Data files have just been uploaded to the AGORA website that provide estimated quantities of building area, building value, contents value, and indoor occupants at 2 PM, 2 AM, and 5 PM, for each combination of US census tract, occupancy type, structure type, and design level. There are also files containing county- and state-aggregate values. The database application and a methodology document are also available. Kishor Jaiswal and Pete Ng of the USGS helped create the data files. Thanks guys.

Download the files (free registration required)...

 

Extracting HAZUS’s Seismic Building Inventory to a Single Table

Draft manuscript, blank database, and text-file results for many US states (soon all)

A draft manuscript entitled  Cracking an Open Safe: Extracting HAZUS’s Building Inventory to a Single Table ,  available here, explains how one can "denormalize" the HAZUS building inventory, i.e., merge 15 linked tables from 2 mdb files on the HAZUS data DVDs into a single table that lists, for each combination of  census tract, occupancy class, earthquake structure type, and seismic design level: the total estimated square footage (1000s of sf), building value ($1000s in 2003), content value ($1000s in 2003), number of occupants at 2 PM, number of occupants at 2 AM, and number of occupants at 5 PM. Only combinations with nonzero values are tabulated.

The process is automated in a beta-version Microsoft Access database, also available here, containing a number of supporting tables, links to the 15 HAZUS-MH tables (the user needs to supply the proper HAZUS-MH DVD and update the links in a process detailed in the manuscript), plus 59 scripted query language (SQL) queries and a macro to perform them all in sequence. It helps to turn off all confirmations before running the macro. 

The author and colleagues at the US Geological Survey have carried out the calculations for a number of states and exported the results to comma-and-quote-delimited text files that are bundled into zip files. Each zip file contains three data files for a state: one each for tract, county, or state-level aggregation. Each file is has column headers in the 1st row for ease of reading.