Illuminating wildfire erosion and deposition patterns with repeat terrestrial lidar

Abstract

Erosion following a wildfire is much greater than background erosion in forests because of wildfire-induced changes to soil erodibility and water infiltration. While many previous studies have documented post-wildfire erosion with point and small plot-scale measurements, the spatial distribution of post-fire erosion patterns at the watershed scale remains largely unexplored. In this study lidar surveys were collected periodically in a small, first-order drainage basin over a period of 2 years following a wildfire. The study site was relatively steep with slopes ranging from 17 ∘ to > 30 ∘ . During the study period, several different types of rain storms occurred on the site including low-intensity frontal storms (2.4 mm h −1 ) and high-intensity convective thunderstorms (79 mm h −1 ). These storms were the dominant drivers of erosion. Erosion resulting from dry ravel and debris flows was notably absent at the site. Successive lidar surveys were subtracted from one another to obtain digital maps of topographic change between surveys. The results show an evolution in geomorphic response, such that the erosional response after rain storms was strongly influenced by the previous erosional events and pre-fire site morphology. Hillslope and channel roughness increased over time, and the watershed armored as coarse cobbles and boulders were exposed. The erosional response was spatially nonuniform; shallow erosion from hillslopes (87% of the study area) contributed 3 times more sediment volume than erosion from convergent areas (13% of the study area). However, the total normalized erosion depth (volume/area) was highest in convergent areas. From a detailed understanding of the spatial locations of erosion, we made inferences regarding the processes driving erosion. It appears that hillslope erosion is controlled by rain splash (for detachment) and overland flow (for transport and quasi-channelized erosion), with the sites of highest erosion corresponding to locations with the lowest roughness. By contrast, in convergent areas we found erosion caused by overland flow. Soil erosion was locally interrupted by immobile objects such as boulders, bedrock, or tree trunks, resulting in a patchy erosion network with increasing roughness over time.

Evolution of Debris‐Flow Initiation Mechanisms and Sediment Sources During a Sequence of Postwildfire Rainstorms

“…Aside from the debris‐flow and sedimentation hazards that it creates, erosion during the period of disturbance following wildfire may contribute substantially to denudation rates over geologic time (Orem & Pelletier, ) with implications for landscape evolution (Istanbulluoglu et al, ; Lavé & Burbank, ). Estimates of hillslope erosion derived here from repeat TLS add to the growing database of studies that utilize high‐resolution topographic data (e.g., DeLong et al, ; Orem & Pelletier, ; Rengers et al, ; Staley et al, ; Wester et al, ) to quantify the spatial patterns of erosion within a recently burned landscape. Likewise, the rainfall records and channel monitoring data provide additional constraints on the precise timing of debris flows within rainstorms, which can be combined with larger data sets of postwildfire debris‐flow activity throughout the western United States to improve empirical models used to estimate debris‐flow likelihood (Staley et al, ).…”

“…A 3‐m resolution DEM derived from prefire airborne lidar, which covers the entire study basin, was interpolated to a 1‐m grid for simulations at the basin scale. Since grid resolution can influence the degree to which water concentrates into preferential flow pathways, the use of a 1‐m grid was selected as a compromise that allowed for reasonable simulation times while also minimizing differences in grid resolution between the TLS‐generated DEM and the airborne‐lidar‐derived DEM (Rengers et al, ).…”

“…Other studies focused on the initiation of runoff‐generated debris flows also highlight observations of levee‐lined rills (e.g., Langhans et al, ; Nyman et al, ), further suggesting that rill erosion plays a role in achieving the high sediment concentrations required for debris flows. Several recent studies, however, using high‐resolution digital elevation models (DEMs) derived from terrestrial laser scanner (TLS) surveys, suggest that erosion from interrill areas may contribute to the majority of erosion on steep, recently burned hillslopes (e.g., Rengers et al, ; Staley et al, ). Despite intense rilling within their study area, Staley et al () used repeat TLS surveys to determine that, during a storm that produced multiple debris flows, roughly 57 % of erosion from their 0.01‐km 2 study area occurred on divergent hillslope locations.…”

Constraining the relative importance of raindrop‐ and flow‐driven sediment transport mechanisms in postwildfire environments and implications for recovery time scales

“…As a result, erosion in one burned area may be dominated by channel incision with relatively small contributions of total soil erosion coming from hillslope locations [Moody and Martin, 2001b]. At other postwildfire sites, soil erosion may occur primarily on hillslopes Rengers et al, 2016a]. The same site may even experience a different geomorphic response depending on the rainfall characteristics.…”

“…More recently, lidar-derived topographic data sets have also been utilized to obtain high-resolution, spatially complete estimates of erosion and deposition patterns following wildfire [e.g., DeLong et al, 2012;Pelletier and Orem, 2014;Staley et al, 2014;Wester et al, 2014;Orem and Pelletier, 2015;Rengers et al, 2016a]. Knowledge of spatial patterns in postwildfire erosion have proven to be useful for deriving spatially complete estimates of sediment yield, which can then be related to important factors such as burn severity or slope over large spatial scales [Pelletier and Orem, 2014], and identifying erosion "hot spots," or sediment source locations, that contribute material to create downstream hazards, such as debris flows .…”