MODELING AND REMOTE SENSING OF POSTFIRE DEBRIS-FLOW SUSCEPTIBILITY AT REGIONAL SCALES

Following wildfire occurrence, reduced tree canopy interception and increased soil water repellency can substantially increase surface runoff and lead to destructive and potentially catastrophic debris flow events. To advance debris-flow susceptibility prediction capabilities, we adapt the National Oceanic and Atmospheric Administration (NOAA)’s National Water Model, WRF-Hydro, to simulate postfire hydrology. In particular, we simulate the hydrological conditions including overland flow and channelized streamflow that caused multiple debris flows within the Dolan burn scar of California in January 2021. These debris flows were triggered by a landfalling atmospheric river that dropped more than 300 mm of rain onto the central California Coast Ranges in four days, with a peak precipitation rate of 240 mm/hr. We use radar-informed Multi Radar/Multi Sensor System (MRMS) precipitation data at 1-km spatial and hourly temporal resolutions as a boundary condition. To replicate the burn scar effects on hydrology, we modify the land cover and soil infiltration parameters in WRF-Hydro. When burn scar conditions are added, our model’s ability to reproduce the U.S. Geological Survey (USGS) stream gage observations increases substantially – Nash-Sutcliffe Efficiency scores increase from negative values to 0.84, 0.73, and 0.53 at three gages located downstream of the burn scar. After validating the model, we divide WRF-Hydro’s simulated peak overland flow and streamflow discharge by the corresponding catchment area to indicate debris-flow susceptibility. Our results show that WRF-Hydro simulated debris-flow susceptibility estimates agree well with the normalized difference vegetation index (rdNVDI) calculated from Sentinel-2 pre- and post-event satellite imagery, i.e., catchments with high simulated susceptibilities correspond well with remotely sensed vegetation loss – likely due to the occurrence of debris flows or debris floods. This study suggests that WRF-Hydro may be a promising tool for process-based hazard assessments and regional-scale prediction of postfire debris-flow risk.