[Proj] Re: Proj Digest, Vol 3, Issue 4

[Melita Kennedy]

Dylan,

I can explain why the origin point doesn't give the expected values.
Michigan GeoRef (and State Plane's Alaska zone 1) use what ESRI 
calls Oblique Mercator Natural Origin. The xy 0,0 location is where
the central line crosses the aposphere's equator rather than the
intersection of the longitude and latitude of center. The origin is 
usually quite close to the ellipsoid's equator. That's why the false 
easting and northing values are so large.

Melita

-- 
Melita Kennedy
mkennedy at pe.net

> Date: Tue, 17 Aug 2004 09:47:19 -0700
> From: Dylan Keon <keon at nacse.org>
> Subject: [Proj] trouble using Michigan GeoRef projection
> To: proj at xserve.flids.com
> 
> Hello,
> 
> I posted this question to mapserver-users and Frank W. replied,
> suggesting that I try this mailing list.  My original question and
> Frank's reply are copied below.  Thanks for any assistance.
> 
> --Dylan
> 
> >> Hi,
> >>
> >> I'm trying to use a layer that's projected in the "Michigan GeoRef"
> >> projection, but I can't get it to line up with other data.  I'm using
> >> the projection parameters provided by the state of Michigan
> >> (http://www.michigan.gov/documents/DNR_Map_Proj_and_MI_Georef_Info_20889_7.pdf).
> >>
> >> Here's what I get when trying to line up a Michigan layer (GeoRef
> >> projection) with a national 24k quad index (geographic), putting
> >> everything into geographic for now:
> >>
> >>    http://gis.nacse.org/mstemp/109267161738211.png
> >>
> >> I haven't found examples of this online, and wonder if I messed up a
> >> PROJ definition or something.  Thanks for any help.  Mapfile is below.
> >>
> >> --Dylan
> >>
> >>
> >> NAME 'test'
> >> SIZE 640 480
> >> STATUS ON
> >> EXTENT -126 34 -66 46
> >> UNITS DD
> >>
> >> IMAGETYPE png
> >>
> >> WEB
> >>   IMAGEPATH '/www/temp/'
> >>   IMAGEURL '/temp/'
> >> END
> >>
> >> PROJECTION
> >>   'init=epsg:4326' #latlon
> >> END
> >>
> >> LAYER
> >>   NAME 'quad_index'
> >>   DATA '/gis/data/us/24kgrid.shp'
> >>   STATUS DEFAULT
> >>   TYPE POLYGON
> >>   CLASS
> >>     COLOR -1 -1 -1
> >>     OUTLINECOLOR 0 0 255
> >>   END
> >>  PROJECTION
> >>     'init=epsg:4326' #latlon
> >>   END
> >> END
> >>
> >> LAYER
> >>   NAME 'ecoregions'
> >>   DATA '/gis/data/mi/ecoreg100'
> >>   STATUS DEFAULT
> >>   TYPE POLYGON
> >>   CLASS #simple for now
> >>     COLOR 198 123 48
> >>   END
> >>   PROJECTION #NAD 1983 Michigan GeoRef Meters <102123>
> >>     'proj=omerc'
> >>     'k_0=0.9996'
> >>     'lat_0=45.30916666666666'
> >>     'alpha=337.255555555556'
> >>     'lonc=-86'
> >>     #'x_0=2546731.496'    # PROJ
> >>     #'y_0=-4354009.816'   # manual
> >>     #'ellps=GRS80'        # says these are not
> >>     #'datum=NAD83'        # needed for oblique mercator,
> >>     #'units=m'            # but I tried them too,
> >>     #'no_defs'            # in various combinations
> >>   END
> >> END
> >>
> >> END
> >>
> 
> > Dylan,
> >
> > Hmm, I made an effort but couldn't get this to work.  I would suggest you
> > take your question to the PROJ.4 mailing list.  I would also include that
> > ESRI defines this projection as:
> >
> > PROJCS["NAD_1983_Michigan_GeoRef_Meters",
> >     GEOGCS["GCS_North_American_1983",
> >         DATUM["D_North_American_1983",
> >             SPHEROID["GRS_1980",6378137,298.257222101]],
> >         PRIMEM["Greenwich",0],
> >         UNIT["Degree",0.017453292519943295]],
> >     PROJECTION["Hotine_Oblique_Mercator_Azimuth_Natural_Origin"],
> >     PARAMETER["False_Easting",2546731.496],
> >     PARAMETER["False_Northing",-4354009.816],
> >     PARAMETER["Scale_Factor",0.9996],
> >     PARAMETER["Azimuth",337.255555555556],
> >     PARAMETER["Longitude_Of_Center",-86],
> >     PARAMETER["Latitude_Of_Center",45.30916666666666],
> >     UNIT["Meter",1]]
> >
> > And stress lat/long value provided for the GeoRef projection origin in the
> > .pdf file referenced.  While a test point at the origin is the worst kind of
> > test point for a projection, it is something to work with.  I couldn't even
> > get it to work out properly.
> >
> > Hopefully Gerald Evenden will know whats up.
> >
> > Best regards,
> > --