[PROJ] Fwd: some help with cm-level GRS80 to WGS-84 transformation?

[Joseph Boardman]

I can do the transform I want in the ESRI PE (used by ENVI) and get a 
dx,dy,dz grid much as in the link I just sent (say +/-1m in x,yz 
differences in conus).

But I'd like to do it in Proj if it can.

ENVI> proj1=envi_proj_create(/geographic,datum='WGS-84')
ENVI> proj2=envi_proj_create(/geographic,datum='North America 1983')
ENVI> lon=-120d
ENVI> lat=40d
ENVI> z=3500d
ENVI> 
envi_convert_projection_coordinates,lon,lat,proj1,lon2,lat2,proj2,input_z=z,output_z=z2
ENVI> lon2
       -119.99998689405381
ENVI> lat2
        39.999994689014549
ENVI> z2
        3500.5506024310116
ENVI> proj3=envi_proj_create(/utm,zone=11,datum='WGS-84')
ENVI> envi_convert_projection_coordinates,lon,lat,proj1,x1,y1,proj3
ENVI> envi_convert_projection_coordinates,lon2,lat2,proj1,x2,y2,proj3
ENVI> x2-x1
        1.0991263525502291
ENVI> y2-y1
      -0.62730476632714272

at 120,40 I get a dx dy dz for WGS-84 to NAD83 of 1.1m, -.62m, -.62m

Can Proj fdo a similar datum shift? If so how?

Thanks!

Joe


On 1/19/2022 4:56 PM, Nicolas Cadieux wrote:
> Hi,
>
> I think he is trying to go from the GRS80 Ellipsoid to the WGS84 
> Ellipsoid...  If this is the case, the difference is already 
> sub-millimeter.   Both have the Semi-major axis of 6378137 but the 
> Inverse of flattening is 298.257223563 for wgs84 and 298.257222101 for 
> GRS80.
>
> Nicolas
>
>
> On 2022-01-19 6:06 p.m., Greg Troxel wrote:
>> Kirk Waters - NOAA Federal via PROJ <proj at lists.osgeo.org> writes:
>>
>>> I'm sure others with more experience will jump in, but I think the main
>>> issue here is that WGS84 is a family of reference frames that are 
>>> all based
>>> on the GRS80 ellipsoid. So, the problem doesn't exactly make sense 
>>> and the
>>> various realizations of WGS84 differ at the cm level. Are you perhaps
>>> trying to go from a NAD83 realization (such as NAD83(2011)) to a WGS84
>>> realization (maybe ITRF14
>>> or something)?
>> I think Kirk has it exactly right.
>>
>> Also, GRS80 is not a datum, but only a name for an ellipsoid. So
>> "converting from GRS80" does not make sense.
>>
>>
>> Further into the weeds, WGS84 realizations are dynamic datums and if you
>> are talking about coordinates of crust-fixed objects you need to pay
>> attention to epoch.  If you care about a few cm, you really need to pay
>> attention.
>>
>> _______________________________________________
>> PROJ mailing list
>> PROJ at lists.osgeo.org
>> https://lists.osgeo.org/mailman/listinfo/proj
>