Erosion and Sedimentation from the Bagley Fire, Eastern Klamath Mountains, Northern CA

The Bagley Fire burned about 19,000 hectares on the Shasta-Trinity National Forest in the late summer of 2012, with soil burn severities of 11% high, 19% moderate and 48% low. Two strong storms in November and December followed the fire. The first storm had a recurrence interval of about 2 years, and generated runoff with a return interval of 10-25 years, causing many road stream crossing failures in parts of the fire. The second storm had a recurrence interval of 25-50 years, and initiated more severe erosion throughout the fire area. Erosional processes were dominated by sheet, rill and gully erosion, and landslides were uncommon. A model predicted high potential for debris flows, but few were documented, and though most stream channels exhibited fresh scour and deposition, residual deposits lacked boulder levees or other evidence of debris flow. Rather, deposits were stratified and friable, suggesting a sediment laden flood flow rather than debris flow origin. The resulting sediment was rich in gravel and finer particles, and poor in larger rock. Soil loss was estimated at 0.5-5.6 cm on most hillslopes. A high resolution DEM (LiDAR) was used to measure gullies, small landslides, and stream scour, and also to estimate sedimentation in Squaw Creek, and Shasta Lake. A soil erosion model was used to estimate surface erosion. Total erosion in the Squaw Creek watershed was estimated at 2.24 million metric tons, which equates to 260 metric tons/hectare. Of this, about 0.89 million metric tons were delivered to the stream system (103 metric tons/hectare). Nearly half of this sediment, 0.41 million metric tons, was temporarily stored in the Squaw Creek channel, and around 0.33 million metric tons of fine sediment were carried into Shasta Lake. Squaw Creek also delivered about 0.17 million metric tons of sand, gravel and cobbles to the lake. This estimate is very tenuous, and was made by measuring the volume of a delta in Shasta Lake from a tributary to Squaw Creek and extrapolating to the entire watershed. LidAR measurements of gully and landslide volume were considered the most reliable values, followed by estimates of channel scour and deposition in Squaw Creek and tributaries. The soil erosion model outputs were calibrated with data from a small debris basin. The most uncertain estimates were those for Shasta Lake sedimentation.