Skip to main content
U.S. flag

An official website of the United States government

Flood Loss Risk and Its Drivers: Evidence from Massachusetts Residential Properties

Presentation Date
Friday, December 11, 2020 at 4:00am



Standard indicators of physical exposure from floods capture only a portion of the probability distribution of potential flood events, in terms of return periods as well as populations and property at risk of inundation and in follow-on assessments are combined with data on economic activity at relatively coarse geographic scales to derive losses, with the potential for aggregation bias. We conduct a comprehensive flood risk assessment at the individual property level on the distribution of inundation exposure due to pluvial and fluvial floods on different return periods over a broad geographic domain from a validated 30 m resolution model, depth-damage functions and property-level information on building characteristics and market prices from Zillow’s ZTRAX database for a sample of over 1 million single family homes in Massachusetts. Of these, 236,897 (26,219) homes are at risk of pluvial (fluvial) flooding with $65 M ($21 M) in expected yearly losses which make up 2.4% (8.9%) of their overall structure value discounted yearly at 3% over the course of a typical mortgage. Overall expected damages are driven disproportionately by homes exposed to extreme (>= 100 cm) flood depths, associated mostly with fluvial floods, but a substantial portion of overall losses is contributed by the large number of high value properties surrounding Boston exposed to nuisance flooding (3-10 cm) in basements from rainfall. Partitioning homes by nuisance, moderate (10-100 cm) and extreme flood depths effectively draws out distinct hazard and asset distributions informing targeted risk reduction strategies, as opposed to return period or event-based planning which dominates current government practice. We find that retrofits and buyouts based on the risk distribution can result in over 55% (30%) in cost effective loss reduction interventions on 16% (8%) of pluvial (fluvial) at-risk homes. In contrast, ex-post disaster buyouts for a historic extreme event aren’t cost effective and ex-post disaster retrofits are cost effective but implemented only after severe losses have been realized. Our representation of objective flood risk over a large spatial scale elucidates the complex relationships under the present distribution of hazard and vulnerability so that we may inform adaptation and development in the face of climate change.


Funding Program Area(s)