Re: How accurate is the elevation on US topo?

Another consideration about vertical accuracy of topo maps comes into
play when you say that a certain contour line is "near" your house, or
some other object on the map. Not only is there some uncertainty
(error) about elevation measureable by photogrammetry, but there's also
uncertainty about the horizontal position of spot elevations and
contour lines plotted on the topos. Horizontal position corresponds to
the old (1940s) NMAS (National Map Accuracy Standard), which offers a
"statistical guarantee" that ~10% of the points will be plotted more
than ~40 feet from where they should be, so the spot elevation or
contour may be displaced from the true position.

Tests usually show that the USGS has been conservative about accuracy
of topos and derived products such as DLGs and DRGs, and they usually
test better than NMAS promises. But remember that there are about
55,000 1:24,000 quad sheets, and until GPS came along there was no
practical (affordable) way to measure the accuracy of any given quad.
Hence, the old NMAS really was more of a production specification/goal
than a specific measure of accuracy like its successor the NSSDA.

.