Satellite vs. aerial imagery in a commercial underwriting file
A commercial property submission lands on your desk with a COPE sheet, maybe a few photos the broker pulled off the applicant's website, and a construction year that may or may not match what's on the roof. Before you bind, you need a current picture of the building, built from an image of the roof itself. That's where the satellite-versus-aerial question matters, and it's less about which looks better in a slide deck and more about what each one gets you into the file before the effective date.
What GSD buys you
Ground sample distance, GSD, is the size of what one pixel covers on the ground. A VHR (very high resolution) satellite image at roughly 0.3 meters per pixel can resolve roof seams, HVAC units, parapet lines and separate roof sections on a multi-building commercial site. That's enough detail to count storeys from shadow length, separate a flat built-up roof from an adjacent metal one, and flag a tarp or visible roof damage before the policy binds instead of at the first claim.
Traditional aerial flyover imagery can run finer than that, and for a single scheduled flight over a specific address it often does. But aerial coverage is captured when someone orders a flight over that parcel, which means it exists where an insurer, a county assessor, or a mapping vendor has already paid for a flyover. VHR satellite coverage, by contrast, gets refreshed on a standing schedule across wide areas, so an address without a recent local flyover still has a current image to pull from.
Annual satellite pass vs. a flyover order
This is really the operational difference, not a quality argument. An annual satellite pass means the imagery for a given address updates on a fixed yearly cadence, regardless of whether anyone requested a look at that specific building. You're not waiting on a flyover contract to cover your renewal book, and you're not stuck with a five-year-old aerial shot because nobody re-flew that county.
A flyover, ordered on demand, gives you a single point-in-time capture for a specific site. It works well when you need one address inspected closely, say a contested claim or a high-value schedule. It doesn't work as a standing process for a submission pipeline where new addresses show up every week and you need something current on each one without placing a separate order every time.
For a book of commercial risk, the annual-cadence approach fits how underwriting actually runs: new submissions arrive continuously, renewals come up on their own schedule, and nobody wants to open a flyover ticket for every address before quoting it.
What goes in the file
What belongs in a property characteristics record pulled from VHR RGB imagery is roof geometry, roof material, storey count and visible site features: the things an applicant's self-reported COPE data gets wrong often enough to matter. A flat roof reported as pitched. A three-storey annex left off the schedule. A parking structure or outbuilding nobody mentioned. None of that requires a site visit or a drone order. It requires a clear, current, high-resolution image of the roof and a structured way to read it.
That's the gap Property Characteristics fills: a structured record per address, built from VHR satellite or aerial RGB imagery on an annual cadence, so the underwriting file reflects what the roof looks like rather than what the application said it looks like. The resolution and cadence trade-off between satellite and aerial comes down to coverage. It's a question of which one covers your submission volume without a special order for every address, and for most commercial books, that's the annual satellite pass.
If your submission file is still built from self-reported roof details, see what a structured, image-based property characteristics record looks like before your next bind.