A farmhouse that burned or was torn down a century ago rarely leaves anything visible from ground level. It almost always leaves a footprint in the terrain.
In short
Homesteads are good LIDAR targets precisely because they're small and low-relief — a stone or poured foundation, a root cellar, a filled-in well, and the lane that connected them to the road all disturb the ground by a few centimeters to a couple of feet, which is exactly the scale LIDAR's bare-earth elevation model is sensitive to. The harder part usually isn't spotting a candidate once you're looking in the right place — it's narrowing down where to look in the first place.
Thin rectangular outline, slightly raised or with a raised rubble edge — clearest in Local Relief or the composite view.
Root cellar / cold cellar
Small circular or oval depression, often just a few meters across — reads as a tight blue spot in Local Relief.
Filled-in well
Small circular anomaly — a shallow depression if it settled over time, or a slight mound if it was backfilled above grade.
Farm lane / cart path
Long, gently sunken linear trace, often curving to follow easiest grade rather than running perfectly straight.
Terracing / field walls
Long, low, straight-ish raised lines — easy to confuse with a foundation edge in isolation; check whether it's part of a longer running line first.
See the terrain-reading guide for how each of these actually renders across hillshade, slope, and Local Relief — a foundation and a field wall can look similar in one style and completely different in another.
2. Narrowing down where to look first
A LIDAR scan only covers the search radius you point it at — running blind searches over open country is a slow way to find a needle in a very large haystack. A bit of desk research first turns "somewhere in this county" into a specific few hundred meters worth actually scanning:
Old county atlases and plat maps — many US counties published illustrated atlases in the late 1800s showing individual landowner parcels and structure locations by name. Local libraries, county historical societies, and state archives are the usual homes for these.
Historical USGS topographic maps — USGS has published topo quadrangles since the late 1800s, and older editions often mark individual structures (small black squares) that don't appear on the current sheet at all, because the structure hasn't existed for decades.
Historical aerial photography — USDA and USGS aerial photo archives go back to the 1930s–40s in a lot of areas; a structure that was demolished after that point may still show up as a roofline in an old photo even though nothing is standing today.
Local historical & genealogical societies — often the fastest route to "we know exactly where the old so-and-so place was," especially for anything with a name attached to it.
Why this matters beyond convenience
Corroborating a LIDAR candidate against an independent historical source is also the difference between "probably a foundation" and "confirmed to have had a structure here as of a documented date" — useful context to have before you invest time getting permission and making the trip.
3. A search workflow
Narrow a candidate area using the research sources above — even a rough "somewhere along this old road" is enough of a starting point.
Run a Lidarman search centered on that area with a generous radius, then tighten the radius once you see where the activity is.
Step through hillshade → slope → Local Relief per the terrain-reading guide to sort real candidates from noise.
Use the present/past flicker comparator to see how a candidate's location sits relative to today's satellite imagery — tree cover, current land use, and access.
Not every rectangle used to be a house. Modern equipment pads, hay bale storage areas, agricultural terracing, old tennis or basketball courts, and even certain drainage features can produce a similar geometric signature to a foundation. This is exactly why Lidarman's auto-detected candidates are ranked by confidence rather than presented as verdicts — treat a flagged candidate as a place worth a closer look, not a place that's confirmed to have had a structure on it. Independent corroboration (an old map, a known family name tied to the parcel, a filled well visible in a different style) is what turns a candidate into something worth the trip.
More guides
Finding Ghost Towns With LIDAR — What a vanished town leaves in the ground, where the list of 16,525 of them comes from, and why the dot is only a starting point.
Detecting Near Civil War Battlefields — The protected core is off limits — the study area around it often isn't. How to tell the two apart before you knock.
Finding Old Mining Camps — 44,677 abandoned metal mines, what the camp left behind as opposed to the workings, and why the two are a short walk apart.
Historical Maps for Metal Detecting — Where to get old maps free, what the symbols mean, and how to line a century-old sheet up with the ground you are standing on.
Detecting Along the Westward Trails — What survives of the Oregon, California, Pony Express and Santa Fe routes, and how ruts and swales read in terrain data.