Accuracy of FEMA-Hazus Single-Family Residential Damage Exposure Data in Houston: Implications for Using or Correcting the Hazus General Building Stock

Single-family residential building damage exposure data from the FEMA-Hazus general building stock data set (HGBS) are compared with corresponding Harris County (Houston, Texas) assessor data. Countywide, the HGBS generates similar house size estimates, but exaggerates replacement new costs by 86% and depreciated costs by 165%. At the census block level of analysis, where most FEMA-Hazus disaster mitigation modeling takes place, both size and replacement cost data had very high and irregular incidences of inaccuracy across individual blocks with no discernable error patterns. These heterogeneous errors, combined with the fact Houston-based HGBS-assessor inaccuracies are markedly dissimilar from prior observations of HGBS inaccuracy in three upper Midwest study locations, demonstrates the difficulties of using simple correction measures for HGBS data.

Automated summary

Comparing FEMA-Hazus general building stock data with Harris County assessor data revealed significant discrepancies in replacement and depreciated costs, while both datasets showed similar house size estimates. Inaccuracies at the census block level were highly irregular with no discernable error patterns, indicating the challenges of using simple correction measures for HGBS data. The discrepancies in Houston-based HGBS-assessor inaccuracies compared to other regions further highlight the complexities of correcting HGBS data.