Search

Seasonal streamflow forecasts are becoming widely used to improve water reservoir operations, especially in areas where climate teleconnections enable predictability on medium and long lead times. Most existing studies have focused on the assimilation of forecasts into operational decision models, an approach that typically banks on predeveloped forecasts to optimize water release decisions. However, this approach may overlook the potential synergies that stand in co-developing forecast and decision-making models. In other words, the opportunities that lie in coupling both forecast and operational decision models have not yet been explored. Here, we address this gap and contribute a novel approach building on the Evolutionary Multi-Objective Direct Policy Search algorithm to design forecast and decision-making models together. The proposed approach is benchmarked against operating policies not informed by any forecast, as well as by forecast-informed policies relying on predeveloped forecasts (data-driven and perfect). Numerical experiments are conducted on the Angat-Umiray water resources system, Philippines, which is operated primarily for ensuring municipal water supply to Metro Manila and irrigation supply to a large agricultural district. Our results show that the integrated design of forecast models and control policies provides a performance gain with respect to policies informed by predesigned forecasts. This result is particularly interesting because the skill of the integrated forecast models is lower than that of the predeveloped ones, thus suggesting that more accurate forecasts do not necessarily produce better water system operations. Overall, our analysis represents a step towards a deeper integration of streamflow forecast and reservoir operation models. [ABSTRACT FROM AUTHOR]

Highlights • We extract typical behaviors of water reservoir operators from observational data. • We cluster the reservoir operators based on their behavioral profile. • We discover behavioral profiles that are vulnerable to drought conditions. Abstract The large number of dammed rivers worldwide emphasizes the need to couple models of natural processes with models describing human behaviors. However, such behavioral models are often simplistic and lack proper validation against observational data. In this work, we contribute a new approach to infer the typical operations of water reservoirs from historical observations, using data-driven behavioral modeling based on eigenbehavior analysis. The approach is demonstrated using monthly storage data from 172 reservoirs in California, USA. Results show that the proposed method identifies four typical behavioral profiles, which are strongly linked to key features of the reservoirs. Moreover, we show how the identified models can be used for discovering behavioral profiles, and associated reservoir characteristics, that are vulnerable to drought conditions. [ABSTRACT FROM AUTHOR]

Ongoing changes in global climate are expected to alter the hydrologic regime of many river basins worldwide, expanding historically observed variability as well as increasing the frequency and intensity of extreme events. Understanding the vulnerabilities of water systems under such uncertain and variable hydrologic conditions is key to supporting strategic planning and design adaptation options. In this paper, we contribute a multiobjective assessment of the impacts of hydrologic uncertainty on the operations of multipurpose water reservoirs systems in arid climates. We focus our analysis on the Dez and Karoun river system in Iran, which is responsible for the production of more than 20% of the total hydropower generation of the country. A system of dams controls most of the water flowing to the lower part of the basin, where irrigation and domestic supply are strategic objectives, along with flood protection. We first design the optimal operations of the system using observed inflows and evaporation rates. Then, we simulate the resulting solutions over different ensembles of stochastic hydrology to partition the impacts of streamflow and evaporation uncertainty. Numerical results show that system operations are extremely sensitive to alterations of both uncertainty sources. In particular, we show that in this arid river basin, long-term objectives are mainly vulnerable to inflow uncertainty, whereas evaporation rate uncertainty mostly affects short-term objectives. Our results suggest that local water authorities should properly characterize hydrologic uncertainty in the design of future operations of the expanded network of reservoirs, possibly also investing in the improvement of the existing monitoring network to obtain more reliable data for modeling streamflow and evaporation processes. [ABSTRACT FROM AUTHOR]

Water resources systems face irreducible uncertainty in supply and demand, requiring policies to respond to changing conditions on multiple timescales. For both short-term operation and long-term adaptation, thresholds or “decision triggers”, where a policy links observed indicators to actions, have featured prominently in recent studies. There remains a need for a general method to conceptualize threshold-based policies in an easily interpretable structure, and a corresponding search algorithm to design them. Here we propose a conceptual and computational framework where policies are formulated as binary trees, using a simulation-optimization approach. Folsom Reservoir, California serves as an illustrative case study, where policies define the thresholds triggering flood control and conservation actions. Candidate operating rules are generated across an ensemble of climate scenarios, incorporating indicator variables describing longer-term climate shifts to investigate opportunities for adaptation. Policy tree optimization and corresponding open-source software provide a generalizable, interpretable approach to policy design under uncertainty. [ABSTRACT FROM AUTHOR]

Robust decision-making is being increasingly used to support environmental resources decisions and policy analysis under changing climate and society. In this context, a robust decision is a decision that is as much as possible insensitive to a large degree of uncertainty and ensures certain performance across multiple plausible futures. Yet, the concept of robustness is neither unique nor static. Multiple robustness metrics, such as maximin, optimism-pessimism, max regret, have been proposed in the literature, reflecting diverse optimistic/pessimistic attitudes by the decision maker. Further, these attitudes can evolve in time as a response to sequences of favorable (or adverse) events, inducing possible dynamic changes in the robustness metrics. In this paper, we explore the impact of alternative definitions of robustness and their evolution in time for a case of water resources system management under changing climate. We study the decisions of the Lake Como operator, who is called to regulate the lake by balancing irrigation supply and flood control, under an ensemble of climate change scenarios. Results show a considerable variability in the system performance across multiple robustness metrics. In fact, the mis-definition of the actual decision maker's attitude biases the simulation of its future decisions and produces a general underestimation of the system performance. The analysis of the dynamic evolution of the decision maker's preferences further confirms the potentially strong impact of changing robustness definition on the decision-making outcomes. Climate change impact assessment studies should therefore include the definition of robustness among the uncertain parameters of the problem in order to analyze future human decisions under uncertainty. [ABSTRACT FROM AUTHOR]

Optimal management policies for water reservoir operation are generally designed via stochastic dynamic programming (SDP). Yet, the adoption of SDP in complex real-world problems is challenged by the three curses of dimensionality, modeling, and multiple objectives. These three curses considerably limit SDP's practical application. Alternatively, this study focuses on the use of evolutionary multiobjective direct policy search (EMODPS), a simulation-based optimization approach that combines direct policy search, nonlinear approximating networks, and multiobjective evolutionary algorithms to design Pareto-approximate closed-loop operating policies for multipurpose water reservoirs. This analysis explores the technical and practical implications of using EMODPS through a careful diagnostic assessment of the effectiveness and reliability of the overall EMODPS solution design as well as of the resulting Pareto-approximate operating policies. The EMODPS approach is evaluated using the multipurpose Hoa Binh water reservoir in Vietnam, where water operators are seeking to balance the conflicting objectives of maximizing hydropower production and minimizing flood risks. A key choice in the EMODPS approach is the selection of alternative formulations for flexibly representing reservoir operating policies. This study distinguishes between the relative performance of two widely-used nonlinear approximating networks, namely artificial neural networks (ANNs) and radial basis functions (RBFs). The results show that RBF solutions are more effective than ANN ones in designing Pareto approximate policies for the Hoa Binh reservoir. Given the approximate nature of EMODPS, the diagnostic benchmarking uses SDP to evaluate the overall quality of the attained Pareto-approximate results. Although the Hoa Binh test case's relative simplicity should maximize the potential value of SDP, the results demonstrate that EMODPS successfully dominates the solutions derived via SDP. [ABSTRACT FROM AUTHOR]

trends and prospects. The course Irc [teaching of catholic religion] Didactics aims to give a professional standing to the future teacher of catholic religion [Idr]. This means it has to help teachers to achieve the required professional skills thanks to: educational, disciplinary, planning, didactic and organizational trainings. The proposal of the course entails a discipline training enabling to use planning categories and proceedings pertaining to the educational system. This begins with the planning of didactic itineraries and ends up with their control and evaluation. At present in the Irc proposal, the comparison with other religions and ways of life is fundamental. Idr formation takes into consideration lab experiences and school apprenticeship which enable to outline educational strategies in schoolrooms. [ABSTRACT FROM AUTHOR]

Worldwide, reservoirs are used to buffer against both extreme floods and extreme droughts. However, operations favoring each of these and other management objectives conflict. Fortunately, tradeoffs in operations can often be mitigated by using hydrologic forecasts to condition release decisions, and significant research has investigated the value of using forecasts for this purpose. However, these studies have struggled with how to quantify forecast value in multi-objective contexts where the benefits may accrue unevenly to different objectives that stakeholders value differently. To address this research gap, we introduce a new metric for quantifying forecast value on multi-objective problems: the average improvement on each objective when moving from a solution that does not use forecast information (a baseline solution) to solutions that do use forecast information and outperform the baseline solution on all objectives. We call this metric "average domination" and use it to investigate whether and how the value of forecast information for multi-objective reservoir operations changes with different flood protection constraints, using the Red River Basin in Vietnam as an example. To assess this, we design multi-objective operations at four reservoirs in the basin both with and without forecast information conditioning release decisions under constraints that the operations ensure protection to downstream water levels with either 100-year or 500-year return levels. Based on the average domination metric, we find that in the Red River Basin, the value of forecast information for all objectives is the same or greater under the more severe 500-year flood design requirement. These findings, identified by our new metric, illustrate that it can be especially beneficial to condition operations on forecast information when stakeholders strongly favor higher levels of risk aversion for flood protection. • We develop a new metric for valuing forecast information on multi-objective problems. • We use this metric to assess forecast value for reservoir operations. • We find forecasts hold more value when constrained to protect to more severe floods. [ABSTRACT FROM AUTHOR]

In state-of-the-art energy systems modelling, reservoir hydropower is represented as any other thermal power plant: energy production is constrained by the plant's installed capacity and a capacity factor calibrated on the energy produced in previous years. Natural water resource variability across different temporal scales and the subsequent filtering effect of water storage mass balances are not accounted for, leading to biased optimal power dispatch strategies. In this work, we aim at introducing a novelty in the field by advancing the representation of reservoir hydropower generation in energy systems modelling by explicitly including the most relevant hydrological constraints, such as time-dependent water availability, hydraulic head, evaporation losses, and cascade releases. This advanced characterization is implemented in an open-source energy modelling framework. The improved model is then demonstrated on the Zambezi River Basin in the South Africa Power Pool. The basin has an estimated hydropower potential of 20,000 megawatts (MW) of which about 5,000 MW has been already developed. Results show a better alignment of electricity production with observed data, with a reduction of estimated hydropower production up to 35% with respect to the baseline Calliope implementation. These improvements are useful to support hydropower management and planning capacity expansion in countries richly endowed with water resource or that are already strongly relying on hydropower for electricity production. [ABSTRACT FROM AUTHOR]

Population growth and rising economic prosperity combined with alterations of hydroclimatic conditions are expected to pose major challenges for water management policies, particularly in fast-developing economies. Growing energy and food demands will likely trigger major investments in large dam projects, whose performance strongly depends on unknown future conditions in the coming decades. In contrast to traditional impact assessments that focus on climate and socio-economic divers separately, in this work we propose an integrated framework to better understand the nested interdependencies of those drivers across sectors and processes at different scales. We integrate an ensemble of climate projections generated from Representative Concentration Pathways with a large factorial dataset of socio-economic scenarios derived from the Socio-Economic Pathways evaluated using the Global Change Assessment Model (GCAM). As a demonstration, we study the Zambezi River, one of the largest river basins in Africa and a paradigmatic of most basins in developing countries. The four largest dams are mainly operated for hydropower production and irrigation supply, with considerable negative effects on the aquatic ecosystems due to the alteration of the natural flow regime.To assess the robustness of the water-energy-food nexus in the Zambezi River basin against future integrated scenarios of changing socio-economic and climate drivers, we first downscaled climate projections to feed local hydrological models and produce projections of water availability in the basin. In parallel, we spatially and temporally downscaled GCAM outputs via the Tethys software to obtain projections of irrigation demands. Numerical results show that the current management of the system is highly vulnerable to changing conditions, with critical risks of failure in terms of food security due to unreliable irrigation supply. Our framework demonstrates that these critical conditions are mostly generated by global socio-economic drivers, namely the alternative regimes of land-use change taxation, rather than predicted changes in water availability due to climate change. [ABSTRACT FROM AUTHOR]

Between 1980 and 2015, Europe experienced 18% of worldwide weather-related loss events, which accounted for over US$500 billion in damage. Early action triggered by timely information on possible emerging impact has the potential to reduce these losses, playing a substantial role on social and economic development. The importance of the early action systems has been recognized by the seventh global target of the Sendai Framework for disaster risk reduction that urges for the development of impact-informed early warning systems.Despite great advancements in flood forecasting technologies, it remains challenging to produce useful impact-based forecast information in a way to trigger early actions that mitigate risks. Overcoming the challenges to act based on the forecast requires an understanding of the existing forecast tools, and a deeper comprehension of the dynamics between hydrometeorological hazards and impacts. Previous studies have shown spatial and temporal relationships between large-scale climate oscillation and flood risk, flood occurrence and flood losses in Europe. However, little is understood about the skill of these oscillations in forecasting flood losses. Using a database of flood losses (NatCatSERVICE of Munich Re), we examined the skill of large-scale climate oscillation - the El Niño Southern Oscillation (ENSO), the North Atlantic Oscillation (NAO), the East Atlantic/West Russian pattern (EA/WR) and the East Atlantic Pattern (EA) and the Scandinavian Pattern (SCA) - in forecasting seasonal and sub-regional Flood Losses in Europe. Using multiple linear regression models to analyse relationships between the indices of climate variability and Flood Losses, we found that at the Pan-European level and sub-regional scales, flood losses are linked to the interaction with all five indices of climate oscillation. The largest effect is found in spring between SCA and NAO, in which large-scale Pan-European flood losses are expected to be 26 times more costly (approximately 2500 million euro) than average when the SCA positive interacts with the negative phase of the NAO. At sub-regional scale, the largest effect is found in winter in Eastern Europe between EAWR and NAO, in which large-scale flood losses of approximately 3600 million euro (292 times higher than average) are expected when the SCA positive interacts with the negative phase of the NAO. Since large-scale indices of climate variability can be predicted with longer lead times than weather variables, sub-regional impact-based forecast could be developed for regions where a significant relationship is found between large-scale indices of climate variability and flood losses. [ABSTRACT FROM AUTHOR]

With more than 3500 hydropower dams currently being planned or built around the world, the need for optimal dam planning is more compelling than ever. Current literature efforts are directed towards strategic sizing and siting of dams to maximize economic return and minimize environmental impacts in regime conditions. However, also the initial transient phase of reservoir filling may have substantial environmental and economic implications and, especially in large projects, it can be responsible for critical sectoral conflicts associated to dam development. In this phase, precaution towards downstream impacts requires transiting high percentages of inflow, resulting in multi-years, even decadal, filling transients. Conversely, upstream interests (e.g., hydropower production) favor fast impoundment of water, albeit generating critical periods of minimal streamflow downstream. In this process, hydroclimatic variability plays a key role: if the filling occurs during a dry spell, augmented impacts must be endured by upstream and downstream sectors. The state of global climate teleconnections represents a valuable source of predictability of local hydrometeorological anomalies, which can provide a dual support to dam planning. First, in scheduling a favorable time to start the filling operations based on local effects of long-term climatic oscillations, and second, in designing effective filling strategies adaptive to medium-term hydroclimatic variability. In this work, we consider the case study of the Omo-Turkana basin (Ethiopia-Kenya) to demonstrate the use of seasonal drought forecasts based on multiple climate teleconnections for modulating reservoir filling operations according to the predicted drought conditions, e.g. by transiting higher fractions of the inflow when a drought is forecasted. This case study is in this respect exemplary, because the impacts due to the construction and filling of Gibe III dam and to the current construction of Koysha dam, raised the strong opposition of local communities and international organizations. We run a retrospective analysis of Gibe III filling and compare the impacts of the historical filling strategy to designed alternative strategies, demonstrating the potential to significantly decrease downstream impacts, for instance by successfully maintaining the level of downstream Lake Turkana up to one meter higher than historically observed. Our approach also shows that 2015 was an unfavorable year to begin the filling operations of Gibe III, and significant damage could be avoided by delaying or anticipating the operations by 2-3 years. Finally, we predict that the planned time to start filling the Koysha dam, 2021, occurs at the beginning of a multi-year drought and we issue recommendations on more favorable timings. [ABSTRACT FROM AUTHOR]

Atmospheric extreme events cause severe damage to human societies and ecosystems. The frequency and intensity of extremes and other associated events are continuously increasing due to climate change and global warming. The accurate prediction, characterization, and attribution of atmospheric extreme events is, therefore, a key research field in which many groups are currently working by applying different methodologies and computational tools. Machine learning and deep learning methods have arisen in the last years as powerful techniques to tackle many of the problems related to atmospheric extreme events. This paper reviews machine learning and deep learning approaches applied to the analysis, characterization, prediction, and attribution of the most important atmospheric extremes. A summary of the most used machine learning and deep learning techniques in this area, and a comprehensive critical review of literature related to ML in EEs, are provided. The critical literature review has been extended to extreme events related to rainfall and floods, heatwaves and extreme temperatures, droughts, severe weather events and fog, and low-visibility episodes. A case study focused on the analysis of extreme atmospheric temperature prediction with ML and DL techniques is also presented in the paper. Conclusions, perspectives, and outlooks on the field are finally drawn. [ABSTRACT FROM AUTHOR]

Many water resources systems include multiple, independent, and distributed decision makers representing different and conflicting interests. In much of the water resources literature, the operation of these systems is studied assuming a fully cooperative attitude by the parties involved and maximizing the global efficiency at the system-level. However, assuming the presence of a social planner might be questionable when multiple institutions are involved, particularly in transboundary systems. At the other extreme, totally uncoordinated strategies among institutionally independent decision makers, acting according to the principle of individual-rationality, are more often experienced in these contexts, yielding a decrease in the system-level performance. In this paper, a novel approach is proposed based on multiagent systems to support the design of regulatory mechanisms, which drive the originally fully independent decision makers towards a more coordinated and system-wide efficient situation. The agent-based model is coupled with tools and algorithms based on distributed constraint reasoning to represent the interactions between the decision makers. The approach is demonstrated on a hypothetical water allocation problem, involving several active human agents and passive ecological agents. Different regulatory mechanisms are explored in three different scenarios of water availability to quantitatively support the discussion about the efficiency-acceptability trade-off. Numerical results show that the proposed approach has a great potential to support the design of distributed solutions balancing system-level efficiency and individual-level acceptability. [ABSTRACT FROM AUTHOR]

In river basin systems, power asymmetry is often responsible of inefficient and unbalanced water allocations. Climate change and anthropogenic pressure will possibly exacerbate such disparities as the dominant party controls an increasingly limited shared resource. In this context, the deployment of cooperation mechanisms giving greater consideration to a balanced distribution of the benefits, while improving system-wide efficiency, may be desirable. This often implies the intervention of a third party (e.g., the river basin water authority) imposing normative constraints (e.g., a minimum release) on the party in the dominant position. However, this imposition will be more acceptable to the dominant party if coupled with some form of compensation. For a public agency, compensation may be burdensome, especially when the allowance is triggered by natural events whose timing and magnitude are highly uncertain. In this context, index-based insurance contracts may represent a viable alternative and reduce the cost of achieving socially desirable outcomes. In this paper, we develop a hybrid cooperation mechanism composed of i) a direct normative constraint imposed by a regulator, and ii) an indirect financial tool, an index-based insurance contract, to be used as a compensation measure. The approach is developed for the Lake Como multi-purpose water system, Italy: a complex Alpine river basin, supporting several hydropower reservoirs and finally flowing into a regulated lake which supplies water to several downstream uses, mostly irrigated agriculture. The system is characterized by a manifest geographic power asymmetry: the upstream hydropower companies are free to release their stored water in time irrespective of the timing of the downstream demands. This situation can lead to financial losses by the downstream users and undesirable social outcomes. Results suggest that financial instruments may offer a reliable and relatively inexpensive alternative to other forms of compensation, and thereby favor more balanced management of multi-purpose water systems characterized by power asymmetry. This finding is especially relevant in times when granting of licenses to use/withdrawal water are often being reviewed with attention to environmental protection and equity issues. [ABSTRACT FROM AUTHOR]

Water reservoir systems may become more adaptive and reliable to external changes by enlarging the information sets used in their operations. Models and forecasts of future hydro-climatic and socio-economic conditions are traditionally used for this purpose. Nevertheless, the identification of skillful forecasts and models might be highly critical when the system comprises several processes with inconsistent dynamics (fast and slow) and disparate levels of predictability. In these contexts, the direct use of observational data, describing the current conditions of the water system, may represent a practicable and zero-cost alternative. This paper contrasts the relative contribution of state observations and perfect forecasts of future water availability in improving multipurpose water reservoirs operation over short- and long-term temporal scales. The approach is demonstrated on the snow-dominated Lake Como system, operated for flood control and water supply. The Information Selection Assessment (ISA) framework is adopted to retrieve the most relevant information to be used for conditioning the operations. By explicitly distinguishing between observational dataset and future forecasts, we quantify the relative contribution of current water system state estimates and perfect streamflow forecasts in improving the lake regulation with respect to both flood control and water supply. Results show that using the available observational data capturing slow dynamic processes, particularly the snow melting process, produces a 10% improvement in the system performance. This latter represents the lower bound of the potential improvement, which may increase to the upper limit of 40% in case skillful (perfect) long-term streamflow forecasts are used. [ABSTRACT FROM AUTHOR]

Model Predictive Control (MPC) has recently gained increasing interest in the adaptive management of water resources systems due to its capability of incorporating disturbance forecasts into real-time optimal control problems. Yet, related literature is scattered with heterogeneous applications, case-specific problem settings, and results that are hardly generalized and transferable across systems. Here, we systematically review 149 peer-reviewed journal articles published over the last 25 years on MPC applied to water reservoirs, open channels, and urban water networks to identify common trends and open challenges in research and practice. The three water systems we consider are inter-connected, multi-purpose and multi-scale dynamical systems affected by multiple hydro-climatic uncertainties and evolving socioeconomic factors. Our review first identifies four main challenges currently limiting most MPC applications in the water domain: (i) lack of systematic benchmarking of MPC with respect to other control methods; (ii) lack of assessment of the impact of uncertainties on the model-based control; (iii) limited analysis of the impact of diverse forecast types, resolutions, and prediction horizons; (iv) under-consideration of the multi-objective nature of most water resources systems. We then argue that future MPC applications in water resources systems should focus on addressing these four challenges as key priorities for future developments. [ABSTRACT FROM AUTHOR]

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. [ABSTRACT FROM AUTHOR]

The integration of quality and quantity issues in the management of water resources systems is key to meet society's long-term needs for freshwater while maintaining essential ecological services and economic benefits. Current water management practices are mostly targeted towards quantitative uses, and quality is usually addressed separately as an independent problem. One of the reasons for the lack of integration lies in the inadequacy of optimization techniques nowadays available to cope with the large, distributed, simulation models adopted to characterize the coupled ecological and biochemical processes in water bodies. In this paper we propose a novel approach based on the conjunctive use of a batch-mode Reinforcement Learning algorithm and a one-dimensional (1D) coupled hydrodynamic-ecological model to design the optimal operation of a multipurpose water reservoir accounting for both quantity and quality targets. We consider up to five operating objectives, including both in-reservoir and downstream water quality parameters, and design efficient operating policies conditioned upon not only the current storage but also water characteristics, such as temperature and total suspended solids at different depths. The approach is applied to a real world case study in Japan consisting of a water reservoir, Tono Dam, equipped with a selective withdrawal structure and used for flood protection, irrigation and recreational purposes. Results show that a potential control over in-reservoir and downstream water quality can be gained without impairing the hydraulic capacity of the reservoir by effectively exploiting-through the operating policy-the operational flexibility provided by the selective withdrawal structures. [ABSTRACT FROM AUTHOR]

Deep uncertainty in future climate, socio-economic and technological conditions poses a great challenge to medium-long term decision making. Recently, several approaches have been proposed to identify solutions that are robust with respect to a large ensemble of deeply uncertain future scenarios. In this paper, we introduce ROSS (Robust Optimal Scenario Selection), a novel algorithm that uses an active learning approach for adaptively selecting the smallest scenario subset to be included into a robust optimization process. ROSS contributes a twofold novelty in the field of robust optimization under deep uncertainty. First, it allows the computational requirements for the generation of robust solutions to be considerably reduced with respect to traditional optimization methods. Second, it allows the identification of the most informative regions of the scenario set containing the scenarios to be included in the optimization process for generating a robust solution. We test ROSS on the real case study of robust planning of an off-grid hybrid energy system, combining diesel generation with renewable energy sources and storage technologies. Results show that ROSS enables computational requirements to be reduced between 23% to 84% compared with traditional robust optimization methods, depending on the complexity of the robustness metrics considered. It is also able to identify very small regions of the scenario set containing the most informative scenarios for generating a robust solution. • A novel algorithm for robust planning under deep uncertainty is proposed. • The algorithm uses active learning for selecting the smallest subset of informative scenarios. • Computational time is significantly reduced with respect to traditional methods. • Most informative scenarios are highlighted and exploited within the optimization process. [ABSTRACT FROM AUTHOR]

• A dynamic, multi-objective approach to optimize off-grid water-energy systems is proposed. • Interdependency between system design and operations is considered. • The nexus between water and energy systems is modelled dynamically. • The approach outperforms the state-of-the-art non-dynamic, least cost approach. • The optimal solutions identified limit investment costs and environmental impacts. Small Mediterranean islands are remote, off-grid communities characterized by carbon intensive electricity systems coupled with high energy consuming desalination technologies to produce potable water. The aim of this study is to propose a novel dynamic, multi-objective optimization approach for improving the sustainability of small islands through the introduction of renewable energy sources. The main contributions of our approach include: (i) dynamic modelling of desalination plant operations, (ii) joint optimization of system design and operations, (iii) multi-objective optimization to explore trade-offs between potentially conflicting objectives. We test our approach on the real case study of the Italian Ustica island by means of a comparative analysis with a traditional non-dynamic, least cost optimization approach. Numerical results show the effectiveness of our approach in identifying optimal system configurations, which outperform the traditional design with respect to different sustainability indicators, limiting the structural interventions, the investment costs and the environmental impacts. In particular, the optimal dynamic solutions able to satisfy the whole water demand allow high levels of penetration of renewable energy sources (up to more than 40%) to be reached, reducing the net present cost by about 2–3 M€ and the CO 2 emissions by more than 200 tons/y. [ABSTRACT FROM AUTHOR]

We built an integrated modeling framework that captures individual power project characteristics within an energy system model that simulates three socioeconomic scenarios that harmonize land-use change, climate impacts on water availability, final energy demands, and climate policy options. Yet, in the short term, especially in the transition to a net-zero emissions energy system, hydropower represents a cheap alternative to displace fossil fuels, especially coal. Cost-optimal hydropower expansion. [Extracted from the article]

Unprecedented population growth and fast economic development in several African countries is driving large infrastructure investments for growing energy, food and water demands. Massive infrastructure expansion required to support rising energy demands and extensive irrigation projects could stress the natural resources leading to social an geopolitical tension. In order to guarantee long-term success and sustainability and to minimize potential negative impacts of these investments, a carefully crafted temporal planning and management of existing and growing infrastructure is required. The challenges of long-term planning under uncertainty have been tackled in the literature through the use of dynamic adaptive pathways which provide a valuable scheme to include time and feedbacks for adaptive policy design. Additionally, a suite of robustness frameworks are now ubiquitous in the water resources field to better deal with planning and management under uncertainty gearing towards exploratory analysis to discover the combination of states that can cause water systems to fail rather than attempting to predict future behavior. We advance these concepts through an integrated approach for optimal time sequencing of infrastructure expansion and its corresponding management strategies. This approach is demonstrated in the Zambezi River Basin, this transboundary basin is key for economic growth and poverty reduction in the region sustaining the basic needs of nearly 30 million people. It provides important environmental services essential to regional food security and energy production. The Zambezi system currently encompasses four hydropower dams and plans to extend its capacity by adding four additional reservoirs. The goal of this study is to generate efficient pathways that allow the temporal and spatial sequence of planning and management alternatives. This refers to the optimization of expansion of hydropower dams and irrigation districts and its corresponding management strategy. For each candidate pathway, an optimal control policy is generated using Evolutionary Multi-objective Direct Policy Search (EMODPS), in which the operation policies are optimized with respect to their performance in meeting energy and irrigation demands as well as environmental requirements. Additionally, the pathways are optimized by balancing the benefits and costs of the additional infrastructure investments which are activated by future projections of growing population rates which are assumed to drive larger water demands. Since rising demands are projected, infrastructure elements will inevitably be added with time. Once the timing pattern is identified, a new selection of optimal operations based on the new system configuration is selected to maximize its benefit. A final step is to test the robustness of the optimal planning/management sequence against a suite of climate scenarios to determine their suitability under potentially pressing future states of the world. This study intends to provide guidance to crucial infrastructure investment planning in developing countries coupled with their optimal management to better absorb expensive infrastructure expansions while meeting growing regional demands. [ABSTRACT FROM AUTHOR]

Nowadays, about one-half of the earth's land surface and more than half of global freshwater is covered by about 260 transboundary river basins worldwide, which are often managed by multiple, institutionally independent decision makers. Most of these international basins are characterized by the presence of many stakeholders with conflicting interests, leading to the rise of water conflicts over distribution, co-management, and utilization of the water resource. In particular, upstream and downstream conflicts arising from the development of new water infrastructures are inevitable and expected to intensify in the near future. This will put additional pressure on water resources, together with growing water/energy demands and increasing water scarcity induced by climate change. Water resources management in transboundary river basins has often been addressed from a single, centralized decision maker perspective, assuming full cooperation and information exchange among multiple stakeholders. Yet, this assumption rarely reflects the real socio-political framework in an international basin. On the other side, game theory has been used to represent individual decision makers and develop more acceptable solutions. However, this approach often overly-simplifies physical and social dynamics shaping transboundary contexts, and the non-linearities between them.In this work, we present a novel, robust decision-analytic framework to explore and compare different cooperation settings among multiple, conflicting stakeholders by means of an optimization approach based on Evolutionary Multi-Objective Direct Policy Search (EMODPS). The framework is tested on the Zambezi river basin in Africa, where we first quantify the differences in system performance achieved under three different cooperation levels under historical hydro-climatic and socio-economic conditions. Then, an a-posteriori robustness assessment step is conducted in order to explore how deep uncertainties in hydrological (e.g., streamflow) and socio-economic (e.g., energy demand, irrigation water demand) variables affect different optimal cooperation strategies in a transboundary context. Our results show the benefits of implementing a centralized rather than a non-centralized management strategy, assessing the value of full information exchange and basin-wide objectives needed to robustly maximize the system performance at the entire basin level under a variety of uncertain futures. [ABSTRACT FROM AUTHOR]

Globally, many countries are actively seeking to maximize the hydropower potential of major river basins, yielding proposals for constructing approximately 3,700 major dams in the near future. At present, cost-benefit analyses are the dominant approach for evaluating candidate dam designs, where the typical problem framing has the following features: (i) a single objective, in which potentially conflicting and heterogeneous objectives are monetized and aggregated into a single Net Present Value (NPV) or Levelized Cost of Energy (LCE) metric; (ii) dam sizing using a pre-defined operating rule; and (iii) a limited accounting for potentially non-stationary external uncertainties. These features in standard cost-benefit analyses strongly bias dam design methods as they fail to explore the interdependent nature of reservoir sizing and operations. In particular, by neglecting a careful consideration of alternative reservoir operating schemes, standard design frameworks are likely to result in over-sized, under-performing dams. Moreover, deeply uncertain changes in hydro-climatic variability and human demands motivate the need for highly flexible assessment frameworks that capture the key tradeoffs between users, how dependencies in dam sizing and operations influence these tradeoffs, and the ultimate robustness of the infrastructure systems. In this work, we combine MORDM (Multi-Objective Robust Decision Making) and EMODPS (Evolutionary Multi-Objective Direct Policy Search) with the specific focus on advancing coupled dam sizing (i.e., discrete reservoir volumes) and operation design (i.e., continuous release policy parameters) problems. We demonstrate the potential of this integrated, highly flexible assessment framework through an ex post analysis of the Kariba dam in the Zambezi river basin. The Kariba dam was built in 1960 and has struggled to follow its specified operational rule curve that originally claimed to maximize hydropower production, independent of irrigation users' demands. Our results show that careful exploration of the coupling of dam sizing and operations yields designs that strongly outperforms the performance achieved under current Kariba system. Moreover, we demonstrate that our integrated sizing and operations framework leads to a significant reduction in capital cost requirements (i.e., smaller reservoir sizes) while simultaneously improving system robustness with respect to changing hydro-climatology and human irrigation demands. [ABSTRACT FROM AUTHOR]

Engineering systems are designed to be robust to low-probability, high-consequence hazards. The concept of robustness has been crucial to the safety of our built environment. However, increasing robustness to one hazard must come at the expense of performance on other objectives. An interesting question is: Can we avert these opportunity costs by better conditioning water management decisions on forecasts? This study explores this question in the Red River Basin of Vietnam, where a system of reservoirs must simultaneously produce hydroelectric power, provide water to meet agricultural, municipal and environmental demands and, most importantly, prevent loss of life by protecting the capital city of Hanoi from devastating floods. The Vietnamese government seeks to use these reservoirs to provide protection to the 500-yr flood and has independently designed their own operating guidelines with this goal. However, optimizing operating policies to meet the multiple sectoral demands while providing protection to the 500-yr flood reveals that meeting this requirement with existing infrastructure is non-trivial. Policies exist that meet this goal under stationary hydrologic conditions, but doing so requires major sacrifices in performance on hydropower and water supply objectives. These tradeoffs are exacerbated if stakeholders also desire robust performance across potential hydrologic and socioeconomic futures; compared to optimizing to the 100-yr flood, it is far more difficult to find policies that also meet minimum performance requirements on hydropower and deficit objectives across a broad range of futures. Conditioning operations on a naïve forecast is partly able to close the gap, though, suggesting more sophisticated forecast products may hold promise in reducing the multi-sectoral consequences of robustness to severe flood events. Future work will explore the potential for seasonal climate forecasts to close the remaining gap and allow for robust flood protection at no cost. [ABSTRACT FROM AUTHOR]

Residential water consumption data collected at the household level with intelligent meters is being increasingly utilized, on the one hand, to inform utilities' operations and demand management strategies, on the other hand, to communicate transparent and personalized reports on water use to water consumers. Coupling intelligent metering systems with data analytics is a core element driving the transition towards digital water utility service providers. By intelligently mining into high resolution water consumption data, utilities can gain knowledge about heterogeneous water consumption routines, segment their customers into profiles based on behavioural similarities and differences, and ultimately tailor demand-side management interventions for the different groups. In this work, we present a data-driven customer segmentation analysis performed on the water use time series of 327 water consumer accounts from Southeast Queensland and Melbourne, each monitored in 2010 with intelligent meters sampling observation with an interval of 5 seconds. Our customer segmentation analysis is composed of three steps. First, we disaggregate the water consumption time series of each household into eight different end uses (i.e., tap, shower, washing machine, dishwasher, toilet, bathtub, irrigation, evaporative cooler) via Autoflow, a state-of-the-art algorithm for end-use disaggregation which combines Markov Models, Artificial Neural Networks, and Dynamic Time Warping. Second, we identify water end use routines for individual accounts (i.e., recurring hourly patterns of daily water use at the end-use level) by Principal Component Analysis. Finally, we cluster the routines to identify groups of accounts revealing similar water end use behaviours. Numerical results show that the data variance of primary routines is mainly explained by shower, washing machine, and irrigation end uses, thus suggesting three main behavioural classes exist among the considered accounts. Second, behavioural differences based on time-of-use, amount of water used for each end-use, weekday vs weekend routines, and regularity in the occurrence or primary routines emerge within each of the above classes. We thus demonstrate that our proposed data-driven customer segmentation analysis reveals the main end uses and consumption patterns driving heterogeneous types of water use behaviours. This brings a two-fold contribution to demand management: first, our customer segmentation analysis successfully mines high-resolution water consumption time series to provide water utilities with a concise profile characterization of their water consumers; second, our findings support the design of effective portfolios of customized feedback and demand management interventions. [ABSTRACT FROM AUTHOR]

Water losses are one of the main challenges posed worldwide to an efficient management and maintenance of urban potable water distribution networks. Large effort has been devoted in the literature to design algorithms to detect and localize water losses via physically-based hydraulic modelling, data-driven modelling, and probabilistic approaches. Most of the state-of-the-art research looks at developing such tools to prevent or mitigate both physical and apparent water losses in water networks and help utilities reduce water waste and costs due to non-revenue water. Along with the digital transition of urban water networks, the deployment of intelligent meters, and the acquisition of high-resolution data of water consumption at the household scale, a few data-driven methods have been also developed to detect post meter leakages, i.e., those leakages that occur within a residential property and that generate extra costs for water consumers. Yet, most of these studies consider single-household water use time series as taken individually and not connected within a network model. Thus, potential correlations among different water use time series, due, for instance, to location-based or behavioural similarities among different water consumption patterns are disregarded. Moreover, the limited availability of accessible water use time series observed via intelligent meters limits the scope and the data-driven nature of such methods to small-scale or theoretical experiments.In this work, we propose a demand- and network-based method for post meter leakage generation and detection. Our method integrates (i) the STREaM stochastic simulation model to generate synthetic time series of residential water demand at the scale of single fixtures (i.e., toilet, washing machine, tap, etc.), (ii) a module for pressure-driven simulation of water supply networks and generation of realistic leakages, and (iii) a data-driven learning algorithm for the automatic detection of post meter leakages. We test our method both with synthetic water demand data including post meter leakages, as well as in a lab-controlled setting where leakage values are set at the end-use notes in order to be comparable to real-world fixture leakages. Numerical results show that by integrating both water demand patterns and a network model, our method exploits spatial and temporal correlations among fine-resolution water demand time series to achieve improved accuracy and reduced amount of false alarms as benchmarked against state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

Global trends in population growth and rising economic prosperity are expected to increase the demand for energy, food, and water, particularly affecting rapidly-growing African economies. To support the production of clean energy and extensive irrigation projects, investment in large dams is increasing. To promote development pathways that are socially inclusive as well as environmentally and economically sustainable it is crucial to assess the social, economic, and environmental impacts of such funding, taking into account the interests of local stakeholders.In this work, we explore the effects of alternative development pathways on the water-energy-food nexus of the Omo River – Lake Turkana basin (Ethiopia and Kenya). In this peculiar region, dam infrastructure development and expanding agriculture might impact the local social, economic, and environmental structures. In the lower Omo valley, seasonal Omo floods support the practice of recession agriculture, which constitutes the livelihood of local indigenous tribes. Further downstream, on the Ethiopian-Kenyan border, the Omo river contributes about 90% of inflow volume to Lake Turkana. This endorheic lake is located in an extremely arid part of the Kenyan Rift Valley and its' water is key to support the local pastoralists and fishermen. We analyze the effects of alternative energy and food development pathways from a multi-sectoral perspective, assessing the trade-offs in respect of hydropower production, irrigation supply, indigenous recession agriculture and other ecosystem services in the Lower Omo as well as fish yield in Lake Turkana. In addition, we test the robustness of the designed pathways against a wide range of potential changes in both flow regimes and irrigation demands. We do so using a two stage framework to first identify optimal pathways using an optimization procedure with a simplified system model, and then testing the pathways identified as optimal using a more detailed (but computationally expensive) model of the nexus interactions. Results show that our approach provides an objective framework to identify solutions that are able to address the water-energy-food nexus tradeoffs and that perform robustly under changing climatic and societal conditions. [ABSTRACT FROM AUTHOR]

Advanced metering technologies coupled with informatics creates an opportunity to form digital multi-utility service providers. These providers will be able to concurrently collect a customers’ medium-high resolution water, electricity and gas demand data and provide user-friendly platforms to feed this information back to customers and supply/distribution utility organisations. Providers that can install low-cost integrative systems will reap the benefits of derived operational synergies and access to mass markets not bounded by historical city, state or country limits. This paper provides a vision of the required transformative process and features of an integrated multi-utility service provider covering the system architecture, opportunities and benefits, impediments and strategies, and business opportunities. The heart of the paper is focused on demonstrating data modelling processes and informatics opportunities for contemporaneously collected demand data, through illustrative examples and four informative water-energy nexus case studies. Finally, the paper provides an overview of the transformative R&D priorities to realise the vision. [ABSTRACT FROM AUTHOR]

Reservoir operations are central to our ability to manage river basin systems serving conflicting multi-sectoral demands under increasingly uncertain futures. These challenges motivate the need for new solution strategies capable of effectively and efficiently discovering the multi-sectoral tradeoffs that are inherent to alternative reservoir operation policies. Evolutionary many-objective direct policy search (EMODPS) is gaining importance in this context due to its capability of addressing multiple objectives and its flexibility in incorporating multiple sources of uncertainties. This simulation-optimization framework has high potential for addressing the complexities of water resources management, and it can benefit from current advances in parallel computing and meta-heuristics. This study contributes a diagnostic assessment of state-of-the-art parallel strategies for the auto-adaptive Borg Multi Objective Evolutionary Algorithm (MOEA) to support EMODPS. Our analysis focuses on the Lower Susquehanna River Basin (LSRB) system where multiple sectoral demands from hydropower production, urban water supply, recreation and environmental flows need to be balanced. Using EMODPS with different parallel configurations of the Borg MOEA, we optimize operating policies over different size ensembles of synthetic streamflows and evaporation rates. As we increase the ensemble size, we increase the statistical fidelity of our objective function evaluations at the cost of higher computational demands. This study demonstrates how to overcome the mathematical and computational barriers associated with capturing uncertainties in stochastic multiobjective reservoir control optimization, where parallel algorithmic search serves to reduce the wall-clock time in discovering high quality representations of key operational tradeoffs. Our results show that emerging self-adaptive parallelization schemes exploiting cooperative search populations are crucial. Such strategies provide a promising new set of tools for effectively balancing exploration, uncertainty, and computational demands when using EMODPS. [ABSTRACT FROM AUTHOR]

The Soil & Water Assessment Tool (SWAT) is a useful model for evaluating socio-ecological tradeoffs and analyzing coupled natural-human system dynamics in agricultural watersheds. However, reservoir operating options in SWAT are limited. This study advances the representation of reservoir operations in SWAT by adding an option for closed-loop, multi-reservoir operating policies that can be optimized using Evolutionary Multi-Objective Direct Policy Search. This enables water managers to evaluate the tradeoffs across more coordinated reservoir operations in SWAT while capitalizing on the model's ability to physically simulate hydrological processes better than traditional reservoir simulation models. Comparing our advanced reservoir operations with SWAT's existing operating options in the Omo River basin of Ethiopia, we find a wider range of policies for managing conflicting stakeholder objectives that better compromise across them and are more robust to climate change. These advances to SWAT's reservoir module show promise for informing integrated water resources management. • We integrate coordinated, closed-loop reservoir operations into the Soil & Water Assessment Tool. • This allows water managers to evaluate tradeoffs of smarter water management options. • It also improves the modeling of hydrologic impacts in traditional reservoir models. [ABSTRACT FROM AUTHOR]

Globally, the pressures of expanding populations, climate change, and increased energy demands are motivating significant investments in re-operationalizing existing reservoirs or designing operating policies for new ones. These challenges require an understanding of the tradeoffs that emerge across the complex suite of multi-sector demands in river basin systems. This study benchmarks our current capabilities to use Evolutionary Multi-Objective Direct Policy Search (EMODPS), a decision analytic framework in which reservoirs’ candidate operating policies are represented using parameterized global approximators (e.g., radial basis functions) then those parameterized functions are optimized using multi-objective evolutionary algorithms to discover the Pareto approximate operating policies. We contribute a comprehensive diagnostic assessment of modern MOEAs’ abilities to support EMODPS using the Conowingo reservoir in the Lower Susquehanna River Basin, Pennsylvania, USA. Our diagnostic results highlight that EMODPS can be very challenging for some modern MOEAs and that epsilon dominance, time-continuation, and auto-adaptive search are helpful for attaining high levels of performance. The ϵ-MOEA, the auto-adaptive Borg MOEA, and ϵ-NSGAII all yielded superior results for the six-objective Lower Susquehanna benchmarking test case. The top algorithms show low sensitivity to different MOEA parameterization choices and high algorithmic reliability in attaining consistent results for different random MOEA trials. Overall, EMODPS poses a promising method for discovering key reservoir management tradeoffs; however algorithmic choice remains a key concern for problems of increasing complexity. [ABSTRACT FROM AUTHOR]

Challenges to efficiently manage water resources for hydropower production include changes and increased variability in climate drivers as well as the pressing needs for sharing the water stored in reservoirs with other water uses to mitigate existing and projected conflicts. In this context, hydropower reservoir operations might benefit from the information coming from a broad range of prediction time scales, from short-term to seasonal and decadal time horizons. The increasing number of climate portals and services for the water sector has brought users of the hydropower sector to ask questions such as: Where is there still room for improvement in typical hydrometeorological forecasting and reservoir management systems designed for the hydropower sector? What do state-of-the-art climate services offer today that actually impacts the performance of hydropower operations? Where, within such systems, should we place our efforts today? What type of investment in the forecasting-reservoir management chain should be prioritized? Which is the dominant driver to inform hydropower dam operations: water availability or energy prices?This brief explores how matching available climate services with local hydrological-energy contexts and needs can be crucial to improve the operation of hydropower systems and impact decision-making. It includes results from four study sites covering the broad EU hydro-climatological and socio-economic contexts: a set of typical mountainous catchments in France, a typical snow-dominated Alpine basin in Italy (the Lake Como), a typical south Mediterranean basin in Spain, with a large share of water for irrigated agriculture (the Jucar River basin), and a typical north European catchment in Sweden (the upper part of the Umeälven River), highly influenced by snowmelt runoff and volumes for planning the hydropower production for the current and next winter seasons.Our main findings show that, when decisions are optimized based on future climate and energy prices, hydropower revenue associated with interannual reservoir storage can increase by 3% to 8%, when seasonal predictions from climate and hydrological models are used. Given the presence of multiple time dynamics in hydropower systems, strategic planning on the medium- (weekly) to long- (monthly to seasonal) term, which are mostly driven by climate and hydrology at the local scale, needs to be aligned with daily and sub-daily decisions, which are mostly driven by the energy market. This highlights how climate-oriented and market-oriented decisions have to cohabit in decision-making in hydropower reservoir operations. Acknowledgement: this work is supported by the IMPREX project funded by the European Commission under the Horizon 2020 framework program (grant 641811) [ABSTRACT FROM AUTHOR]

Climate change is expected to affect substantially future water availability and temperature across many watersheds in Europe potentially impacting hydropower generation and cooling efficiency of thermo-nuclear power plants. On the other hand, increasing Renewable Energy Sources (RES) penetration will substantially modify energy portfolios across Europe. As a consequence, prices will become more volatile, requiring more system flexibility. This study aims to explore the following key questions for supporting the definition and implementation of EU transition policies driving investments in the energy sector: 1) How climate change impacts on the hydrological system propagate to the energy system? 2) Can hydropower provide additional flexibility to a high RES-based energy system? To answer these questions, we developed an integrated modelling approaches combining a conceptual hydrological model (HBV) and an energy system model (PRIMES-IEM). Based on daily air temperature and precipitation, HBV computes daily streamflow for each power plant. Water temperature is obtained from air temperature using an empirical formula. These two variables are used to define the constraints to the energy model PRIMES-IEM, a power market simulator of the European electricity markets.Our analysis focuses on Iberian Peninsula and Danube region in 2040. Both regions' energy mix have a significant share of hydropower and rely significantly on thermo-nuclear power plants. In addition, both systems have already been under stress conditions due to heatwaves and droughts over the last years. On the climate side, we downscaled projected air temperature and precipitation projections for the RCP2.6 and RCP8.5 scenarios, covering the full range of possible climate futures from very low emission pathways to very high ones. On the energy side, we also consider two scenarios representative of the current system and of a future decarbonized system complying with European climate targets at 2030 and 2050. Despite some differences across the two regions, numerical results suggest some general trends: hot and dry climate futures significantly impact on the energy system because either increasing water temperature produces heavy stress due to thermonuclear generation curtailment or declining water availability limits hydropower generation forcing the energy system to shift to other technologies in order to maintain its reliability. [ABSTRACT FROM AUTHOR]

Small Mediterranean islands represent a paradigmatic example of remote, off-grid systems, where water supply and electricity generation are intrinsically coupled and interdependent. Distance from the mainland, lack of accessible water sources, and high seasonal variability of both water and electricity demand strongly and critically influence the operations of both water and energy systems. Energy security is nowadays based on unsustainable carbon intensive diesel generators, usually oversized to meet peaking summer electricity demand driven by high touristic fluxes. Drinking water is produced by energy intensive desalination technologies, which strongly impact on the electricity system increasing air pollution and greenhouse gas emissions. In order to improve the overall sustainability of small islands, the introduction of renewable energy sources coupled to power storage technologies and an efficient management of the water-energy system represent a viable and interesting solution to produce clean energy at lower costs. However, the identification of the optimal hybrid system design is strongly affected by the deep uncertainty in future climate and socio-economic drivers.The aim of this work is to identify robust hybrid renewable energy systems by directly including different plausible future scenarios of the main exogenous variables (e.g., solar radiation, wind speed, water/electricity demand) within the optimal system design phase. By adaptively selecting a concise set of future scenarios as the most informative, our multi-objective optimization approach allows to identify robust system designs that do not degrade when re-evaluated over a large out-of-sample set of scenarios.Results show the effectiveness of including the deep uncertainty in the main external drivers within the optimization process for generating solutions that are robust with respect to future changing conditions. Moreover, the adaptive selection of the most informative set of future scenarios allows to obtain satisfactory solutions in terms of system performance by considerably reducing the computational requirements. [ABSTRACT FROM AUTHOR]