Re: NCL vs. WRF_NCL

Hi, a couple of weeks ago I posted an inquiry about the
desirability of using NCL vs. WRF_NCL to plot a lot of
output from our WRF operational runs for detailed analysis.

A couple of you responded, and the general message was
that NCL would provide greater access to control of details,
whereas WRF_NCL would provide a set of higher-level interfaces,
that might be harder to tune for specific needs. It was also
suggested that NCL was a good place to start (one reason being
that when you're looking for help, the NCL community is very
responsive!), and once some confidence is gained, it might make
sense to at least look at WRF_NCL.

I agree with the above and would like to add what I "think" is
a critical observation. WRF output files contain some pretty
funky variables (at least for folks like me who only pretend to
understand atmospheric sciences). The vertical coordinate system
is terrain-following, and if you want the temperature for a given
level, for example, well, all you can get directory from the WRF output
file is the perturbation potential temperature, which you would then
convert to total potential temperature, then you gotta start using other
variables to get a "real" temperature. Further, because of the
nature of the vertical coordinate system, it's not very clear on how
you would plot something as basic as 500mb temperatures (or anything
at 500mb, for that matter).

My observation (I fully realize I might be missing critical pieces) is that NCL
is great for plotting the basic variables from the WRF output files, and even
some variations of those. It's not that hard, especially with guidance from
the NCL community. However, there doesn't appear to be an "easy" way
to use NCL for plotting some of the common variables (again, something like
500mb temperatures or winds). I realize all the necessary info is available
in the WRF output file, and obviously it's "possible" to write NCL scripts to
do this, and maybe it's not even that hard, but to a dummy like me, it's
hard to know where to start. The WRF_NCL scripts, on the other hand,
as advertised, provide a higher-level interface that can do a lot of this
dirty work for you.

So, my personal conclusion is that NCL is a great place to start, and I think
it's a good idea to understand how to plot the basic variables with plain
NCL, and then you're in a good position to start using some of the higher
level features of WRF_NCL (hopefully having gained enough expertise
to modify those routines as necessary).

Thanks to those who offered advice, and helped me with some of the
peculiar issues I was facing.

Don Morton

-- 
Arctic Region Supercomputing Center / U. Montana
http://weather.arsc.edu/