[GRASSLIST:6284] GPS points off by about 120m to the west

GRASS GRASS user
1

Greetings!

after retrieving some waypoints from my Garmin 12 GPS, i noticed that when
imported into GRASS, they were offset by about 120 meters to the west....!

here is a picture of the bad (red) points, and corrected (blue) points:
http://basho/~dylan/temp/bad_gps.png

i am nearly positive that the data in GRASS is good: in NAD83/WGS80 UTM zone
10N projection.

the GPS was set for the same projection parameters...

any ideas on what could have caused this 120meter shift to the west?

could my GPS really be in a NAD27 datum, and still be reporting NAD83 ?

any ideas? .

thanks in advance!

--
Dylan Beaudette
Soil Science Graduate Group
University of California at Davis

2

after retrieving some waypoints from my Garmin 12 GPS, i noticed that
when imported into GRASS, they were offset by about 120 meters to the
west....!

here is a picture of the bad (red) points, and corrected (blue)
points: http://basho/~dylan/temp/bad_gps.png

i am nearly positive that the data in GRASS is good: in NAD83/WGS80
UTM zone 10N projection.

the GPS was set for the same projection parameters...

any ideas on what could have caused this 120meter shift to the west?

could my GPS really be in a NAD27 datum, and still be reporting NAD83
?

The GPS is exporting waypoints in WGS84;
the error you are seeing is surely due to a datum shift;
v.in.garmin should take care of all that for you.

i.e. regardless of what you have the GPS datum set to display on the
screen, it will output WGS84. At least that what my Garmins do.

were you importing with v.in.garmin or v.in.ascii?

note v.in.garmin in 6.0 is broken, get the latest version of the script
from CVS.

Hamish

3

On Mar 29, 2005, at 11:58 PM, Hamish wrote:

after retrieving some waypoints from my Garmin 12 GPS, i noticed that
when imported into GRASS, they were offset by about 120 meters to the
west....!

here is a picture of the bad (red) points, and corrected (blue)
points: http://basho/~dylan/temp/bad_gps.png

i am nearly positive that the data in GRASS is good: in NAD83/WGS80
UTM zone 10N projection.

the GPS was set for the same projection parameters...

any ideas on what could have caused this 120meter shift to the west?

could my GPS really be in a NAD27 datum, and still be reporting NAD83
?

The GPS is exporting waypoints in WGS84;
the error you are seeing is surely due to a datum shift;
v.in.garmin should take care of all that for you.

i.e. regardless of what you have the GPS datum set to display on the
screen, it will output WGS84. At least that what my Garmins do.

were you importing with v.in.garmin or v.in.ascii?

note v.in.garmin in 6.0 is broken, get the latest version of the script
from CVS.

Hamish

Hamish and others who replied,

thanks for the information on this -- i was using v.in.ascii to import these points ... thinking that the GPS would be exporting my points in NAD83/UTM as the GPS unit was configured to use. i will give it a try with the CVS copy of v.in.garmin a little later this week.

on a related note, if i am sending points to the GPS- they would need to be transformed to WGS84 as well?
maybe cs2cs can do this for me?

thanks!

--
Dylan Beaudette
Soils and Biogeochemistry Graduate Group
University of California at Davis
530.754.7341

4

thanks for the information on this -- i was using v.in.ascii to import
these points ... thinking that the GPS would be exporting my points in

NAD83/UTM as the GPS unit was configured to use. i will give it a try
with the CVS copy of v.in.garmin a little later this week.

on a related note, if i am sending points to the GPS- they would need
to be transformed to WGS84 as well?

I think so, but I'm not 100% sure. I usually just try to keep the GPS in
lat/lon WGS84 to keep things sane. I see that 'gpstrans -s' lets you
pick a datum and UTM format. The top of the gpstrans output, it lists
what you are using I think. Also note Garmin protocol outputs GMT, not
local time as displayed on GPS display.

gpsbabel?

maybe cs2cs can do this for me?

well sure. v.in.garmin uses cs2cs to do just this.

I've been meaning to port the m.proj2 module from GRASS 5.4 to GRASS 6
(well a functional replacement anyway) which will do WGS84 import/export
from a list of coordinates. Still haven't got to it, partly because I
haven't figured out a good name for it (g.reproj, g.pointproj ??),
partly because I haven't figured out how to get a shell script to accept
input piped from stdin (as well as a file). The main part of the script
is actually very simple so the only reason to do it in C is if I can't
figure out how to support the dual input methods.

Hamish