[Proj] NAD83(Original) Transformation with Heights

[Rob Skelly]

Hello List,

I'm trying to convert some LiDAR data from NAD83(Original), with
orthometric heights based on CGVD28, to WGS84 with ellipsoidal heights, and
to NAD83(CSRS) with orthometric heights based on CGVD2013.

I've been verifying my output using NRCAN's online GPS-H tool (
http://webapp.geod.nrcan.gc.ca/geod/tools-outils/gpsh.php), and not getting
good results.

To begin with, I use las2las to convert my data to WGS84, using parameters:

--a_srs '+proj=utm +zone=12N +ellps=GRS80 +geoidgrids=HT2_0.gtx'
--t_srs EPSG:4326

Then I go from WGS84 to NAD83(CSRS) with the following:

--a_srs EPSG:4326
--t_srs +proj=utm +zone=12N +geoidgrid=CGG2013n83.gtx +ellps=GRS80
+datum=NAD83

Now, I don't actually have a decent way of verifying my output, other than
with GPS-H, and the disagreement seems significant. For example, the
original elevation at 494618.88mE, 6306815.69mN is 469.95m, and I get an
ellipsoidal height at -111.088356°W, 56.90501°N of 444.52m. GPS-H claims
that the orthometric height at that position (for HT2_0, the closest
approximation to CGVD28) should be 470.004m, a difference of 54cm.

I realize there are a lot of points where this process could be going
wrong, or that it's not going wrong at all. I wonder if anyone can give me
any insight or fill in any obvious holes in my reasoning. If I've left out
any pertinent information, let me know.

Note that I could just use GPS-H to process the points, but I need to
automate the conversion of a lot of data, and establish a reusable process
to do so. And, I'd like to use this opportunity to learn about proj!

Thanks,
Rob
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.maptools.org/pipermail/proj/attachments/20140213/7de7ff70/attachment.htm