At BuildingFootprintUSA™ we enjoy how our highly accurate spatial data can solve problems across industries. Lately we have spoken with a number of businesses involved in the flood insurance and certification industry, and we wanted to understand for ourselves how building footprints could provide higher accuracy. What we learned was eye opening – if you are not making use of building footprint-based location intelligence to answer flood-related questions you are making a lot of mistakes.
We wanted to see how building footprint data can enhance the accuracy of the flood determination process. For those who aren’t aware, there is an entire industry that decides which houses or other buildings require flood insurance and which don’t. You take in an address, get the most accurate location for the building and do a spatial comparison against government defined flood plains.
State of the art right now for geocoding is to use address points – point locations that are the best approximation for a house or building. How would address point-based decisions compare against addressed building footprint-based decisions? We chose the Raleigh, NC area for study, very topical given recent flooding.
Knowing that most commercial address point providers use local sources of address point data we obtained the Wake County, North Carolina address point dataset which is likely about 99% the same as the address points that commercial data providers use. For every address we identified the address point location (x/y) and the building footprint location (x/y) for that address, and compared the resulting locations against FEMA flood data that defines the flood plains.
Let’s give a visual. Imagine a house on Brookfield Dr. Going out into the backyard you would enter a shallow ravine where in flood conditions you will find moving water. The ravine is considered a flood prone area by FEMA and coded accordingly. But the house itself is on high ground.
Flood Plain Analysis
Now it’s time to determine whether the house needs flood insurance. If you use an address point to locate the house, the address point is imprecise and is located in the ravine which happens to be Flood Zone A, one of the highest risk flood zones. If you do your flood determination based on a high accuracy building footprint your analysis will correctly show that the home is in Flood Zone X, low risk.
Does this matter? We were stunned that if you look at the universe of addresses for Wake County, North Carolina and rely on address point-based analysis over 500 times you will misidentify a property as needing flood insurance when it actually does not. The number might be as high as 700 times but we’re not flood plain experts so we’ll be conservative.
Furthermore, we looked at the inverse – using address point-based analysis over 50 times you will tell people that they would not need flood insurance when in actuality they do – and building footprint-based analysis can correctly determine this. Imagine that conversation – your customer gets flooded out, you have told them that they don’t need insurance, and they do their due diligence to find out you were wrong.
Maybe flood cert companies expect this and have an army of people to unwind bad automated decisions. Or maybe they don’t as these bad automated decisions get enacted. Higher accuracy will obviate either of these scenarios.
As we said at the beginning, we enjoy finding new ways that building footprint data can solve new geospatial problems, or existing problems at a more accurate level. Thanks for reading and feel free to drop us a note if you agree or disagree with the findings.