18 February 2022 in Science
Proposed hydropower dams at more than 350 sites throughout the Amazon require strategic evaluation of trade-offs between the numerous ecosystem services provided by Earth’s largest and most biodiverse river basin. These services are spatially variable, hence collective impacts of newly built dams depend strongly on their configuration. We use multiobjective optimization to identify portfolios of sites that simultaneously minimize impacts on river flow, river connectivity, sediment transport, fish diversity, and greenhouse gas emissions while achieving energy production goals. We find that uncoordinated, dam-by-dam hydropower expansion has resulted in forgone ecosystem service benefits. Minimizing further damage from hydropower development requires considering diverse environmental impacts across the entire basin, as well as cooperation among Amazonian nations. Our findings offer a transferable model for the evaluation of hydropower expansion in transboundary basins.
18 May 2021 in Proceedings of the AAAI Conference on Artificial Intelligence
With the success of machine learning, integrating learned models into real-world systems has become a critical challenge. Naively applying predictions to combinatorial optimization problems can incur high costs, which has motivated researchers to consider learning augmented algorithms that can make use of faulty or incomplete predictions. Inspired by two matching problems in computational sustainability where data is abundant, we consider the learning augmented min weight matching problem where some nodes are revealed online while others are known a priori, e.g., by being predicted by machine learning. We develop an algorithm that is able to make use of this extra information and provably improves upon pessimistic online algorithms. We evaluate our algorithm on two settings from computational sustainability -- the coordination of unreliable citizen scientists for invasive species management, and the matching between taxis and riders under uncertain trip duration predictions. In both cases, we perform extensive experiments on real-world datasets and find that our method outperforms baselines, showing how learning augmented algorithms can reliably improve solutions for problems in computational sustainability.