Journal of Water Resources Planning and Management

Forum papers are thought-provoking opinion pieces or essays founded in fact, sometimes containing speculation, on a civil engineering topic of general interest and relevance to the readership of the journal. The views expressed in this Forum article do not necessarily reflect the views of ASCE or the Editorial Board of the journal.

Aligned with the goals of achieving higher sustainability in water systems, the minimization of energy costs is a top priority for the operators and owners of such systems. Early research contributions on the rate of energy dissipation due to head losses, defined as energy slope $J$, demonstrated that this parameter can be optimized so that such cost minimization is achieved. However, an important factor that was not considered in previous research is that water demand might be variable during the lifetime of water systems. Whether due to daily variation patterns, seasonal changes, or long-term water demand growth, variations on the flow rate complicates the determination of the optimized energy slope $J_{\mathrm{opt}}$ that will minimize energy costs. This work addresses this knowledge gap, performing an evaluation of how demand variation impacts $J_{\mathrm{opt}}$ and how deviations from the optimum energy slope affect costs for varying water demands. In addition, this work also provided insights into pipe cost installation terms and pipe cost exponents using real-world data of existing water systems in Spain. Such exponents are very important in the computation of optimum energy slopes. It is hoped that this research will provide practical guidance to designers and operators of water systems regarding the minimization of energy costs, leading to greater sustainability of such systems.

Stochastic optimization is an important tool employed to manage salt intrusion and increased freshwater production in coastal aquifers by estimating the optimum well locations and well operation parameters. In karst aquifers, the shape and location of the caves in the aquifers are often uncertain parameters. Thus, it becomes necessary to take into consideration the uncertainties when optimizing water production from such aquifers. The uncertainty associated with the parameterization of aquifers is often handled by creating several equiprobable realizations of aquifers through stochastic simulations. These realizations jointly describe the uncertainty in the aquifer model and as such are used as a means to manage uncertainty when performing optimization and simulation studies of such aquifers. However, owing to the large number of stochastic realizations often created to describe the uncertainty in an aquifer model, performing optimization under uncertainty becomes computationally expensive. In this paper, we propose a freshwater production optimization strategy that uses two separate clustering strategies to identify a small set of realizations (from the total ensemble of aquifer model realizations) upon which the optimization study can be conducted. In this study, a clustering strategy is adopted to reduce the computational expense associated with conducting the optimization study. The $k\text{-}\mathrm{means}++$ algorithm was used as the clustering algorithm, and a modified form of the Darcy model with optimized permeability distribution (DMOPD) was selected as the forward model that describes the flow of fluid in the aquifer. Furthermore, the DMOPD was connected to an advection-dispersion-adsorption equation that describes the transport of salt with the fluid phase. A synthetic aquifer example was used to illustrate the optimization strategy and the results obtained show that the clustering algorithm proves to be a useful tool in selecting representative samples for the optimization case study. Also, the optimization algorithm was found to be a viable tool to limit saltwater intrusion in karstic aquifers while maximizing freshwater recovery.

Climate change impacts on irrigated agriculture reveal the necessity of improving water and energy use efficiency and integrating renewable energies in this sector. Pressurized networks are typical in the irrigation sector. In these networks, nodes with excess pressure are frequent and the installation of pressure-reducing devices is required for their proper operation. As these devices imply an energy dissipation previously used, this work focuses on determination of the optimal number and location of pumps as turbines (PATs) in irrigation networks to maximize the energy recovery, provided that the feasibility of installation of each PAT separately is guaranteed considering the large flow fluctuations typical in these networks. The proposed methodology considered theoretical PAT curves and was evaluated in two irrigation networks located in Southern Spain. The results showed three and four potential sites for PAT installation in each network, with total annual energy recovery estimations of 43.1 and 91.8 MWh, a payback period lower than 10 years in most of them, and potential annual ${\mathrm{CO}}_2$ emissions savings up to 13.4 and $28.5{\mathrm{tCO}}_2$. The developed model could be an interesting tool to offer irrigation district managers or technicians a baseline hydropower potential scenario in irrigation networks.

The pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in 2020 led to a significant change in human behaviors, mainly because of the quarantine to avoid the spread of the virus. Measures affected both economic activities and citizens’ behaviors as they developed more intense hygiene habits to avoid contamination and switched to home offices. These exceptional behaviors also affected the way that water is consumed and need to be fully understood to manage supply systems. Therefore, this study aims to investigate changes in residential and commercial water consumption in 31 municipalities in the state of São Paulo, Brazil, during SARS-CoV-2. To do this, the expected consumption for the first half of 2020 was forecasted using the Holt-Winters multiplicative method and compared with the data observed for the same period. In addition, we compared monthly records of new contaminations and the social distancing index to establish a correlation with changes in water consumption. The results show an average difference between forecasted and observed consumption equal to $+6.23\%$ and $-18.59\%$ for residential and commercial activities, respectively. For the first one, the consumption per capita increased at the rate of ${8.44\mathrm{L}.\mathrm{person}}^{-1}{.\mathrm{day}}^{-1}$. The observed changes in consumption seem to be a consequence of hygiene habits, social distancing and the closing of nonessential services in commerce.

Waterway restoration in urban areas can be an important environmental policy topic. However, there have been few studies of monetary benefits to help justify expenditures and inform priority setting for remediation works. We investigated values for improving river health in two catchments in Sydney, Australia: the Cooks Catchment, which is smaller and urban-based, and the Georges Catchment, which is larger and located in both urban and periurban areas. Using choice experiments, we found that households are willing to pay Australian Dollars (AUD) 10.34 per year for 5 years in the Cooks Catchment, and AUD 2.64 in the Georges Catchment, to restore 1 km of urban waterway. We found that willingness to pay is moderated by the time until project outcomes are achieved. Comparison with typical costs of urban stream remediation suggested that there is a broad range of projects that will produce positive net benefits in the Cooks Catchment, whereas in the Georges Catchment the economically viable projects are those with lower cost and that can achieve their outcomes relatively quickly.

Accurate estimates of reference evapotranspiration (${\mathit{ET}}_0$) are critical for estimating actual crop evapotranspiration and agricultural water use. This study uses observations from the Nevada Integrated Climate and Evapotranspiration Network (NICE Net) to validate forecasts of ${\mathit{ET}}_0$ and its driving variables from the National Weather Service’s National Digital Forecast Database (NDFD). Daily NDFD ${\mathit{ET}}_0$ at lead times of 1 to 6 days were compared against 18 NICE Net stations. Correlations between NDFD and observations generally ranged between 0.4 and 0.9, with lower correlations at longer leads and a notable drop in skill during July and August. Systematic arid biases (high bias for temperatures and low bias for humidity) were found in NDFD with a strong warm minimum temperature bias and low vapor pressure bias most prominent during the growing season. Some of the largest relative biases were found in wind speed, although they were systematic and varied greatly by location. A case study revealed that NDFD consistently underestimates the variability found in observed minimum temperature, solar radiation, wind speed, and ${\mathit{ET}}_0$. Cloudy days during summer were not well represented in the NDFD estimated solar radiation, which had a cascading impact on temperature, vapor pressure, and ${\mathit{ET}}_0$ estimates. A monthly ratio-based bias-correction was applied to NDFD ${\mathit{ET}}_0$, which reduced the root-mean squared error by 5%–30% for most locations. Bias-corrected ${\mathit{ET}}_0$ forecasts from NDFD or other forecast systems show potential as a guide to develop weekly irrigation schedules for agricultural producers, with the ultimate goal of reducing applications of excess irrigation water.

Water productivity promotion is critical for achieving a balance maintaining the livelihood of large poor agricultural population and protecting ecological water in developing countries in arid and semiarid regions, and government financial payments play a significant role in this process. According to different water rights, government financial payment can be categorized as permanent water right payment and temporary water right payment, whereas the impacts of the two policies and their interrelationship are still unclear. This study aimed to explore the impacts of the two water right payments on the balance between agricultural and ecological interests in developing countries in arid and semiarid areas. Taking the Hami Prefecture in east Xinjiang, China, as the case study area, an agent-based model was developed to simulate farmers’ decisions and agricultural and ecological results under different policy conditions. Results showed that the temporary water right payment is useful in promoting water right trade but hard to drive technology adoption individually or protect agriculture economic profit. The single permanent water right payment is better under the crop production preference and ecology preference, whereas the proper combination of the two financial payment policies could reinforce each other and cost less fiscal expenditure if the government prefers a balance between agricultural production and ecology interests.

Effective detection and classification of abnormalities, such as pipe bursts, leakage, illegal water use, and sensor failures, are critical for assisting water utilities in decision making, rapid response, and minimizing damage and disruption. This work presents a new flow data-based anomaly detection and classification method in water distribution networks. The method first establishes hourly nodal water demand forecasting models, then uses a unique integration of feature extraction technique of flow curve and convolutional neural network method to enable anomaly detection and classification from continually updated time window flow data. Verification progress from real and synthetic data of the case network shows that the proposed method can identify four common types of abnormal patterns in a fast and reliable manner with high recognition accuracy. The established models have self-learning capabilities, can process flow data in real time, and do not require hydraulic models to assist in analysis, which can be promising for wide practical applications in smart management of water distribution systems.

Nowadays, drinking water demand forecasting has become fundamental to efficiently manage water distribution systems. With the growth of accessible data and the increase of available computational power, the scientific community has been tackling the forecasting problem, opting often for a data-driven approach with considerable results. However, the most performing methodologies, like deep learning, rely on the quantity and quality of the available data. In real life, the demand data are usually affected by the missing data problem. This study proposes an analysis of the role of missing data imputation in the frame of a short-term forecasting process. A set of conventional imputation algorithms were considered and applied on three test cases. Afterward, the forecasting process was performed using three state-of-the-art deep neural network models. The results showed that a good quality imputation can significantly affect the forecasting results. In particular, the results highlighted significant variation in the accuracy of the forecasting models that had past observation as inputs. On the contrary, a forecasting model that used only static variables as input was not affected by the imputation process and may be a good choice whenever a good quality imputation is not possible.

Environmental constraints in hydropower systems serve to ensure sustainable use of water resources. Through accurate treatment in hydropower scheduling, one seeks to respect such constraints in the planning phase while optimizing the utilization of hydropower. However, many environmental constraints introduce state-dependencies and even nonconvexities to the scheduling problem, making them challenging to represent in stochastic hydropower scheduling models. This paper describes how the state-dependent maximum discharge constraint, which is widely enforced in the Norwegian hydropower system, can be embedded within the stochastic dual dynamic programming (SDDP) algorithm for hydropower scheduling without compromising computational time. In this work, a combination of constraint relaxation and time-dependent auxiliary lower reservoir volume bounds is applied, and the modeling is verified through computational experiments on two different systems. The results demonstrate that the addition of an auxiliary lower bound on reservoir volume has significant potential for improved system operation, and that a bound based on the minimum accumulated inflow in the constraint period is the most efficient.

Abstract

Analysis of water distribution networks (WDNs) is considered to be indispensable for water supply design. Over the last decades, several methods have been developed to formulate and solve the steady-state hydraulic analysis of WDNs. For large WDNs, modeling the networks is complex and takes a great deal of time, and using the traditional methods becomes practically inefficient. This paper presents a parallel paradigm based on cellular automata to accelerate the hydraulic analysis of WDNs. The structure of the computing system is generated according to the structure of the WDN using the concept of cellular automata. Two methods are proposed for solving the hydraulic models. The performance of the proposed approach was compared with that of EPANET 2.0 software for networks with different complexity and topologies. The results indicate that, despite requiring more iterations to converge, the proposed parallel algorithm can accelerate the hydraulic analysis as much as 105 times using the proposed methods.

Practical Applications

WDNs analysis under steady-state conditions is used widely in applications such as network design and leak detection in WDNs. In these applications, it is necessary to solve hydraulic equations of the WDN under different scenarios to obtain the desired answer. As the size of the WDN increases, the time required to solve the equations increases accordingly. This paper presents a parallel paradigm to accelerate the hydraulic analysis of WDNs. The results indicate that the proposed parallel algorithm can accelerate the hydraulic analysis, thus reducing the network design time and determining the location of leaks in the WDN with greater speed and accuracy. Consequently, more money, time, and water can be saved. In addition, this acceleration may allow engineers to find new ways to solve their problems, or even to provide solutions to problems that in the past could not be solved with existing facilities.

Visual detection and classification of water and waterbodies provide important information needed for managing water resources systems and infrastructure, such as developing flood early warning systems and drought management. But water itself is a challenging object for visual analysis because it is shapeless, colorless, and transparent. Therefore, detecting, tracking, and localizing water in different visual environments are difficult tasks. Computer vision (CV) techniques provide powerful tools for image processing and high-level scene analysis. Despite the complexities associated with water in visual scenes, there are still some physical differences, such as color, turbidity, and turbulence, affected by surrounding settings, which can potentially support CV modeling to cope with the visual processing challenges of water. The goal of this study is to introduce a new image data set, ATLANTIS Texture (ATeX), which represents various water textures of different waterbodies, and evaluate the performance of deep learning (DL) models for classification purposes on ATeX. Experimental results show that among DL models, EffNet-B7, EffNet-B0, GoogLeNet, and ShuffleNet $\mathrm{V}2\times 1.0$ provide the highest precision, recall, and F1 score. However, by considering the training time, total number of parameters, and total memory occupied by these models, ShuffleNet $\mathrm{V}2\times 1.0$ is presented as the most efficient DL network for water classification. Finally, results from this study suggest that ATeX provides a new benchmark to investigate existing challenges in the field of image analysis, in particular for water, which can help both water resources engineers and the computer vision community.

Climate change exacerbates environmental challenges caused by urban growth. This study offers a multifaceted approach to understanding socioenvironmental vulnerability in cities by considering fine-scale spatial distribution of social and physical metrics overlaid with runoff and surface temperature (ST) data. Stormwater runoff depth and surface temperature data for micro-subbasins in Philadelphia were estimated and integrated to create an environmental vulnerability index. Social variables representing poverty, the elderly population, population density, and buildings with basements were used to develop a social vulnerability index. Simple analysis based on medians of combined environmental and social vulnerability indexes identified specific subbasins of the city that are most vulnerable to heat and flooding impacts of climate change. Hotspot analysis of combined social and environmental vulnerability data identified regions of the city that would benefit most from a focus on implementation of new green-blue infrastructures. Results indicated that neighborhoods in southern Philadelphia are most vulnerable to flooding and heat. The outlined approach can be used for any city, and provides city managers with a city-scale visualization for prioritizing areas in immediate need of climate mitigation practices.

To examine the characteristics of future precipitation under climate change is of great significance to urban water security. In this paper, multiple machine learning techniques, i.e., statistical downscaling model (SDSM), support vector machine (SVM), and multilayer perceptron (MLP), were used to downscale large-scale climatic variables simulated by the General Circulation Models (GCMs) to precipitation on a local scale. It was demonstrated in Shenzhen city, China, through multisite downscaling schemes based on projections from the Max Planck Institute Earth System Model (MPI-ESM1.2-HR), Meteorological Research Institute Earth System Model Version 2.0 (MRI-ESM2.0), and Beijing Climate Center Climate System Model (BCC-CSM2-MR). The obtained results showed that the downscaled precipitation would provide good monthly simulations against observations at 10 discrete stations. Regardless of superior performance of SVM and MLP over SDSM, the daily precipitation simulations should be further improved, and downscaling of heavy daily precipitations would be promoted by quantile mapping corrections. Due to the relatively poor simulation performance of BCC-CSM2-MR, the other two climate models were considered under the Shared Socioeconomic Pathways (SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios) for ensemble precipitation projections for 2015–2100. Under the SSP1-2.6 scenario, the amounts of annual average precipitation for 10 stations were estimated to be higher relative to the historical period (2.7%–17%), and 9 out of 10 stations presented an increasing trend. However, downward trends also existed at three stations when it comes to scenarios SSP2-4.5 and SSP5-8.5. Moreover, a significantly positive trend was found to dominate the trend changes of annual extreme daily precipitation during 2015–2050, but the detected trends at stations were greatly dependent on the downscaling techniques and climate models. Besides, the increase in daily extreme precipitations for various return periods as well as statistically different precipitation characteristics for discrete stations would further shed light on urgent demands on urban resilient strategies for climate change adaptation.