The frequency, magnitude and connectivity of post-wildfire rainfall-runoff and sediment transport

Abstract

Wildfire increases the likelihood of runoff, erosion, and downstream sedimentation in many of the watersheds that supply water for communities across the western U.S. The goal of this research was to examine the complex interactions between fire, rainfall and landscape properties (e.g., burn severity, topography) across scales from hillslopes to watersheds. The research combines both regional data analysis and field monitoring to examine the frequency, magnitude and connectivity of post-fire rainfall-runoff events and associated sediment delivery. In the first part of this study (chapter 2), the goal was to quantify rainfall thresholds that cause runoff and sediment delivery across multiple fires, years post-fire, spatial scales, and mulch treatments in the Colorado Front Range. Rain intensity thresholds were identified for plots, hillslopes, and watersheds across three Colorado Front Range fires. Thresholds did not significantly differ among fires for any year post-fire, but were significantly different between spatial scales and years post-fire. Thresholds increased with time since burn likely due to vegetation regrowth, litter accumulation and recovery of soil infiltration capacity. The frequency of storms exceeding thresholds for runoff and erosion was mapped across Colorado to provide a tool for identifying areas most vulnerable to post-fire runoff and sediment delivery and prioritizing post-fire treatments. In chapter three, the goal was to improve understanding of the catch efficiency of sediment fences commonly used to measure post-fire hillslope erosion. During post-fire year two (2014) of the 2012 High Park Fire four sediment fences were modified to collect and measure both the sediment deposited behind the fence and the amount of runoff and sediment that overtopped the fence. Sediment fence catch efficiency ranged from 28-100% for individual events and from 38-94% across the sampling season. Increasing rainfall intensities were correlated with greater runoff and total sediment loads and lower sediment fence catch efficiencies. Enrichment ratios indicate that the sediment behind the fence was significantly enriched in sand relative to the hillslope soil samples. These results indicate that sediment fences underestimate sediment yields and demonstrate how sediment particle sizes may be sorted en route to the stream network. In chapter four, the goal was to examine connectivity between hillslopes and channel networks. Runoff and sediment from nested hillslopes (n = 31) and catchments (n = 12) were assessed for two rainfall events with different duration and intensity during post-fire year three (2015) of the High Park Fire to determine the factors affecting connectivity. The first event had a return interval of <1 year with low intensity rainfall over an average of 11 hours, whereas the second event had high intensity rainfall that lasted for an average of 1 hour with a maximum return interval of 10 years. The lower intensity event led to low hillslope sediment yields and widespread channel incision. The higher intensity event led to infiltration excess overland flow, high sediment yields and in-stream sediment deposition and fining. During both events, the percent of a catchment that burned at high severity was positively correlated with sediment delivery ratios and area-normalized absolute channel change. Overall, this research demonstrated that the rainfall events and thresholds associated with the generation of post-fire runoff and sediment transport vary with spatial scale and time since burn. In addition, not every threshold-exceeding event will produce the same type of response due to the complex and transient nature of post-fire responses from hillslope to watershed scale. Increasing our understanding of post-fire responses and connectivity will therefore require nested multi-scale monitoring over time to determine how sediment moves to and through channel networks.