Search for: All records

Abstract. In groundwater pumping optimization (GPO), offline-trained data-driven surrogates can be used to replace numerical-intensive simulators in order to save computing time. The traditional offline training approach involves building surrogates prior to optimization, fitting training datasets that cover the input space uniformly or randomly, which can prove inefficient due to the potential oversampling of low-gradient areas and under-sampling of high-gradient areas. This study proposes an offline machine-learning (ML) algorithm that ranks candidate training points by scoring them based on their distance to the closest training point and on the local gradient of the surrogate estimate and then choosing the highest-rank point. This method is applied to develop surrogates for solving a two-objective GPO problem formulated on a three-dimensional (3D) island aquifer, using hydrogeological conditions representative of San Salvador Island, Bahamas. The objectives are to minimise the supply cost (fOC) resulting from groundwater pumping and desalination and maximise fresh groundwater supply (Qp), subject to constraints on seawater intrusion (SWI) control expressed in terms of aquifer drawdown Δs at pumping locations and aquifer salt mass increase ΔSM. Gaussian Process (GP) is the technique applied to construct surrogates of objectives and constraints, alongside the estimation of uncertainties. Using GP models, it is possible to estimate the probability of “Pareto optimality” for each pumping scheme by Monte Carlo simulation. Pareto optimal pumping schemes (POPS) are then characterized by a probability of occurrence, which can be verified by numerical simulation. The GP training strategy's effectiveness in generating POPS is compared to traditional training approaches, showing that such a strategy can efficiently identify reliable POPS. 

Reduced river flows and groundwater depletion as a result of climate change and population growth have increased the effort and difficulty accessing and processing water. In turn, residential water costs from municipal utilities are predicted to rise to unaffordable rates for poor residential water customers. Building on a regional conjunctive use model with future climate scenarios and 50-year future water supply plans, our study communicates the effects of climate change on poor people in El Paso, Texas, as water becomes more difficult and expensive to obtain in future years. Four scenarios for future water supply and future water costs were delineated based on expected impacts of climate change and groundwater depletion. Residential water use was calculated by census tract in El Paso, using basic needs indoor water use and evaporative cooling use as determinants of household water consumption. Based on household size and income data from the US Census, fraction of household income spent on water was determined. Results reveal that in the future, basic water supply will be a significant burden for 40% of all households in El Paso. Impacts are geographically concentrated in poor census tracts. Our study revealed that negative impacts from water resource depletion and increasing populations in El Paso will lead to costly and difficult water for El Paso water users. We provide an example of how to connect future resource scenarios, including those affected by climate change, to challenges of affordability for vulnerable consumers. 

Abstract The middle Rio Grande is a vital source of water for irrigation in the region. Climate change is impacting regional hydrology and is likely to put additional stress on a water supply that is already stretched thin. To gain insight on the hydrologic effects of climate change on reservoir storage, a simple water balance model was used to simulate the Elephant Butte–Caballo Reservoir system (southern New Mexico). The water balance model was forced by hydrologic inputs generated by 97 climate simulations derived from CMIP5 global climate models, coupled to a surface hydrologic model. Results suggest that the percentage of years that reservoir releases satisfy agricultural water rights allocations over the next 50 years (2021–70) will decrease relative to the past 50 years (1971–2020). The modeling also projects an increase in multiyear drought events that hinder reservoir management strategies to maintain high storage levels. In most cases, changes in reservoir inflows from distant upstream snowmelt is projected to have a greater influence on reservoir storage and water availability downstream of the reservoirs than will changes in local evaporation and precipitation from the reservoir surfaces. 

Abstract Centralized water infrastructure has, over the last century, brought safe and reliable drinking water to much of the world. But climate change, combined with aging and underfunded infrastructure, is increasingly testing the limits of—and reversing gains made by—this approach. To address these growing strains and gaps, we must assess and advance alternatives to centralized water provision and sanitation. The water literature is rife with examples of systems that are neither centralized nor networked, yet meet water needs of local communities in important ways, including: informal and hybrid water systems, decentralized water provision, community‐based water management, small drinking water systems, point‐of‐use treatment, small‐scale water vendors, and packaged water. Our work builds on these literatures by proposing a convergence approach that can integrate and explore the benefits and challenges of modular, adaptive, and decentralized (“MAD”) water provision and sanitation, often foregrounding important advances in engineering technology. We further provide frameworks to evaluate justice, economic feasibility, governance, human health, and environmental sustainability as key parameters of MAD water system performance. This article is categorized under:Engineering Water > Water, Health, and SanitationHuman Water > Water GovernanceEngineering Water > Sustainable Engineering of Water