Enhanced Methods for Building Inventory Development for Regional Risk Analysis

Graduate Researcher: Mia Lochhead
Advisors: Professor Greg Deierlein, PhD and Adam Zscarnóczay, PhD

While it is well accepted that natural hazard risks result from the intersection of hazard, exposure, and vulnerability, progress in characterizing these factors has been uneven. In the context of earthquake engineering, significant advances have been made in modeling the hazard and vulnerability components, but the development of exposure models (building inventories), has received comparatively limited attention in academic research.

There are many building inventory data sources available in the United States, including both nationally available datasets and locally specific sources such as tax parcel or address data; however, these datasets often contain limitations or gaps. Limitations of individual datasets can be addressed by synthesizing multiple data sources, though this introduces a unique set of challenges. This work aims to 1) develop a systematic workflow for creating footprint-level building inventories through the synthesis of multiple data sources and 2) test the hypothesis that the data sources and methods used to develop regional inventories significantly affect both the inventory and the resulting risk estimation.

The proposed workflow is implemented to develop and compare several inventories representing the case study city of Hayward, California using a M7 scenario earthquake on the Hayward Fault. The findings indicate that different inventory development methods and data sources can lead to significantly different quantifications of seismic risk. Furthermore, these differences cluster both geospatially and by building feature, potentially introducing inaccuracies and biases in the quantification of seismic risk. Synthesizing inventory data from multiple sources is proposed as an effective way to identify potential biases and resolve disagreements in the data to more faithfully capture the built environment.