[GRASS-user] Feet-to-Meter Conversion

[Hamish]

Rich wrote:
>   The standard conversion factor
> from feet to meters is 1 ft = 0.3048 m.
> However, the conversion factor for US Survey Feet
> (frequently the distance unit in the US State Plane Coordinate
> System) is 1 US ft = 0.3600 m.

"frequently" is not so much anymore. I think that only 2-8 or
so states still have the US Survey Foot on the books for their
modern state plane definitions, and it is less each year. Almost
all of them now use the converts-to-meters-exactly version. When
dealing with historical maps it turns up of course.

Almost all "feet" in the US since the last 30-40 years is the
converts-to-meters-exactly version. (stupid UNIX `units` command
bases which to use on the current locale setting! argh!)

>   I don't know the magnitude of the error when GRASS
> assumes international feet (0.3408) as the conversion factor.

In general GRASS assumes nothing and only works from official
sources. So you get what the defined location is supposed to
use and very rarely need to touch it.

> When I have a project location in SPCS is there any harm in
> manually changing the PROJ_UNITS from 0.3048 to 0.3600?

Yes, as if it is a modern SP it is most likely a mistake to do
that, as it will be properly using the converts-to-meters-exactly
version these days.


A while back* I did an extensive audit of the GRASS codebase for
unit conversion factors. If you know of a specific error please
let me know.

(* audit was done a while before the unit consolidation effort in
trunk's libgis, I haven't deeply reviewed those new changes)


Hamish