UC Santa Barbara

Landslides are natural events that occur in many parts of the world in both developed and developing countries. They can be triggered by rainfall, earthquakes, volcanos, and a combination of post-fire and rainfall. However, the impact they cause in a community can be very different, based on how the community can anticipate and respond. They can cause material losses and some minor property damage, but also, they can become natural disasters, resulting in significant material and human losses.One way to avoid and mitigate landslide disasters is by implementing Disaster Risk Reduction measures, such as risk mapping or developing an evacuation plan. These measures help municipalities plan before an event occurs, thereby protecting their communities and being proactive. In this doctoral dissertation, the focus is on two of these Disaster Risk Reduction measures: risk and vulnerability assessments. In the first part, the study is conducted in the Metropolitan Area of Sao Paulo, Brazil. In these urban areas, shallow landslides are frequent in low-income neighborhoods on hillslopes, especially during summer months, when intense rainfall frequently occurs. Chapters 2 and 3 propose two distinct methodologies to quantify an inventory-based shallow landslide risk mapping. The quantification is intended to reduce bias and standardize risk mapping methodology. In chapter 2, experts’ knowledge and the Analytical Hierarchical Process (AHP) method is applied to compute weights for variables, and application is developed and used to calculate the risk level automatically. In this study, variables that vii illustrate instability in the terrain were the ones with the highest contribution for a higher risk. In chapter 3, a large dataset from a previous mapping of Sao Paulo city and Ordinal Logistic Regression (OLR) was used to select essential variables and to compute their weights. The equation calculated can be used in the developed application to compute the risk level automatically. In chapter 4, a spatial analysis of the slope stability using SHALSTAB and Factor of Safety and saturated hydraulic conductivity was computed in two hillslopes of São Bernardo do Campo, in the Metropolitan area of Sao Paulo. In one of these sites, the hillslope is disturbed by human activities (previous homes, cut and filling, landfill), and in the second site, the hillslope is undisturbed. The saturated hydraulic conductivity and the slope stability analysis illustrate that the disturbed site has less homogeneous soil than does the undisturbed site. Additionally, based on the soil characteristics and results from SHALSTAB and Factor of Safety analysis, the undisturbed hillslope is more stable than the disturbed hillslope. The second part of the study was conducted in Montecito, California. The study was proposed after the 2018 Montecito debris flows that killed 23 people and damaged more than 400 homes. In this study, I used a parallel mixed-methods approach and a temporal-spatial analysis to understand the main factors that made the community vulnerable to debris-flows and to determine who were the most vulnerable people. The study concludes that the most vulnerable people were informal workers (gardens, housekeepers, caretakers, nannies), renters, and residents of areas that were in the voluntary evacuation zone. Lack of education about debris-flows and previous debris-flows in the region by the community increased the vulnerability of the entire community. Moreover, the institutional vulnerability was high for the entire community, but county measures in the aftermath contributed to the community resilience.