[Proj] Geoid models egm96 et egm08 in PROJ4

[Charles Karney]

Thanks.  Although I'm confident that the new files are correct, it
would be useful if anyone who is in a position to do so also confirms
this.

   --Charles

On 02/15/2014 11:21 AM, Howard Butler wrote:
> Charles,
>
> I have updated the files at http://download.osgeo.org/proj/vdatum/ with your set and renamed the old ones with a scary filename so that they can stay around for posterity and other archaeological operations if necessary.
>
> https://trac.osgeo.org/proj/ticket/126
>
> Thanks for digging through this. Sorry it took so long to attend to it.
>
> Howard
>
> On Feb 14, 2014, at 9:36 AM, Charles Karney <charles.karney at sri.com> wrote:
>
>> Corrected versions of egm96_15.gtx and egm08_25.gtx are available at
>>
>>    https://sarnoff.sharefile.com/d/sb23d64908da48ffb
>>
>> These were produced directly from the NGA spherical harmonic expansions
>> using the GeoidToGTX program (a sample program distributed with
>> GeographicLib).  These files correct the following issues with the PROJ4
>> files:
>>
>> (1) +/- 5mm quantization errors in egm96.
>> (2) incorrect origins in egm96 and egm08
>>
>> GeoidToGTX faithfully reproduces the grids given by the harmonic
>> synthesis programs (in Fortran) provided by NGA.
>>
>> (The link above should be valid "indefinitely".  However, I can't be
>> sure that it will work beyond about a year.)
>>
>>    --Charles
>>