by John Kirk » Mon Sep 16, 2024 9:03 am
I updated this on 9/12, dGPS in the field is considerably more accurate. We didn't miss the summit boulder as some speculated (elevations match), rather the vegetation at the contender points has been misclassified in the point clouds as ground. Andrew Kirmse has been working with USGS to better understand and potentially adapt ground point classification algorithms, which vary on the continuum of conservative vs. aggressive classification.
Long story short, dGPS when operated properly is going to supercede LiDAR given its superior accuracy and capability of ruling out non-ground objects.
One minor point on NGVD29 vs. NAVD88 conversions... whether archaic paper quads use NGVD29 or not shouldn't matter. Datum consistency with unanalyzed peaks is not the goal on LoJ as LiDAR is not intended to be used as an exception handler. The LoJ goal is comprehensive restatement with the most up-to-date datum and accuracy.