Using Hydrologic and Hydraulic Simulations to Predict Near-term Effects of Beaver Habitat Restoration on Flood Flows

Author

Luke S. Stewart, California Polytechnic State University, San Luis ObispoFollow

Available at: https://digitalcommons.calpoly.edu/theses/3048

Date of Award

6-2025

Degree Name

MS in Civil and Environmental Engineering

Department/Program

Civil and Environmental Engineering

College

College of Engineering

Advisor

Stefan Talke

Advisor Department

Civil and Environmental Engineering

Advisor College

College of Engineering

Abstract

There is a growing need to predict the complex hydrologic and hydraulic effects of beaver-maintained wetlands. This study uses relatively simple inputs and a replicable methodology to predict flood flow impacts of beaver habitat restoration at the site design scale using the engineering industry’s customary hydrology and hydraulic modeling approach. This study focuses on Toro Creek, the primary drainage for a 15.2 square mile watershed on the California Central Coast.

A hydrologic model of Toro Creek watershed was developed in HEC-HMS to approximate saturated soil conditions typical of late-spring flooding events observed in the region. The hydrologic model was calibrated to an observed runoff event on February 19^th, 2024, using stage, flow, and rainfall data collected during the 2023-2024 water year and PRISM 1991-2020 annual normal precipitation data to estimate rainfall distribution patterns throughout the watershed. Two-dimensional hydraulic models of baseline conditions and a near-term beaver habitat restoration scenario were developed in HEC-RAS. These models were used to simulate observed, historical, and synthetic storm events. The historical storm events – January 9^th and March 10^th, 2023 – had estimated peak flows of 4722 cfs and 4738 cfs, respectively. The 100-year and 500-year estimated return interval storms were based on NOAA Atlas 14 rainfall estimates and had estimated peak flows of 7125 cfs and 9299 cfs, respectively.

Modeled beaver habitat with 8 dams over a 2.2km reach showed a small but significant peak flow attenuation (< 1%) and increase in lag time (7 minutes on average) relative to baseline conditions for storms smaller than an approximately 25-year event with effects diminishing for larger storms. The scale of modeled attenuation does not suggest that beaver reintroduction alone would be an effective flood control measure for Toro Creek. The scale of attenuation from modeling physical obstructions was similar to the scale of effects expected from flow through dam faces, hyporheic flow, and geomorphologic changes, suggesting that these factors will be necessary considerations to develop this modeling approach into a practicable methodology in industry.