Large-scale remote sensing of geomorphic change in mulched and unmulched watersheds burned in the 2020 East Troublesome Fire, Colorado
Date
2023
Journal Title
Journal ISSN
Volume Title
Abstract
Elevated levels of sediment transport in post-wildfire landscapes can degrade the hydrologic and geomorphic processes of a river system, damage aquatic habitat, and pose a threat to downstream infrastructure. Hillslope mulching applications have proven to be effective at mitigating runoff and erosion at plot and hillslope scales but the impacts of mulching at the watershed scale remain generally unknown. We conducted repeat aerial surveys of one unmulched and five partially mulched watershed outlets (0.61-1.44 km2) to quantify erosion and deposition in the East Troublesome Fire burn scar. The objectives of the study were (1) to quantify volumes of erosion and deposition for hillslopes and channels for a variety of sites at a range of elevations (2) to identify and quantify the drivers of erosion and deposition and their relative contributions within and across watersheds (3) to determine the impact of slope, width, and vegetation cover on sediment storage and transport within watersheds; and (4) to assess the impacts of a large-scale aerial mulching operation at scales from hillslopes to watersheds. Multiple drone flights were conducted for each study site between July and October 2022. The earliest and latest surveys were differenced to produce DEM of Difference (DoD), with spatial resolutions ranging from 3.8 to 4.4 cm. Vertical uncertainties calculated from measurement uncertainty and Structure from Motion (SfM) errors were filtered out of the DoD at a 95% confidence interval (CI), resulting in maximum and mean detection thresholds of 11 and 4 cm, respectively. A supervised classification algorithm was used to filter out changes due to vegetation growth and decay, which varied in effectiveness across the six study sites. Hillslope erosion and deposition volumes were at least three times higher than near-channel volumes, with most sites being an order of magnitude higher. However, near-channel erosion and deposition magnitudes normalized by area were higher than normalized hillslope magnitudes at all sites. A bootstrap forest regression model was used to determine relationships between various site-specific parameters and erosion and deposition for each watershed individually, and for all six sites combined. The model indicated mean slope, absence of vegetation, mean differenced normalized burn ration (dNBR), and hillslope length to be strong drivers of erosion and deposition for the individual models. Total precipitation accumulation and maximum 60-minute rainfall intensity were stronger contributors in the combined models. Near-channel storage and transport was influenced by local relationships between width, stream power, and absence of vegetation. Mulch coverage area was found to be weakly correlated with erosion and deposition at the watershed scale, with contributions possibly being dependent on coverage rate. These findings emphasize the importance of applying mulch in areas where it is both necessary and can have a measurable impact on reducing erosion rates.
Description
Rights Access
Subject
geomorphology
remote sensing
wildfire
mulch
erosion
structure-from-motion