Integration of In Situ and Remotely Sensed Data for Post-Fire Flood Risk Modeling

Abstract:

Flood risks are growing below the wildland-urban interface. Urban development continues to expand across the southwestern U.S. into areas below mountain wildlands that experience cycles of drought, fire, flooding, and debris flows. Events such as the Thomas Fire in Southern California and the subsequent Montecito debris flows in 2018 starkly illustrate the potential for fatalities and property damage. Predicting sediment fluxes from recently burned catchments is a critical need for emergency preparedness and the design of stormwater infrastructure including debris basins and flood control channels. However, previous research shows that it is difficult to predict sediment fluxes without order-of-magnitude levels of uncertainty [1]. Reductions in uncertainties could help to make more reliable forecasts of dangerous conditions warranting evacuation, and help to avoid the costly and environmentally damaging over-design of flood control infrastructure. This calls for an improved process-based understanding of the connections between wildfire effects and the catchment-scale hydrogeomorphic response [2]. Herein we present a synthesis of in situ and remotely sensed data characterizing the catchment-scale hydrogeomorphic response of several catchments in Riverside County, California, affected by the Holy Fire of 2018. Riverside County Flood Control and Water Conservation District deployed instrumentation to monitor precipitation, runoff, and debris basin capacity, and this data was supplemented by remotely sensed observations of gauge-adjusted radar precipitation, burn severity, vegetation cover, and soil moisture as well as soil property data from soil surveys. Synthesis of remotely sensed and in situ measurements provides insight into soil conditions and vegetation levels that yield hazardous erosion and flooding from relatively small storm events. Additionally, multiple regression analysis provides an improved understanding of the remotely sensed data that is most useful for constraining estimates of sediment volumes deposited by sediment-laden floods.

[1] Gartner, J. E., Cannon, S. H., & Santi, P. M. (2014). Empirical models for predicting volumes of sediment deposited by debris flows and sediment-laden floods in the transverse ranges of southern California. Engineering Geology, 176, 45–56.

[2] Hyde, K. D., Riley, K., & Stoof, C. (2016). Uncertainties in Predicting Debris Flow Hazards Following Wildfire. In K. Riley, P. Webley, & M. Thompson (Eds.), Geophysical Monograph Series (pp. 287–299). John Wiley & Sons, Inc.