Abstract

To learn from past earthquake events, a critical task is validating observations with data. After a significant seismic event like the 2023 Kahramanmaras Türkiye earthquake series, various organizations perform reconnaissance efforts to collect building response data. Some reconnaissance missions focus on visiting fewer buildings and recording detailed information (high-resolution, HR), while others aim to visit as many buildings as possible recording basic information (low-resolution, LR). Both dataset types are vital for learning, and the next step is analyzing and visualizing the data to validate in-field observations. This paper has three purposes: 1) examine a HR dataset by Degenkolb with 138 buildings, 2) summarize a LR dataset by Thornton Tomasetti with thousands of images, and 3) compare a Degenkolb HR reconnaissance dataset with a Degenkolb LR dataset of image metadata.

The Degenkolb dataset included 35 collapsed and 87 damaged buildings, observational findings reported no apparent predominant pattern in the cause of this damage. To better understand relationships between damage and site, seismicity, and building parameters this paper presents geospatial maps and quantitative graphs of the reconnaissance data. Specifically, the authors examine correlations of building damage with building location and building properties (e.g., number of stories, occupancy, material, and lateral system). The Thornton Tomasetti images each capture building damage and a record of geospatial location for buildings. This paper summarizes the qualitative relationship between these two parameters. Improving reconnaissance data analysis is also important in learning from earthquakes, and a tool that automates the workflows used in this paper is in development.

Disciplines

Architectural Engineering

Number of Pages

11

Download

Share

COinS
 

URL: https://digitalcommons.calpoly.edu/aen_fac/155