Re: Southern Hemisphere wind barbs in panel plots

From: Kyle Griffin <ksgriffin2_at_nyahnyahspammersnyahnyah>
Date: Thu Jul 11 2013 - 12:09:06 MDT

Hi Dave,

If you're working on chasing down this bug, some plots I made today led me
to narrow down this bug search significantly, I hope. It seems the problem
only exists if the plots are entirely in the SH (i.e., mpMinLatF and
mpMaxLatF are both negative). Below, I've linked to two test images I
created just now - same data, but the first one has mpMaxLatF = 0.0, where
the second has mpMaxLatF = -0.5. This results in the barbs on the second
plots being incorrect in the second and third panels, while all three
panels are correct on the first image.

Bounded at 0, correct barbs:
http://www.atmos.albany.edu/student/kgriffin/research/swio/iwetkg/images/dt_panel_test/dt_panel_test_1.png
Bounded at -0.5, incorrect barbs:
http://www.atmos.albany.edu/student/kgriffin/research/swio/iwetkg/images/dt_panel_test/dt_panel_test_2.png

Hope this helps the hunt,

Kyle

----------------------------------------
Kyle S. Griffin
Department of Atmospheric and Oceanic Sciences
University of Wisconsin - Madison
Room 1421
1225 W Dayton St, Madison, WI 53706
Email: ksgriffin2@wisc.edu

On Tue, Jul 9, 2013 at 12:54 PM, David Brown <dbrown@ucar.edu> wrote:

> Hi Kyle,
> I will look into this problem as soon as possible.
> -dave
>
> On Jul 8, 2013, at 5:27 PM, Kyle Griffin <ksgriffin2@wisc.edu> wrote:
>
> Hi all,
>
> Although NCL does automatically account for Southern Hemisphere wind barbs
> in single-map applications, I believe I've found a bug in trying to use
> this feature in panel applications. My experiments suggest that the
> automatic flipping of the wind barbs only occurs on the first panel of the
> plot, leaving the other panels with NH wind barbs (see attached). The first
> thought for a workaround is to manually set vcWindBarbTickAngleF = -62.0,
> but this results in flipping the direction for *all* panels, making the
> first panel plot NH-oriented barbs. I've included the resource settings and
> plotting commands I've used for completeness, since my script is likely too
> complicated/messy to include here and probably not necessary for diagnosing
> this problem. Thanks!
>
>
> Kyle
>
> ; All vector resources, even those that shouldn't be relevant here
> vres@vcGlyphStyle = "WindBarb"
> vres@vcRefLengthF = 0.022
> vres@vcRefMagnitudeF = 10.0
> vres@vcRefAnnoOn = False
> vres@vcMinDistanceF = 0.022
> vres@vcWindBarbLineThicknessF = 3.0
> ;vres@vcWindBarbTickAngleF = -62.0
>
> ;............. plot data below, looping over i until numPanels-1
>
> mpid(i) = gsn_csm_map(wks,mpres)
> pwflid = gsn_csm_contour(wks,theta({curdate},:,:),pwres)
> cnid = gsn_csm_contour(wks,vort({curdate},:,:),cres)
> vcid = gsn_csm_vector(wks,u_pv(i,:,:),v_pv(i,:,:),vres)
>
> overlay(mpid(i),pwflid)
> overlay(mpid(i),cnid)
> overlay(mpid(i),vcid)
>
> ;.....set some generic panelRes options here, then....
>
> gsn_panel(wks,mpid,(/ntimes,1/),panelres)
>
>
> ----------------------------------------
> Kyle S. Griffin
> Department of Atmospheric and Oceanic Sciences
> University of Wisconsin - Madison
> Room 1421
> 1225 W Dayton St, Madison, WI 53706
> Email: ksgriffin2@wisc.edu
> <dt_panel_fig.png>_______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
>
>

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Thu Jul 11 12:09:35 2013

This archive was generated by hypermail 2.1.8 : Fri Jul 12 2013 - 16:37:39 MDT