We are pleased to share a new publication in Water Resources Research developed in collaboration with colleagues from Duke University and the University of California Davis.

The paper explores a fundamental question in water resources management: what information should guide operational decisions? While advances in monitoring, remote sensing, forecasting, and environmental modeling have dramatically increased the amount of data available, most reservoir operating policies still rely on a limited set of variables, such as reservoir storage and seasonality.

This paper proposes a new perspective in which information selection becomes an integral part of policy design. We introduce a framework that classifies existing approaches to selecting policy inputs and discusses how observations, forecasts, and environmental indicators can be jointly optimized with decision policies. The paper also highlights the potential of these methods to improve monitoring systems, forecasting products, and adaptive water management under climate change and deep uncertainty. By bridging decision science, machine learning, and water resources management, this work identifies new research directions for developing more informative, adaptive, and resilient environmental decision-support systems.

The full paper is available here: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2025WR042745