Characterizing uncertainty of general building stock exposure data

Effective disaster risk management (DRM) and disaster risk reduction (DRR) require modeling potential and post-event impacts using building exposure data. The data used to develop building exposure databases will influence the accuracy of risk assessments and the appropriateness of subsequent decisions. This article proposes a framework for classifying approaches of developing building exposure databases into levels. To examine the uncertainty introduced through using various approaches to exposure development, a probabilistic seismic risk assessment was run with the exposure data corresponding to each proposed level using the County of Los Angeles as the study area. A factor of similar to 2.5 was observed in the final loss estimates. The variance was less dependent on the spatial scale of data than on key values, most notably estimates of building size and replacement cost.