Post-fire debris-flow hazard analysis for interstate 80, Truckee River Canyon, near the California-Nevada state line, USA

Abstract

On July 9, 2017, west of the California-Nevada state line, USA, the 2.8 km² Farad fire burned steep slopes above Interstate 80. On the evening of August 18, 2017, a localized convective storm produced short-duration, moderate-intensity rain, which in turn triggered debris flows in a historically inactive basin. These flows impacted Interstate 80. At least four additional debris flows, not related to wildfire, have occurred along this section of road since 2013, but during much higher intensity rainfall. Here we utilize the history of pre- and post-fire debris flows along this section of Interstate 80 to explore the impacts of wildfire on debris flows. Specifically, we combine pre- and post-fire rainfall data and field measurements with empirical debris-flow models to quantify the impacts wildfire had on debris-flow generation and to estimate the likelihood and magnitude of future events. A characteristic pre-fire debris flow occurred on July 25, 2013 in a susceptible path ~30 minutes after rainfall began and during peak 15-minute intensities of ~50 mm/hr. This event closed both east and westbound lanes of Interstate 80. No other nearby paths had debris flows during this rainstorm. In contrast, one month post-fire, on August 18, 2017 a debris flow occurred in a historically inactive path, but within an area of high burn severity. Debris-flow initiation occurred ~30 minutes after the beginning of rainfall, but with peak 15-minute intensities of only ~26 mm/hr. This amplified rainfall-runoff response is consistent with fire-induced changes in soil hydraulic properties for which we measured post-fire decreases of a factor of 2 in field-saturated hydraulic conductivity and a factor of 4 in sorptivity. From field measurements, total volume estimates for the August 18, 2017 post-fire debris-flow event ranged between 1270 and 4700 m3 depending on assumptions regarding pre-event channel geometry and volume of hillslope sediment transported. A shallow landslide that liquefied and flowed into the channel contributed ~450 m³ of material and was apparently triggered by concentrated overland flow off an old road into the toe slope of a much older deep- seated landslide. Debris flows eroded most of the travel path above the fan to bedrock and contributed >850 m³ of debris, at a nearly uniform spatial rate, both of which suggest this event was likely limited by sediment availability. Just 100-150 m above Interstate 80 the flow transported boulders with maximum diameters in excess of 1 m, at peak velocities of ~2-5 m/s. We used the analysis of the August 18, 2017 debris-flow event to verify empirical equations developed by the USGS for predicting the probability, total volume, and runout distance of post-fire debris flows. We found good agreement between model output and observations and hence used these equations to predict characteristics of debris flows likely to occur in the near future.