Re: gsnMaskLambertConformal

From: Mary Haley <haley_at_nyahnyahspammersnyahnyah>
Date: Mon, 4 Dec 2006 09:36:11 -0700 (MST)

On Fri, 1 Dec 2006, David B. Reusch wrote:

> I have been creating maps for years like the one below using
> gsnMaskLambertConformal and for the most part this is fine.
> my understanding/experience is that the masking process inherently (?)
> produces Postscript such that this figure is actually a rectangle and the
> areas outside my masked area are white-filled objects covering up the map
> underneath. this leads to problems when I load the postscript file into
> Illustrator (for example, all versions I've used over the last few years) and
> put these figures on a non-white background such as this:
> now the white backgrounds show up and it's not particularly pleasing to the
> eye. I've been dealing with this for some time by just not using colored
> backgrounds, but now that I've figured out how to do this as I'd like (if not
> perfectly) with another plot (see below, "ocean" is transparent), I'm
> wondering if there's a way to accomplish something similar with this plot?

Hi Dave,

I've never been happy with the way that these masked lambert conformal
plots are produced. Your understanding of how these plots are created
is correct: since we don't have a way to generate these kind of map
projections automatically, we use the kludgy way of drawing the full
circular map and then masking out roughly half of it by drawing a white
filled background over the area we're not interested in.

Unfortunately, as you saw, this causes problems with other tools as
this masked white area may show up unexpectedly, or else you'll see
remnants of the border of the area we're trying to mask.

I think the only way around this problem is to have NCL be updated so
that it can create these kind of projections automatically, without
any masking having to take place.

I'll see if this can be added to the list. Meanwhile, is it possible
in your tool to tell it to make "white" a transparent color?

> a possibly significant difference between these plots is that the one I'd
> like to fix contours a much larger 2-D grid that only has data defined within
> the North Atlantic basin. the plot I prefer (even if it's still a bit weird)
> is contouring non-gridded point data from Greenland only. (note: the
> "weird" white box mid-way up the west coast of Greenland does not always
> appear exactly like this, though I think something's usually there.) I think
> it might be possible to change my plots to 1-D if that would help. I'm not
> sure why I went to 2-D with this data (it's read in as 1-D) unless that used
> to be a contouring requirement and no longer is?

That is true, but it somewhat depends on what kind of data you are
trying to plot. If you can email me personally a sample data set, or a
detailed description of your grid, we can probably tell you the best
way to contour it. You can also take a look at our "contouring
non-uniform grids" page at:

Click on any one of the links to see some examples of what kind of
grids you can generate contours on.


> I've looked at the mp resources but have yet to convince myself that there's
> a path there to where I want to be, so I was hoping someone else might have
> some suggestions.
> thanks,
> dave
> p.s. code for the above is available on request.
> --
> 517 Deike Building
> Earth & Environmental Systems Institute
> Penn State University
> University Park, PA 16802
ncl-talk mailing list
Received on Mon Dec 04 2006 - 09:36:11 MST

This archive was generated by hypermail 2.2.0 : Tue Dec 05 2006 - 13:29:24 MST