Re: temperature contours look strange

From: Mary Haley <haley_at_nyahnyahspammersnyahnyah>
Date: Wed Sep 12 2012 - 11:05:37 MDT

Vivian,

In a case like this, just looking at your script is not very helpful. I don't know if the problem is with the data or your plot. Also, which edge are you talking about, and which contours are "strange"? Are you talking about the line contours in the upper left of the plot?

As a side, I noticed that in your plot, you don't have enough contour labels for the labelbar. It looks like you are trying to select discrete contour levels, but not providing enough labels.

When you contour with NCL, it *always* does its contouring based on values that fall above and below the contour levels, and never based on levels exactly equal to some value.
If you want to make NCL behave like it is contouring at exact values, then you need to pick contour levels that straddle the values that you're interested in.

So, in your case, instead of this:

  res1@cnMinLevelValF = 1
  res1@cnMaxLevelValF = 7
  res1@cnLevelSpacingF = 1

which I assume
Try:

  res1@cnMinLevelValF = 1.5
  res1@cnMaxLevelValF = 6.5
  res1@cnLevelSpacingF = 1

This will only work as expected if you do indeed have discrete values like 1.0, 2.0, etc.

Meanwhile, if you want me to look at this problem further, can you provide your dataset? You can send this to me offline if it is small (< 10 MB), or you can use our ftp:

http://www.ncl.ucar.edu/report_bug.shtml#HowToFTP

--Mary

On Sep 10, 2012, at 7:52 PM, Yi Huang wrote:

> Dear ncl experts,
>
> I am trying to use ncl to plot some variables for a cross section from my WRF outputs (innermost domain). However, the temperature contours look very strange near the edge. (Please see both the figure and my script attached). This happened during certain periods of time of the simulation, but not always. I am sure that the two end points of the cross section are within the domain. What could be the cause of this problem? How can I fix it?
>
> Thanks very much for your helps.
>
> Cheers,
> Vivian
> <test.ps><cross_section_categories_matched.ncl>_______________________________________________
> 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 Wed Sep 12 11:05:52 2012

This archive was generated by hypermail 2.1.8 : Thu Sep 13 2012 - 14:22:31 MDT