Stream flow modeling tools inform environmental water policy in California

Abstract

Management of California's vast water distribution network, involving hundreds of dams and diversions from rivers and streams, provides water to 40 million people and supports a globally prominent agricultural sector, but it has come at a price to local freshwater ecosystems. An essential first step in developing policies that effectively balance human and ecosystem needs is understanding natural stream flow patterns and the role stream flow plays in supporting ecosystem health. We have developed a machine-learning modeling technique that predicts natural stream flows in California's rivers and streams. The technique has been used to assess patterns of stream flow modification, evaluate statewide water rights allocations and establish environmental flow thresholds below which water diversions are prohibited. Our work has informed the statewide Cannabis Cultivation Policy and influenced decision-making in more subtle ways, such as by highlighting shortcomings in the state's water accounting system and building support for needed reforms. Tools and techniques that make use of long-term environmental monitoring data and modern computing power — such as the models described here — can help inform policies seeking to protect the environment while satisfying the demands of California's growing population.