Zoomed to the edge: is asphalt really continuing under the green?

The strongest case. Pavement-height returns continue ~0.9 m outboard of the detected edge, within 0–3 cm of the crossfall-extended plane, with a continuous surface line and no step. The scanner sees ground through the canopy (264 blue canopy+ground columns), so nothing is hidden — but the continuing surface is ~3.5× sparser and ~7× greener than reference asphalt with a small low bush (0.25–0.55 m tall) overhead. Reading: the outer road margin here is thin, vegetation-tinted, and partially overgrown; the detector's edge retreated from it.

The original "money shot", demoted to a lip by measurement. A dense low canopy (bulk 0.1–0.35 m above the plane, top ~0.6 m) fills the notch; asphalt-level returns continue only ~0.2 m outboard of the detected edge — the thin blue strip on the class map — after which the notch interior is canopy-only or empty out to the verge row at −0.85 m (green low growth with a separate grey post/rail-like structure above it, and a delineator post at −0.9 m). Consistent with vegetation occlusion causing the notch, but the pavement-under-canopy evidence is 0.2 m of lip, not a continuing apron.

Not a rising grass bank and not clean occlusion. Ground returns run continuous to 2 m outboard with no shadow; they sit 11–21 cm above the fitted plane and step up ~10 cm onto a shoulder berm at 0.5–0.8 m before dropping back. Colour turns brown-olive within the first 0.15 m and strongly grass-like by 0.2–0.35 m. Canopy overhangs the pavement only on the flanking metres — at the notch centre it stops ~0.17 m outboard of the edge. Reading: grass/soil encroachment on a raised shoulder; the notch tracks the vegetated berm, not hidden asphalt.

Returns continue 0.74–0.78 m outboard but from near-plane up to ~15–18 cm above it, and their colour is unambiguous: excess-green median 0.11 vs 0.004 on reference asphalt (~30×). There is no second return layer at plane level beneath the mat. True canopy-over-ground columns are only 3.4% of the tile, a thin stripe at −0.6 m, entirely outboard of the notch. Reading: a low vegetation mat growing over the edge zone; pavement continuing underneath is not demonstrated — the judges also flagged that the class-tinted views cannot prove what material is under the mat.

My initial "pavement continues 0.5 m" was refuted by measurement: the outboard returns climb steadily off the plane (+0.06 m at the edge to +0.12 m at 0.45 m out, ~25% slope, no flat shelf), are warm-olive against the road's cool grey, and are ~3× sparser than reference asphalt. The verge starts rising at the detected edge. The elevated cluster at −0.75 m that I read as a delineator post is a road-parallel linear structure 0.36–0.74 m above plane — possibly a rail or fence line in the hedge; identity uncertain from these views.

The judges split here (one upheld my caption, one wanted it stronger); the measured version: ~0.3 m of ground continues outboard, climbing from plane level at the edge to 0.15–0.20 m above it — a bank toe, not a pavement-level mat — while the pavement itself falls away from the edge. No canopy+ground column falls inside the bite; the canopy band starts ~0.6 m outboard. Clean asphalt continuation is not demonstrated.

Zero pavement-level returns in the bite. The outboard strip is a green raised band 0.25–0.6 m above the fitted plane (note: the plane fit sits ~0.2 m below the actual pavement here, so read heights against the adjacent road surface), with a separate foliage layer ~1.2–1.4 m up hanging over the strip — so "some overhead foliage" exists, but there is no observed surface beneath any of it. The detected edge floats inside a vegetated area; whether asphalt physically ends here or is fully buried, the notch is not recoverable from the point evidence.