[GRASSLIST:8434] Re: problems in gnomonic projection cont'd
Milan Salek
salek at chmi.cz
Tue Sep 27 03:59:10 EDT 2005
Sorry, the attachments are not allowed in the list. If someone is interested in
resolving of the the issue (which is definitely appreciated), ask me for the
attachment and I will send it directly to the e-mail address.
Milan Salek
Milan Salek napsal(a):
> Hello all,
>
> several weeks I am struggling with reprojection from/to gnomonic projection and
> I am more and more convinced that something is wrong in r.proj/v.proj dealing
> with gnomonic projection.
>
> I was trying to give you the simplest example of the problem that still persists.
>
> Enclosed you will find two locations, rad_m and fi_lambda, both roughly for the
> same region, the former in gnomonic projection and the latter in lat/lon.
>
> In the gnomonic projection there is an artificial array with some distinct
> values creating a cross at a given location (it is the reference point of the
> projection but it could be any arbitrary point)
>
> d.proj -l -d works nicely, the point/cross gives accurate values:
>
> EAST: NORTH: LON: LAT:
> -62.5 237.5 14.44612648 50.01013321
>
>
> When using proj for the LON/LAT:
> proj +proj=gnom +lat_0=50.008 +lon_0=14.447 +a=6379000 +x_0=301500 +y_0=-217500
> +es=0. +no_defs -f %6.0f <<EOF
> 14.44612648 50.01013321
>
> it gives
>
> 301438 -217262 which, in respect to above mentioned results of d.proj -l -d,
> and the false easting/northing given by proj parameters, gives nice match.
>
> Here it works OK.
>
> But, when I am in location fi_lambda and want to reproject 00test map into it,
> it imports well but everything is by 0.2 degree to north.
>
> Reprojection:
> r.proj input=00test location=rad_m mapset=salek output=00test method=nearest
>
> Then, when pointing to the little point/cross:
>
> d.where -l -d
> LON: LAT:
>
> 14.447125 50.20371875
> ------------
>
> The latitude should be near 50.01. The little difference in LON does not bother
> me, but the LAT does.
>
> Similar results are in Spearfish area, even the difference is the same.
>
> If you want to import the data yourself, try in rad_m:
>
> r.in.bin input=<PATH>/input_gnom.int output=00test bytes=2 north=217500
> south=-310500 east=426500 west=-301500 rows=528 cols=728
>
> Has someone an idea why the latitude is wrong?
>
> Best regards
>
> Milan
>
More information about the grass-user
mailing list