[Proj] Proj4 datum transform assumption Question

[Frank Warmerdam]

Joel_Cusick at nps.gov wrote:
> Hello,
> Brand new subscriber, but long time user of DNRGarmin.  I am trying to
> resolve the datum transform signage in Proj 4 projection engine that is
> being used by Minnesota's DNRGarmin software.  One can open up the ESRI or
> EPSG files and directly observe the 7 parameters for datums.  Since
> DNRGarmin loads to and from Garmins using Garmin proprietary protocols, the
> Proj4 text files are used to transform to or from WGS84 depending on data
> flow.  Loading from a Garmin, its expected to transform from WGS84 to
> NAD83.  When Sending to a Garmin from a NAD83 shape, its expected to
> transform from NAD83 to WGS84.
> 
> question:  Does Proj4 assume data is going to WGS84?  Using the same _5
> ESRI transform parameters, but written in DNRGarmin, the syntax used is
> "+towgs84".  This infers that Proj4 is written as a default From NAD83 to
> WGS84.  This is counterintuitive to documentation presented by Soler/Snay
> article (see Table 1, Page 51, last column
> http://www.ngs.noaa.gov/CORS/Articles/SolerSnayASCE.pdf ) for the transform
> parameters from ITR96 to NAD83.  When comparing the articles's 7
> parameters, they are exactly opposite in signage to the DNRGarmin program.
> Just want to confirm how exactly Proj4 is coded.

Joel,

I must confess to being confused, perhaps partly because I'm not
familiar with DNRGarmin.

The +towgs84 transformation is used to transform from other
datums to WGS84 (in geocentric space).  The transformation is
reversed when going from WGS to other datums.

I am aware there are various sign conventions for towgs84
parameters, but PROJ.4 uses the same conventions used by
mapinfo, OGC, and NIMA MUSE (from which this transformation
was derived).

> # NAD_1983_To_WGS_1984_5
> <1515> +proj=longlat +ellps=GRS80
> +towgs84=-0.991,1.9072,0.5129,-0.02579,-0.00965,-0.01166,0 +no_defs
> no_defs <>

Note that by default PROJ.4 considers NAD83 and WGS84 identical.
It is only by providing specific transformations, such as above,
that any distinction will be made between these two datums in PROJ.4.
The above definition for 1515 is not in the normal PROJ.4 definitions
list, and must have been introduced by the folks who produced DNRGarmin.

Best regards,
-- 
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | Geospatial Programmer for Rent