Sustainable water management in the agricultural sector under deep uncertainty

Resumen

The world is experiencing a global water crisis, caused by the compounded effects of overexploitation, population growth, mismanagement, and climate change. The design of water-adaptation strategies to address this crisis has traditionally relied on consolidative models that offer decision-makers point predictions on economic impacts, water conservation, etc. However, in recent decades, nonlinearities (e.g., in climate change, adaptive behavior) have challenged the reliability of these models and overwhelmed existing policies, which have systemically failed to achieve their targets due to new correlations across complex and interconnected socioeconomic and ecological systems that were not previously anticipated. In this context, planning for the future is characterized by a high degree of uncertainty, or deep uncertainty. Under deep uncertainty, we cannot associate probabilities to outcomes (as with risk), and therefore we cannot individuate with confidence a single strategy that is expected to outperform the alternatives. Instead of looking for optimality, under deep uncertainty, decision-makers should prioritize robustness, i.e., the identification of the strategy(ies) that achieve the objective of sustainable and equitable economic growth in the most plausible futures. This calls for water reallocations from economic to environmental uses in order to guarantee the good ecological status of ecosystems, complemented by reallocations among economic uses to enhance both efficiency and equity. Most reallocation strategies will target the agricultural sector, which is the largest water user and concentrates the marginal (i.e., least valuable) use of the resource. This thesis presents a modeling framework to design and inform robust adaptation strategies in the agricultural sector, structured in 5 chapters. In the first chapter, we introduce the topic, state the objective, and present the structure of the thesis. The second chapter presents the socioeconomic methodology used to assess farmers' behavior (Mathematical Programming Models – MPMs) and introduces a new model that allows deficit irrigation as an adaptation strategy to water scarcity, usually not considered in conventional MPMs. In the third chapter, we introduce a novel multi-model ensemble of MPMs to sample uncertainty and thus inform robust decisions. In chapter four, we explicitly include the water systems coupling the multi-model ensemble of MPMs with a decision support system model used to manage water at basin level. Finally, chapter five includes the conclusions and recommendations of the author. The main purpose of our modeling framework is to deliver actionable science. Thus, it is designed to be modular, updatable, and ready to apply by policymakers. Thanks to the collaboration with Italian and Spanish regulatory authorities, we have applied our modeling framework to 5 policy cases: we tested the performance of two pricing policies, and a water bank to buyback water for the environment, we assessed a pecuniary compensation scheme designed to sustain irrigation-dependent ecosystem services, and we calculated the resource cost of agricultural water. Our multi-model ensemble can inform the identification and adoption of robust strategies that contribute to the targets of equitable and sustainable economic and welfare growth. Moreover, the explicit inclusion of the water system allows for the consideration of the co-evolution of the water and human systems, in order to avoid unfavorable outcomes triggered by the possible two-way feedback between these systems.