[GRASSLIST:1871] Re: Importing USGS SDTS DLG maps from differing datums? I'm confused. [very long]

[Roger S. Miller]

On Thu, 24 May 2001, Tom Russo wrote:

> Simple statement of the question: How can I import two DLGs that were created
> using different geodetic data (NAD27 and NAD83) into the same grass location?
> v.proj seems to be the way to do it, but I tried and it didn't work as I
> expected it to.

Tom,

I think you found an actual problem.  I'm nearby, so I took a local
interest in your problem.

I reproduced the problem that you had, then tried an alternative
procedure.  I converted both sets of road data from UTM coordinates to the
corresponding lat-long coordinates, then tried to convert the NAD83 roads
from lat-long NAD83 to lat-long NAD27 so the two data sets could be
compared.

This procedure parallels the process that you have to go through to
convert the data using the National Coast and Geodetic Survey's (NCGS)
software.  The conversion from lat-long NAD83 to lat-long NAD27 failed.
v.proj produced:

Creating dig file ...
Error in pj_do_proj

and the output file was empty.

I believe you were correct that the boundaries of the two maps should
*not* have overlain after the conversion.  So I did a complete conversion
manually to check GRASS' results.

I dumped the NAD83 roads to an ascii vector file and wrote a small program
that converted the data to a format that could be read by the NCGS
programs.  I ran that data through th NCGS' "utms" program to convert from
UTMs to lat-long on the GRS80 spheroid, then used the NCGS NDCON210
program to convert the latitudes and longitudes from NAD83 to NAD27.  I
again used "utms" to convert the NAD27 lat-long records back to UTM
coordinates on the Clark '66 spheroid.  Finally, I wrote a small program
that merged the converted coordinates back into the original ascii vector
file.

I've been using this procedure to convert NAD83 UTM values to NAD27 UTM
values for several years, and I'm pretty sure that I got it right.

I imported the new ascii vector file back into GRASS and compared it to
the USGS' NAD27 roads.  The two files were pretty much the same except for
the obvious differences caused by new road development.

So, I think you have found a fairly serious bug in GRASS.  Do you want to
do the official bug report?

As a product of all that I have this ascii vector file that has the new
roads system in NAD27 UTMs.  If you want it just send me an email.


Roger Miller
Lee Wilson and Associates