Skip to main content
Author(s):
Alicia M. Kinoshita, Brenton A. Wilder
Year Published:

Cataloging Information

Topic(s):
Fire Effects
Ecological - Second Order
Water
Post-fire Management
Post-fire Rehabilitation
Erosion Control

NRFSN number: 23362
Record updated:

Fire frequency and severity in southern California and across the western United States is increasing, posing a concern to the safety and well-being of communities and ecosystems. Increased aridity coupled with water stressed vegetation from prolonged droughts are leading to a higher propensity for larger, more intense fires that impact ecohydrological processes. Accurate characterization of these processes will improve rapid response efforts and long-term resource management to promote resilient communities along the wildland-urban interface. This work investigates prediction tools for small watersheds, where post-fire effects occur at a disproportional rate, by presenting methods to improve rapid predictions of post-fire streamflow and long-term monitoring of ecohydrological recovery. A random forest machine learning algorithm with 45 watershed parameters was created to predict post-fire peak streamflow for 1920 to 2019. This flood forecasting technique incorporated additional characteristics about meteorological and watershed properties to improve predictions of peak streamflow compared to flood frequency methods such as Rowe et al. (1949). The time elapsed after fire, peak hourly rainfall intensity, and drainage area were important factors that represented realistic conditions and increased accuracy of the random forest predictions. We used the case of the 2018 Holy Fire in southern California to characterize pre-fire climate and vegetation interactions and monitor post-fire recovery of ecohydrological processes (rainfall- runoff and evapotranspiration) for unburned (Santiago) and burned (Coldwater) catchments. ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station (ECOSTRESS), Operational Simplified Surface Energy Balance Model (SSEBop), satellite-based vegetation indices, and local rainfall-runoff data were incorporated into our analyses. Consistent with the drought conditions in California from 2012 to 2018, we observed low precipitation and evapotranspiration prior to the fire. Further, large pre-fire vegetation biomass and areas containing montane hardwood species were more likely to be classified as high soil burn severity. Between ECOSTRESS and SSEBop there was larger variability in evapotranspiration estimates after fire compared to pre-fire, which had implications for post-fire vegetation recovery and water storage. The water balance highlighted variability in predicted storage between burned and unburned catchments, which was dependent on the evapotranspiration model used. ECOSTRESS PT-JPL model was more sensitive to parameters such as land surface temperature, net radiation, slope aspect, soil burn severity, and vegetation species due to higher spatial resolution. The findings of this research improves upon our current methods in modeling post-fire peak flows and post-fire vegetation recovery in southern California and has potential for future applications in management and planning.

Citation

Kinoshita AM, and Wilder BA. 2021. Improving post-wildfire peak streamflow predictions for small watersheds and communities. Final Report for Joint Fire Sciences Project 19-1-01-55, June 2021, 37 pages.

Access this Document

The website encountered an unexpected error. Please try again later.