Method for quantitative economic risk assessment of post-fire debris flows, A

Abstract

This thesis presents a method to estimate economic risk from post-fire debris flows. This method combines new geographic information system data-extraction strategies that expand on previously existing post-fire debris-flow hazard models to quantify damages and economic risk from individual burned basins and/or entire burned areas. The method was used to model damage and economic risk from post-fire debris flows based on two storm-scenarios at three case-study sites in the western United States. Charts comparing damage, risk given storm occurrence, and annual risk show that for a given basin, damage and economic risk associated with the 10-year storm are moderately higher than for the 2-year storm, assuming that the storm will occur. However, when the annual probability of the storm is included in the calculation, risks associated with the 10-year storm are significantly less than those associated with the 2-year storm. For each of the three case-study sites, results from all basins were combined to create site-wide estimates of damage and risk. Site-wide estimates were compared and contrasted with each other using graphical analysis; the hypothesis that intensity of development is a significantly more important contributor to economic risk than debris-flow hazard was tested using one-way analysis of variance. Results of graphical analysis suggest that differences in intensity of development and value of elements-at-risk are important, but results of the statistical analysis indicate that there is insufficient evidence at a significance level of 0.1 to support the hypothesis with the limited data set used. Debris-flow runouts for the studies described in this thesis were modeled using LAHARZ (Schilling, 1998). The LAHARZ model utilizes a pair of semi-empirical relationships between debris-flow volume and the cross-sectional (LAHARZ A parameter) and planimetric (LAHARZ B parameter) areas inundated by a lahar or debris flow to model the expected runout and footprint in a GIS. Uncertainty in these parameters, as well as uncertainty in the location of the onset of debris-flow deposition can have significant influence on the expected damages from a modeled debris-flow event. A sensitivity analysis was performed by modeling damage and risk on a subset of three basins from one of the case-study sites in order to quantify how uncertainty in the input parameters to the LAHARZ model influences estimated damages and associated economic risk. Evaluated parameters included debris-flow volume, LAHARZ A and B parameters, and location of onset-of-debris-flow deposition. Results of the analysis indicate that variability of estimated damages resulting from uncertainty in the debris-flow volume model and planimetric area (LAHARZ B parameter) greatly exceed the variability observed between the 2-year- and 10-year-storm baseline models. The location of onset of deposition also strongly influenced the estimated damages; however, variability in cross-sectional area of flow (LAHARZ A parameter) had relatively little influence on estimated damages for the basins considered in the study. At the locations evaluated, the greatest damage occurred when onset of deposition started at the basin mouth; however, relatively large damages were estimated even when onset of deposition started several kilometers upstream from the basin mouth. Results of the sensitivity analysis suggest that future research aimed at reducing uncertainty in debris-flow volume estimates, the LAHARZ B parameter, and location of onset of deposition could significantly improve risk estimates based on these models. Emergency managers need to make rapid decisions to ensure sufficient time for selection, design, and implementation of mitigation measures before a possible debris-flow triggering storm occurs, which can be as little as a few weeks after a fire. Prior work to by others has aimed to address these issues; however, previously developed post-fire risk-management frameworks do not provide specific guidance regarding estimation of debris-flow occurrence or runout, and procedures developed to specifically evaluate post-fire debris-flow hazard often provide little to no information about expected runout and lack specific guidance for quantitative evaluation of risk. The work described in this thesis makes several new contributions to the existing frameworks. Specifically, the method described in this thesis expands on previously existing post-fire debris-flow hazard models by adding quantitative assessment of damage and economic risk, and adds to previously existing post-fire risk management frameworks by providing methods for evaluating debris-flow specific damages. The method described in this thesis can be used to model damage and economic risk for individual basins and entire burned areas in a period of days to weeks following a fire; these modeled damages and estimated risk values then provide inputs to a new model that can be used to select cost-optimized debris-flow management strategies, and can help emergency managers allocate funds to reduce economic risk from post-fire debris flows. Use of the method to evaluate modeled damages and estimated economic risks at a case-study site identified several key areas of research that can be studied in the future to reduce uncertainty in the modeled results and improve future versions of the method.