundwater age and water quality evolution for all three carbon isotopes. The approach also facilitates incorporation of multiple environmental tracers and combination with parameter optimization techniques. This article is protected by copyright. All rights reserved.

Description: Water scarcity is likely to increase in the coming years, making improvements in irrigation efficiency increasingly important. An emerging technology that promises to increase irrigation efficiency substantially is a wireless irrigation sensor network that uploads sensor data into irrigation management software, creating an integrated system that allows real-time monitoring and control of moisture status that has been shown in experimental settings to reduce irrigation costs, lower plant loss rates, shorten production times, decrease pesticide application, and increase yield, quality, and profit. We use an original survey to investigate likely initial acceptance, ceiling adoption rates, and profitability of this new sensor network technology in the nursery and greenhouse industry. We find that adoption rates for a base system and demand for expansion components are decreasing in price, as expected. The price elasticity of the probability of adoption suggests that sensor networks are likely to diffuse at a rate somewhat greater than that of drip irrigation. Adoption rates for a base system and demand for expansion components are increasing in specialization in ornamental production: Growers earning greater shares of revenue from greenhouse and nursery operations are willing to pay more for a base system and are willing to purchase larger numbers of expansion components at any given price. We estimate that growers who are willing to purchase a sensor network expect investment in this technology to generate significant profit, consistent with findings from experimental studies. This article is protected by copyright. All rights reserved.

Description: Flow and transport through fractured geologic media often leads to anomalous (non-Fickian) transport behavior, the origin of which remains a matter of debate: whether it arises from variability in fracture permeability (velocity distribution), connectedness in the flow paths through fractures (velocity correlation), or interaction between fractures and matrix. Here we show that this uncertainty of distribution- vs. correlation-controlled transport can be resolved by combining convergent and push-pull tracer tests because flow reversibility is strongly dependent on velocity correlation, whereas late-time scaling of breakthrough curves is mainly controlled by velocity distribution. We build on this insight, and propose a Lagrangian statistical model that takes the form of a continuous time random walk (CTRW) with correlated particle velocities. In this framework, velocity distribution and velocity correlation are quantified by a Markov process of particle transition times that is characterized by a distribution function and a transition probability. Our transport model accurately captures the anomalous behavior in the breakthrough curves for both push-pull and convergent flow geometries, with the same set of parameters. Thus, the proposed correlated CTRW modeling approach provides a simple yet powerful framework for characterizing the impact of velocity distribution and correlation on transport in fractured media. This article is protected by copyright. All rights reserved.

Description: Average water travel times through a stream network were determined as a function of stage (discharge) and stream network properties. Contrary to most previous studies on the topic, the present work allowed for stream flow velocities to spatially (for most of the analyses) as well as temporally. The results show that different stream network mechanisms and properties interact in a complex and stage-dependent manner, implying that the relative importance of the different hydraulic properties varies in space and over time. Theoretical reasoning, based on the central temporal moments derived from the kinematic-diffusive wave equation in a semi-2D formulation including the effects of flooded cross-sections, shows that the hydraulic properties in contrast to the geomorphological properties will become increasingly important as the discharge increases, stressing the importance of accurately describing the hydraulic mechanisms within stream networks. ‘ Using the physically based, stage-dependent response function as a parameterization basis for the stream flow routing routine (a linear reservoir) of a hydrological model, discharge predictions were shown to improve in two Swedish catchments, compared to when using a conventional, statistically based parameterization scheme. Predictions improved for a wide range of modeled scenarios, for the entire discharge series as well as for peakflow conditions. The foremost novelty of the study lies in that the physically based response function for a streamflow routing routine has successfully been determined independent of calibration i.e . entirely through process based hydraulic stream network modelling. This article is protected by copyright. All rights reserved.

Description: One–dimensional analytical heat transport equations based on temperature time series data have become popular tools to quantify groundwater–surface water interactions. The influence of non–ideal field conditions on the use of these equations has been assessed for non–sinusoidal stream temperature signals, uncertainty in thermal parameters, sensor accuracy and multidimensional flow. Given that streambeds are often highly heterogeneous, the influence of streambed heterogeneity on flux estimates from temperature time series requires further investigation. Synthetic streambed temperatures were generated using two–dimensional numerical models with heterogeneous hydraulic conductivity distributions. Streambed temperatures were used to calculate fluxes using methods based on amplitude ratios ( A r ), phase shifts (Δ Φ ) and both ( A r Δ Φ ). Calculated fluxes were compared to known fluxes from the numerical models for flow fields analogous to losing streams. The influence of streambed structure, degree of heterogeneity, depth of the sensor pair, and location along a flow path were assessed. Errors in calculated fluxes increased with sensor pair depth, position along a flow path, and with the degree of heterogeneity. These errors were larger for streambeds with isotropic structures compared with anisotropic structures, and of the three methods tested; the Δ Φ method produced the largest errors. The simultaneous estimation of strong fluxes using Δ Φ , and an inability to obtain a flux estimate from A r can suggest the presence of low hydraulic conductivity zones. Given the large errors and inability to determine flow direction from the Δ Φ method, the A r and A r Δ Φ methods are recommended for downwelling fluxes. This article is protected by copyright. All rights reserved.