[GRASS-dev] r.watershed at poles

[Andy Wickert]

Hi Soeren and all,

I appreciate the opinion, and in fact have split continents in the past,
but the project that I want to undertake requires many calculations with
variable sea-levels and would be much easier to run on a full globe (albeit
with null values in the ocean at each particular time-step).

That, and this unfortunately still doesn't address doing Antarctica without
reprojecting.

So really what I am saying is that I would like to hear ideas about the
technical side of r.watershed (maybe I should just email Markus Metz
directly) and how to change where it creates sinks. I know that it is able
to go across the 180 meridian, so this boundary has been modified for
lat/lon, and it therefore in my mind would be feasible to tell it that the
north and south poles are not a sink for flow.

If there is really no way to do this with r.watershed, I will write code
myself, but I am hoping that I can instead just improve r.watershed.

Thanks,

Andy

On Fri, Oct 2, 2015 at 2:04 AM, Sören Gebbert <soerengebbert at googlemail.com>
wrote:

> Hi Andy,
> running r.watershed for the whole globe is IMHO not necessary. Just
> run r.watershed for each continent (except Europe/Asia) and isles
> (greenland, UK, ...). Everything that is parted by an ocean or a
> channel can be processed independently.
>
> Best regards
> Soeren
>
> 2015-10-02 0:27 GMT+02:00 Andy Wickert <andrewwickert at gmail.com>:
> > On Thu, Oct 1, 2015 at 3:52 PM, Markus Neteler <neteler at osgeo.org>
> wrote:
> >>
> >> On Thu, Oct 1, 2015 at 7:10 AM, Andy Wickert <andrewwickert at gmail.com>
> >> wrote:
> >> > Dear GRASS developers,
> >> >
> >> > Would you know of an easy way to make r.watershed work in
> >> > latitude/longitude
> >> > over the pole of a planet (in my case, over Antarctica) by specifying
> >> > that
> >> > the polar boundary in lat/lon is not a sink?
> >>
> >> Wouldn't be a stereographic projection more appropriate?
> >>
> >>
> >>
> https://grass.osgeo.org/uploads/images/Gallery/raster/polar_stereographic.png
> >>
> >> Best
> >> Markus
> >
> >
> > Thanks for the suggestion. However, my goal is to run r.watershed across
> the
> > whole globe, so a stereographic projection would an ad-hoc solution.
> >
> > (I have also had problems in the past projecting between polar
> stereographic
> > and lat/lon, but this is a secondary issue.)
> >
> > Andy
> >
> > _______________________________________________
> > grass-dev mailing list
> > grass-dev at lists.osgeo.org
> > http://lists.osgeo.org/mailman/listinfo/grass-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/grass-dev/attachments/20151002/1e9d44be/attachment.html>