IN-WOP project WP3 - Seine River study case: Risk-based management rules and assessment of fairness under climate change
2023-09-13 | UMR G-EAU
Chapter 1 Introduction
1.1 The IN-WOP project
Current practice in integrated water management predominantly use multi-objective optimization approaches with aggregated objectives. This biases results towards the status quo and against innovative solutions, can foster stakeholder resistance when they do not recognize their values and objectives in the optimization formulation, while also raising ethical concerns related to the inclusion of undesirable and/or hidden trade-offs. In contrast, many-objectives optimization approaches can consider many non-aggregated objectives, which has the potential to enrich the solution space with alternative courses of action that better reflect the diverging perspectives of stakeholders, and align better with ethical concerns. From the viewpoint of ethics, disaggregated assessment criteria are preferred as these may avoid undesirable and hidden trade-offs.
The overarching aim of this project is to investigate the contribution of many-objective optimization approaches to IWRM. For this, we use three water management cases in a comparative evaluation of many-objective approaches in diverse hydrological and cultural setting. In Italy our focus is on the Lake Como Basin located in the Italian Southern Alps, serving irrigated agriculture and competing demands from navigation, fishery, energy production environmental and flood protection. In the Seine River, France, we focus on the coordinated regulation of the Seine river discharge to reduce both floods and droughts. In Tunisia we address the anthropogenic impact on the management of water resources in the Merguellil Basin.
1.2 Work package 3: The Seine River Basin study case
Our approach of many-objective optimization consists in performing many single-objective optimizations and to derived management rules from prioritization of these objectives.
To do so, we first evaluate the risk of non-achievement of each objective independently for a given climatology and a given state of the system. Then, we derive management rules by prioritizing the riskiest objectives in the daily decision making.
The main tasks performed for the work package 3 are summarized in Figure
First, catchment uninfluenced flows are modeled at a daily time step with a semi-distributed GR4J model based on the airGRiwrm R package (Dorchies, Delaigue, and Thirel 2022) and forced by 11 GCM/RCM scenarios for both RCP4.5 and RCP8.5 between 1950 and 2100.
Then, these flows are used to assess the risk of non-achievement of each objective taking into account the current reservoir volume, the day of the year and a selection of climate scenarios and periods. This assessment is derived from the statistical distribution of the minimum (resp. maximum) volume required in the reservoirs for a given drought (resp. flood) objective calculated by a single objective dynamic programming optimisation.
The result of this assessment is available to the public through an interactive Shiny website (http://irmara.g-eau.fr) that allows to experiment management scenarios in real time.
Finally, “risk-based” management rules are derived by prioritising the riskiest objectives and balancing proactive and reactive decisions taking into account a hedging policy.
At the same time, we have modeled the current management rules over historical and future periods and we use these outputs to compare this management over the “risk-based” management rules. Both modelling results are specifically analysed for their distributive justice.
1.3 Abstract
This report aims at demonstrating the merits of many-objective optimization explicitly accounting for ethics and equity concerns in the Seine River basin. Mitigation of drought and flood rely on objectives that are often a combination of several flow thresholds to be respected for different locations downstream the reservoirs. Especially in the context of several reservoirs in parallel, multiple-objective optimization techniques show their limits because of the curse of dimensionality and because the aggregation into composite objectives results in potential resistance by stakeholders(Pianosi, Dobson, and Wagener 2020).
To tackle this issue, for the Seine River case study, we propose an approach in which we first compute the limit management of each independent objective using dynamic programming from uninfluenced flow time series obtained by a hydrological semi-distributed model. Then, we assess the risk of non-achievement of each objective for a given climatology and a given state of the system. Finally, we derive management rules based from reactive and proactive decisions combining the prioritization of the riskiest objectives and a hedging policy based on risk aversion decided by the stakeholders. The hedging policy consists in discarding objectives subject to reactive decision when objectives with higher priority exceed a predefined risk level undermining future proactive decision.
The hydrological model forced by GCM/RCM projections allows carrying out the risk assessment on sliding periods, which provides the benefit to take care of climate unstationarity in the long-term decision process. Even in the complex context of many conflicting objectives, this decision support system provides comprehensive algorithm that allows the water manager to justify the taken decisions leading to more transparency and justice in the decision process.