DHS, FIMA, FEMA’s Response Geospatial Office, Oak Ridge National Laboratory, and the U.S. Geological Survey collaborated to build and maintain the nation’s first comprehensive inventory of all structures larger than 450 square feet for use in Flood Insurance Mitigation, Emergency Preparedness and Response.

Abstract

DHS, FIMA, FEMA’s Response Geospatial Office, Oak Ridge National Laboratory, and the U.S. Geological Survey collaborated to build and maintain the nation’s first comprehensive inventory of all structures larger than 450 square feet for use in Flood Insurance Mitigation, Emergency Preparedness and Response.

FEMA and our federal partners identified a need to create a building outline (the polygon representation of the structure) and an updated address database for the nation that could help ensure that critical infrastructure and residential buildings are accounted for in the disaster response and recovery decision-making processes.

To respond effectively, we need to understand population and the built environment—where people live, work and the critical infrastructure they rely on.

Many residential structures have an address where the occupants receive mail, but the address is not associated with the structure’s physical location. This can happen when people use a Post Office box, or when the mailbox is located in a central location, which is often the case in mobile home parks, apartment buildings, and rural communities.

The lack of structure information can limit our ability to adequately characterize a disaster’s potential impacts since parts of the community are missed by the predictive models. Furthermore, the homes that lack a footprint with an associated address are occupied by the most vulnerable in the community, delaying aid to those who are most in need. How? Since the location and address are not coupled, FEMA and our response partners can struggle to determine which damaged home is associated with the address presented by the survivor seeking assistance.

Methodology

To create the building outline inventory, FEMA, in conjunction with DHS Science and Technology, partnered with the Oak Ridge National Laboratory (ORNL) to extract the outlines via commercially available satellite imagery. We then worked to determine the building’s usage or occupancy type (e.g., residential, commercial, industrial) which is noted as an attribute for each structure.

In the past, geographers have relied on satellite imagery as a high-coverage and low-cost data source to create building-location inventories; however, identifying individual buildings is labor-intensive and had been difficult to automate due to large variations of building appearances. Our processes included some new machine learning techniques and a collection method to obtain data from multiple sources, including from local governments who agreed to share it, and open data from the National Geospatial-Intelligence Agency (NGA).