[GRASSLIST:4175] Re: Map import cluelessness

[Ian Macmillan]

Chris,
I think I understand your problem, it sounds like you are trying to georectify
your mercator map in degrees.  However, mercator locations use measurement
units of meters only.  Degrees are not a valid way of measuring distances in
mercator locations in GRASS (kind of makes sense).  So for example a location
on the earth with a mercator map would be something like 546783, 3156734 as
easting, northing, NOT 28.5E, 75.2N.  One thing you could try if you only have
degrees on your map is to use i.target, etc. from your xy location straight to
your lat long location (make sure that the datum and ellipsoid match on your
lat-long location and your scanned map).  That should do the trick I think. 
Try to use at a minimum 6 points in i.points, there will be a lot of distortion
in the map, and the more points you define, the more correct the distortion
will be.

-Ian

Quoting "C.S. Cornuelle" <bob at xmission.com>:

> Ian,
>
> See below, please.
>
> Thus spake Ian MacMillan (ian_macmillan at umail.ucsb.edu):
>
> > So as I understand it, I think you set up the mercator projection
> > correctly.  One question, are you sure you want a mercator projection,
> > and not a UTM (universal transverse mercator) projection?  Generally
> > mercator projections are only used with large areas of the world.  With
> > a UTM, you just need to input the zone number, a map can be found here.
> > http://www.dmap.co.uk/utmworld.htm
>
> Yes, I need mercator as the map is in that projection and covers
> Eurasia and Africa.  Too big, I agree.  :^/
>
> > If you do want a mercator projection, then the typical parameters for a
> > world map would be pretty much what you put
> > Latitude of true scale : 0
> > Cenral meridian : whatever is appropriate for your map
> > Scale factor at central meridian : 1
> > plural form of units : meters
> >
> > This location differs from an XY location in that as you move away from
> > the latitude of true scale, the vertical and horizontal scale do not
> > stay equal to each other.  For example, world mercator maps really
> > distort the north and south poles.  An xy location works essentially
> > like Photoshop, the x axis and y axis are always the same scale.
> >
> > From what I know, all types of reprojecting like you are trying can be
> > done completely within GRASS.
> >
> > Does this help?
>
> Hmm.  I'm trying eventually to get to lat-lon.  As you nicely (thanks
> again) describe it, it is necessary to read the file into an XY
> location, rectify it (i.rectify and related commands) into a Mercator
> location, and finally import it into a lat-lon location.
>
> All well and good.  However, the warning in i.rectify where the
> projections do not match (well, of course they don't ...) is still
> there.  This in fact leads to null maps on the output for me.
>
> What confuses me more is that once the Mercator location is established,
> setting the projection seems only to leave the option of making the map
> the same size and resolution as the corresponding XY location.  It does
> not seem possible to enter into DEFINE THE DEFAULT REGION any degree-
> based coordinates, e.g. 71N and such.  If the result is to be a fully
> geocoded database, then I would have guessed that the Mercator location
> would be defined in terms of longitude and latitude.
>
> Clearly, again, I'm missing something here ...
>
> <snip>
>
> --
> Adios,
> Chris Cornuelle
> bob at xmission dot com
>
>


-----------------------------------------------------
Ian MacMillan
Geological Sciences-UCSB

"insert witticism here"