Earthquake risk: A geographic information system-based model and sensitivity analysis, Salt Lake County, Utah

This thesis presents research results on a probabilistic assessment of expected losses to residential and commercial buildings due to seismically induced ground shaking within developable Salt Lake County, Utah. Additionally, results on expected casualties to residential and employee populations as a result of expected level of property loss are presented. Salt Lake County is subject to the risk of earthquake ground-shaking from events likely on segments of 21 nearby faults, including major segments of the Wasatch Fault. The geographical information system (GIS)-based analysis includes an algebraic combination of (1) a mapping of probabilistic ground-shaking intensities, (2) an inventory of buildings by location, value, and frame type, (3) seismic damage functions that define structural performance as a function of ground-shaking intensity, (4) residential and employee population data, and (5) seismic casualty functions that define casualty risk as a function of building damage. The numerous effects of error and its propagation within the GIS and mathematical models are discussed. Additionally, a methodology to randomly perturbate ground shaking boundaries was developed. This technique was used in a sensitivity analysis that modeled the effect of spatial uncertainty in ground-shaking zonation through Monte Carlo simulation. This test suggests that the property damage model can accommodate potential misclassification in ground-shaking zonation and still provide a reliable loss estimate. Seismic mitigation options, principally the effect of a seismic retrofit policy for commercial unreinforced masonry structures, are also evaluated.