Publication Date

8-2017

Date of Final Oral Examination (Defense)

5-10-2017

Type of Culminating Activity

Thesis

Degree Title

Master of Science in Geosciences

Department

Geosciences

Major Advisor

Jennifer Pierce, Ph. D.

Advisor

Eric Lindquist, Ph.D.

Advisor

Brittany D. Brand, Ph.D.

Abstract

Wildfires increase erosion in mountainous landscapes. The most catastrophic form of post-fire erosion is the debris flow, viscous slurries of water and sediment capable of scouring and entraining larger sediment and rafting boulders. Post-fire debris flows are particularly hazardous when fire- and debris flow-prone landscapes intersect the Wildland Urban Interface (WUI). Homes built into the edge of the flammable WUI are at high risk of both wildfire and subsequent debris flows in mountainous landscapes of the western US, yet the WUI is expanding at an extraordinary rate. There are predictive models that inform citizens, land managers, and local governments of post-fire debris flow hazards they face, but they are rarely used at the WUI, where their use may be particularly beneficial.

Wildfire significantly increases the ability of landscapes to erode; post-fire soils are damaged, ash-laden and potentially hydrophobic. Damaged hillslopes previously protected by vegetation are directly exposed to rainfall where, on steep slopes, soil and ash are easily mobilized, channelized and capable of entraining larger and greater amounts of sediment as runoff moves downslope, forming a debris flow. Vegetation, soils and slopes vary across ecosystems; forested slopes have larger fuels that burn at higher severity, deeper, finer soils, and steeper slopes than those of rangeland ecosystems. While both ecosystems produce post-fire debris flows, more sparsely vegetated rangelands slopes may not be limited by fire to erode. Instead, rangeland systems may erode more continually and at lower magnitudes than forested slopes, whereas punctuated disturbance by fire on burned, previously forested slopes more often may lead to catastrophic failures, often by debris flows.

This thesis compares model estimates of post-fire debris flow probability and volume between forest and rangeland ecosystems within the fire- and erosion-prone rangeland-forest ecotone of the Boise Foothills above the Boise Metropolitan Area, Idaho USA. Models developed by the United States Geological Survey estimate post-fire debris flow probability and volume using soil, burn severity, topography and rainfall attributes, which have distinct characteristics between forest and rangeland ecosystems. This thesis also compares post-fire debris flow model estimates to a historic post-fire debris flow event that occurred within burnt range-grassland slopes after a summer convective storm in the Boise Foothills. We compare modeled volume and probability estimates to recorded debris flow locations and volumes of the 1959 “Boise Mudbath” to determine if models can accurately predict a real-world event.

Our findings show that models estimate higher post-fire debris flow probability and volume for forested basins vs. rangeland basins. The average modeled sediment yield is ~1.4x higher for forested basins than rangeland basins under both the low (2 yr) and high (100 yr) precipitation recurrence interval scenarios. The average post-fire debris flow probability is ~15% and ~32% greater for forested basins than rangeland basins under the 2yr and 100yr recurrence rainfall events, respectively. We also find that models over-predict sediment yields and under-predict probability of debris flow occurrence compared to the 1959 Mudbath event. We found that the post-fire debris flow model volume estimates were ~2-6x greater than those actually produced by the 1959 post-fire debris flow event. True 1959 debris flow yields are similar to those calculated for regional depositional records of sparsely vegetated drainage basins. Interestingly, modeled 1959 debris flow yields more closely match (~50% probability of debris flow occurrence under the 1959 post-fire debris flow scenario, despite the fact that all basins did in fact produce debris flows. These findings show that debris flow sediment yields appear to be distinct between forest and rangeland basins and conclude that post-fire debris flow models are more suited to forested slopes, as sediment yields appear to be distinct between burned rangeland and forested drainage basins.

The science produced in the geology section of this thesis was provided to City, County and State land and hazard managers to inform decision making regarding post-fire debris flow hazards. This transfer of knowledge from science to decision-maker lends itself to a seemingly simple question: how will this science be used to make decisions? There is a growing supply and demand of science addressing wildfire hazards at the Wildland Urban Interface, yet what makes science usable and how it is used to make policy decisions is not well understood. In Chapter IV of this thesis, we merge quantitative and qualitative social science methods with public policy theory to identify how stakeholders at the Boise, Idaho WUI use science to inform wildfire hazard policy. We hypothesize that how a manager defines a wildfire problem will determine how that manager uses science to inform a policy solution to that problem. To test this hypothesis, we performed content analysis on policies of wildfire stakeholders at the Boise WUI, and coded the policies into distinct categories that classify how they address wildfire hazards. We then conducted interviews with managers representing local, state and federal stakeholders in the Boise WUI to discuss how new, local science may address wildfire hazards they identify as needing policy solutions. Our findings show that stakeholders at the Boise WUI address the similar wildfire hazards with unique policy solutions. Interviews reveal that science is most useful when it is quickly understood, and when it can help draw boundaries from which wildfire hazard funding can be allocated and prioritized. We recommend the framework used in this study to provide policy context to scientists as they discuss their results with interested stakeholders, and to managers requiring policy context to the wildfire science they are asked to consider.

DOI

https://doi.org/10.18122/B2WB04

Share

COinS