Field notes: the week a hundred-parcel region came back with four survivors

We keep a running tally of regions we've screened for clients this year, and one county kept coming up in conversation because the before-and-after was so stark. A hundred parcels went in, flagged off tax-roll acreage and a rough county boundary. Four came out the other side still viable. That ratio sounds brutal until you walk through what took the other ninety-six down.

Slope did the first cull, and it wasn't close. The county's eastern third rolls off a ridgeline that doesn't show up as "hills" on a county GIS viewer zoomed to the township level, but it reads immediately once you're looking at a slope raster built from decent elevation data. Thirty-some parcels sat on grades past what a single-axis tracker array tolerates without civil work nobody budgeted for. A few were borderline, benched enough in spots that someone will probably walk them eventually. Most weren't close enough to argue about.

Land cover took the next bite, and this is the one that burns the most hours when you're doing it by hand. A chunk of the remaining acreage was pivot-irrigated row crop, the kind of ground that looks clean on a plat map and turns out to be center-pivot corn with the circle tracks still visible from altitude. Whether that's disqualifying depends on the jurisdiction, but combined with lease pricing that assumes active ag use, it moved those parcels to the bottom of the stack fast. Another dozen turned out to be mixed hardwood that hadn't been logged in a generation, which is its own conversation with the landowner and usually not one that ends in a signed option within a quarter.

Then the water did what water does. A floodplain overlay knocked out a strip along the creek bottom that had looked, on the parcel map alone, like some of the best ground in the county. Flat, cleared, good road frontage. Also sitting in a mapped hundred-year zone with a drainage pattern that made the adjacent parcels iffy too. That's the kind of thing a GIS analyst catches eventually, cross-referencing FEMA layers against a shapefile one parcel at a time, but it's exactly the step that gets rushed when a team is trying to clear a list before a landowner outreach deadline.

What was left after slope, cover, and water was shape and size, and this is where the last handful of parcels either held together or didn't. Two irregular slivers that technically met the acreage threshold on paper turned out to be unbuildable once you accounted for setbacks and the actual usable footprint inside an odd property line. One parcel got split by an aspect problem, a north-facing shoulder that would never pencil for a fixed-tilt layout even though the rest of the tract was fine.

Four parcels made it through clean. Not the four anyone on the team would have guessed walking in, which is usually how it goes. The ones that looked best on the tax map were the floodplain strip and the pivot ground, both gone by the second filter.

The part worth sitting with is how early all of this was knowable. None of it required a truck, a landowner call, or a survey crew. It's the kind of cross-referencing a solar site screening workflow is built to run against a whole region before field reconnaissance starts, so the ninety-six get set aside and the four get a real look. The county didn't change between the tax map and the slope raster. The information was just sitting in different layers nobody had stacked yet.

If your next region looks anything like that one on paper, it's worth finding out before the team's in trucks which parcels are worth the drive.