[R-sig-Geo] Good projection for N/S America?

Sun Apr 8 16:56:03 CEST 2007

Denis,

That's really useful. It occurs to me that we only really need a
half-dozen basic projections to cover 90% of user cases. Perhaps these
could be incorporated into the 'sp' group somewhere and relieve the
dependence on proj4. (It could be packaged separately for R  for the
other 10% of cases where its needed.)

THK

On 4/6/07, White.Denis at epamail.epa.gov <White.Denis at epamail.epa.gov> wrote:
> Thanks, Roger.  There was a request to see the R code for these figures.
> Attached is the script for the second PDF file plus the input boundary
> file I used for the hemisphere.  The three projection functions are for
> simple spherical, rather than ellipsoidal, models of the earth.  The
> graticule generating function could be more elegant.  I'm not yet up to
> speed with sp and the many new spatial capabilities in R so please
> excuse the old style "lines()" format encoding and graphics.
>
> Tim, I don't know whether proj4 could do the interrupted sinusoidal.
>
> (See attached file: whemi.projs.r)(See attached file: whemi.lin)
>
> r-sig-geo-bounces at stat.math.ethz.ch wrote on 2007-04-06 04:51:53:
>
> > Since this topic is of general interest, I've made an exception and
> > allowed (this once!) a posting of more than 200K. In general, if
> graphics
> > are big, please consider either an alternative device (png is often
> OK),
> > or posting just a URL to the real file.
> >
> > With apologies to list members on dial-up connections in the field,
> >
> > Roger
> >
> >
> > On Thu, 5 Apr 2007 White.Denis at epamail.epa.gov wrote:
> >
> > > Yes, for many uses that is my choice also.  For the conterminous US
> for
> > > example, the Lambert azimuthal has lower mean distortion than the
> > > commonly used standard projection, the Albers conical equal area,
> > > although Albers was chosen by USGS as a standard because of lower
> > > extreme distortion than many other possible projections.
> > >
> > > For our hemispherical application, because we were gridding the
> data, we
> > > wanted parallels of latitude to be parallel in the projected
> coordinate
> > > space, which we wouldn't get with the Lambert azimuthal.
> > >
> > > (See attached file: whemi.projs.pdf)
> > >
> > > Tim Keitt <tkeitt at gmail.com> wrote on 2007-04-05 10:56:09:
> > >
> > > > Thanks. My application is not that demanding. Really, I just want
> it
> > > > to look reasonable. My plan is to lay out the postings in the
> > > > projected coordinates and then back transform into geographic
> > > > coordinates for analysis. I tried lots of projections and found
> > > > Lamberts Azimuthal Equal Area to be quite good. I like the look of
> the
> > > > Azimuthal Equidistant better, but figured equal area was a good
> > > > choice.
> > > >
> > > > THK
> > > >
> > > > On 4/4/07, White.Denis at epamail.epa.gov
> <White.Denis at epamail.epa.gov>
> > > wrote:
> > > > > Tim,
> > > > >
> > > > > It depends on which kind of distortion is of most concern.  For
> many
> > > > > types of extensive data, especially counts, for example, the
> equal
> > > area
> > > > > property is desirable.  We used the Lambert cylindrical equal
> area
> > > > > projection with standard parallels of +/- 30 degrees for some
> > > western
> > > > > hemispherical work, see reference below.  (The center longitude
> > > could be
> > > > > -80 west, but that is less important than the choice of
> parallels.)
> > > > >
> > > > > Before falling back on the Lambert as an easy to use projection,
> I
> > > tried
> > > > > to get several ESRI products to implement an interrupted
> projection
> > > > > using the sinusoidal projection, in part for reasons given in
> the
> > > second
> > > > > reference.  I used a separate center longitude for north and
> south
> > > of
> > > > > the equator and the appearance is certainly more satisfactory
> than
> > > the
> > > > > Lambert in my opinion.  I'll attach a PDF of an illustration of
> this
> > > > > approach generated in R that I hope you will get but not the
> rest of
> > > the
> > > > > list unfortunately.  I can send PDFs of the references also if
> > > needed.
> > > > >
> > > > > Denis
> > > > >
> > > > > Lawler JJ, White D, Neilson RP, Blaustein AR.  2006.  Predicting
> > > > > climate-induced range shifts: model differences and model
> > > reliability.
> > > > > Global Change Biology 12:1568-1584.
> > > > >
> > > > > White D.  2006.  Display of pixel loss and replication in
> > > reprojecting
> > > > > raster data from the sinusoidal projection.  Geocarto
> International
> > > > > 21(2):19-22.
> > > > >
> > > > > (See attached file: whemi.sinus.pdf)
> > > > >
> > > > > r-sig-geo-bounces at stat.math.ethz.ch wrote on 2007-04-04
> 12:17:39:
> > > > >
> > > > > > Anyone know of a particularly good map projection for showing
> all
> > > of
> > > > > > North and South America without too much distortion?
> > > > > >
> > > > > > THK
> > > > > >
> > > > > > --
> > > > > > Timothy H. Keitt, University of Texas at Austin
> > > > > > Contact info and schedule at http://www.keittlab.org/tkeitt/
> > > > > > Reprints at http://www.keittlab.org/tkeitt/papers/
> > > > > > ODF attachment? See http://www.openoffice.org/
> > > > > >
> > > > > > _______________________________________________
> > > > > > R-sig-Geo mailing list
> > > > > > R-sig-Geo at stat.math.ethz.ch
> > > > > > https://stat.ethz.ch/mailman/listinfo/r-sig-geo
> > > > >
> > > >
> > > >
> > > > --
> > > > Timothy H. Keitt, University of Texas at Austin
> > > > Contact info and schedule at http://www.keittlab.org/tkeitt/
> > > > Reprints at http://www.keittlab.org/tkeitt/papers/
> > > > ODF attachment? See http://www.openoffice.org/
> >
> > --
> > Roger Bivand
> > Economic Geography Section, Department of Economics, Norwegian School
> of
> > Economics and Business Administration, Helleveien 30, N-5045 Bergen,
> > Norway. voice: +47 55 95 93 55; fax +47 55 95 95 43
> > e-mail: Roger.Bivand at nhh.no
> >
> > _______________________________________________
> > R-sig-Geo mailing list
> > R-sig-Geo at stat.math.ethz.ch
> > https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>

-- 
Timothy H. Keitt, University of Texas at Austin
Contact info and schedule at http://www.keittlab.org/tkeitt/
Reprints at http://www.keittlab.org/tkeitt/papers/
ODF attachment? See http://www.openoffice.org/