[GRASS-dev] barscale in latlong region

[Hamish]

Michael Barton wrote:
> 
> I know this has come up before, but I wanted to see if there might be
> a way to fix this now.

see  http://thread.gmane.org/gmane.comp.gis.grass.devel/2794/focus=2841

> I don't work much with latlon locations, but I know that a lot of
> people do. For large areas especially, it is one of the main ways to
> manage geospatial data. GRASS has pretty good tools for dealing with
> latlon and its idiosyncrasies. However, a peculiar gap is that the
> barscale tool won't work in latlon. Given that I'm told that there are
> GRASS libraries already in existence to deal with latlon to metric
> conversions, it seems odd that we can't put a scale on a latlon map.
> 
> Is this actually a very difficult problem to solve, or just one that
> no one has yet gotten around to doing--I know that you all have sooo
> much free time on your hands ;-)


It's a problem of correctness, not a technical/time issue.
A horizontal plate carrée scale is only valid at a single lat, and
becomes more wrong as you zoom out or look up and down the map. See a
National Geographic map for a solution which has a series of scales in
the shape of a horn, where you choose a scale depending on the lat.


Glynn's suggestion in the above thread of changing the units to degrees
(or min,sec) could work. But then you could use d.grid for that.


note "d.barscale -n" will work in lat/lon as that doesn't need scale.
(north arrow only). Better: replace "d.barscale -n" with a symbol so you
can optionally draw the compass or fancy_compass under the "N" text
instead of the arrow-head.


Hamish