[GRASS-user] Re-projected Data Position Mismatch

[Jeshua Lacock]

> On Aug 27, 2017, at 8:54 PM, Anna Petrášová <kratochanna at gmail.com> wrote:
> 
> On Tue, Aug 22, 2017 at 7:46 PM, Jeshua Lacock <jeshua at 3dtopo.com> wrote:
>> 
>> To hopefully help troubleshoot; I just reprojected one of the raster tiles (from epsg: 3857), into the source location of one of the lonlat vectors (reverse projections from my OP), and the datasets are offset by the same distances. Since the dimensions of the raster are being changed (by r.proj), it leads me to think it must be a datum or coordinate system misalignment and not a projection issue.
> 
> I have the same problem, working with NAIP imagery. It is related to:
> https://trac.osgeo.org/grass/ticket/2229
> 
> I have to remove nadgrids: @null from the PROJ_INFO file to be able to
> reproject into that location, but then it is shifted. gdalwarp helps.

Hi Anna!

Thank you very much for confirming that! I am working with the NAIP imagery as well.  :)

I have found that their original Geotiff assets work perfectly.

In fact, I was very happy to stumble on to the fact that the complete NAIP archive (~250 terabytes) is available as a bucket drive on Amazon Web Services (AWS). So I setup GRASS instances to process the tiles, then download the processed, reprojected images compressed as JP2s. I am paying for the bandwidth and compute time, but I think its worth it for my purposes. I’ll be able to process and download the imagery I need in about 60 days compared to over 300 days without AWS!


Cheers,

Jeshua Lacock
Founder/Engineer
<3DTOPO.com>
GlassPrinted.com