Desktop study vs field reconnaissance: when to send a crew to ground-truth an alignment

Every corridor project hits the same fork early: you've got three or four candidate alignments sketched on a base map, a survey budget that covers maybe one and a half of them properly, and a client who wants to know which one to chase first. The desktop study is supposed to answer that. Too often it doesn't, because it gets treated as a formality instead of a filter.

What a desktop study is for

A desktop study answers a different question than field recon: which alignments are worth the cost of putting a crew on the ground at all.

Done right, a desktop pass tells you where the slope kills a grade, where the land cover flags wetland or heavy timber clearing, and where a built-up footprint forces a realignment before anyone drives out. That's three separate datasets, usually sitting in three separate files or three separate consultants' desks, and most teams still overlay them by eye on a plan sheet. It works, but it's slow, and it's easy to miss where two constraints stack on the same half-mile.

The output you want from this stage is a ranking, not a go/no-go on the whole project. Alignment B clears the slope and land-cover screen clean but crosses more built-up area near the interchange. Alignment C is the opposite. That ranking is what justifies spending survey dollars on one corridor instead of three.

Signals it's time to send a crew

A desktop study flags slope, land cover, and built-up constraints from imagery, but it can't confirm a culvert's actually there, or catch that the "open field" on the land-cover layer is a horse paddock with a fence line nobody digitized. Field reconnaissance closes that gap, but it costs real money and real time, so the trigger to deploy a crew should be specific, not a calendar date.

Send a crew when:

• The desktop screen shows a constraint sitting right at the edge of a threshold: slope that's borderline buildable, a built-up mask that lands right on the edge of a structure. That edge is exactly where satellite-derived masks lose precision and a walked line earns its cost.
• Two candidate alignments score close enough on the desktop pass that the tiebreaker is something a remote dataset can't see: drainage pattern on the ground, soil condition, an access road that didn't show up in the imagery.
• A client or agency needs a defensible record that someone physically checked the preferred alignment before design work starts. Desktop screening narrows the field; it rarely satisfies that requirement on its own.

If none of those are true yet, sending a crew is spending survey budget to confirm what the desktop pass already told you.

The order most teams get backwards

The common failure mode is running field recon too early, on too many candidates, because the desktop screen wasn't trusted enough to cut the list first. A crew spends a week walking an alignment that a slope layer alone would have eliminated in an afternoon.

Sequencing fixes this. Screen every candidate against land cover, slope, and built-up constraints at project start, before anyone books a crew. Rank the candidates on that combined picture. Then send field recon at the one or two alignments that survive the screen, where the remaining questions are things only a walked line can answer.

That's the gap Route Selection Map is built for: a constraint-scored map of your candidate alignments, built from land cover, slope, and built-up masks at project start, so the ranking exists before you've committed anyone to a truck. It won't replace your crew. It should decide which alignment gets them first.

If you're staring at three corridor sketches and a survey budget that only covers one, that's the question worth answering before the trucks leave the yard.