A Climate Change Informed Decision Making Framework for Mono Lake Diversions and Recovery

Mono Lake is a saline lake located in California’s Sierra Nevada that is incredibly important ecologically and culturally. The lake has no natural outflow, but in the 1940s, the LA Department of Water and Power (LADWP) began diverting water from Mono Lake as part of the Los Angeles Aqueduct system which provides crucial water resources to the city of Los Angeles. Since the inception of these diversions, the lake level has declined substantially, increasing the lakes salinity and altering the aquatic and shoreline habitats that support the lakes unique ecosystem. The exposed alkali lakebed has also caused an influx of PM10 particulates into the surrounding region, resulting in some of the worst air quality conditions in the US. In 1994, LADWP was ordered to protect Mono Lake and a new diversion policy was selected using lake level models, which showed the lake would reach the target recovery level within 20 years. However, this new policy, in tandem with anomalously dry recent conditions, has not yielded the expected outcome and the lake levels are still well under the target recovery level.

As a result, regional stakeholders, water managers and researchers have been working to address the failures of past lake level models and implement a new diversion policy. This incredibly difficult task is even further complicated by the uncertainty surrounding the impact of climate change on recent and future hydrological conditions, which current lake level models neglect to consider. In this study, we aim to improve future Mono Lake level predictions through the development of a climate change driven lake level model. We then use this model to implement a flexible policy optimization in order to support decision makers as they balance numerous and often conflicting stakeholder objectives.

To evaluate future conditions, our model predicts annual Mono Lake levels from 2024 to 2100 based on the atmospheric and hydrologic conditions from a suite of downscaled CMIP6 data. Our modeling approach was informed by technical experts from multiple agencies and organization that specialize in Mono Lake hydrology, infrastructure and water management. We use single realizations from 11 GCMs under SSP245, SSP370 and SSP585 to capture a wide range of possible future climates and natural variability sequences. To account for varying water management strategies, we implement flexible diversions based on lake level and hydrologic conditions, which allows us to evaluate a wide range of policies. In collaboration with multiple stakeholders, we select key performance objectives and use an iterative optimization algorithm which generates and tests thousands of potential policies in order to narrow in on the best performing policies. Our approach is designed to optimize performance across several objectives as well as balance conflicting objectives and quantify any compromises. The intention of this is to provide decision makers with a flexible framework to understand the limits of the system and explore different ways of balancing various constraints and priorities.