Can little ponds fight big floods?: A comprehensive analysis on the utility of activated distributed storage networks for flood peak reduction

Many studies have shown that flooding is increasing across the United States. As the planet warms and heavy rainfall becomes more common, this is likely to worsen. In the state of Iowa, the combined cost from the five largest floods since 1990 is over 20 billion dollars in damage, suggesting that current flood fighting methods, like building levees or storing rainfall in large reservoirs, may not be enough. While we can’t control where, when, or how much rain falls, we can focus on finding new tools to combat flooding. Reducing flood impacts requires additional places to store excess rainfall. The United States has over 90,000 dams, most of which are small and allow stored water to drain freely, thus squandering their potential for fighting floods. This dissertation investigates whether installing gated outlets at hundreds of ponds across a watershed and operating them as a system can meaningfully reduce stream flows after heavy rainfall, a technique called “activated distributed storage”. By modeling this approach to water management at 130 existing ponds in a small Iowa watershed, this dissertation aims to determine i) how an activated distributed storage system responds to rainfall of different durations and intensity, ii) whether operating the gates preemptively using forecasts can meaningfully impact flow outcomes, and iii) whether a dynamic operation framework that responds to the unique characteristics of individual rainfall events can lessen flood impacts. The result is a foundational investigation of active control of distributed storage networks that identifies many opportunities and limitations for flood reduction.