Subject: Re: How can we build own input datasets?
From: Erik Kluzek (erik@cgd.ucar.edu)
Date: Mon Dec 14 1998 - 02:04:04 MST
On Fri, 11 Dec 1998, Patrick Marquis wrote:
>
> We would like to run CCM3.6.6 with "live" initialization datasets. I have searched for and cannot find any references to creating our own input datasets. Is there a data source on the web we can utilize (can initialization from another model be used in part)? If so, are there existing programs to correctly format this data for the CCM?
The CCM Users Guide talks about taking the standard input datasets and
doing vertical and horizontal interpolations on them, to get datasets are
non-standard resolutions. But, we don't really have much on creating your
own datasets. But, this isn't really a problem, take one of the standard
datasets interpolate to the resolution desired, and then replace the
data with your own data. Of course making sure that you provide all the
data types on the input dataset, and provide data in a physically reasonable
and consistent manner. There are various ways of dealing with netCDF
files in this manner, but I don't know of existing programs that will
format generic data to CCM netCDF format. One simple way, would be to use
"ncdump" to convert the netCDF file to ASCII, manipulate the ASCII file
and then convert back to netCDF with ncgen. Or of course you could use Yorick,
or NCL, or any of the other host of programs that can work with netCDF data.
>
> Maybe the question I should ask is how were the included input datasets created and where did the raw data come from?
>
The input datasets came as the result of output from a previous CCM3
simulation. See the CCM3 web page on input datasets for more details...
http://neit.cgd.ucar.edu/cms/ccm3/initial/
Of course now you want to know where the previous input dataset came
from? But, will it's the same answer -- from an even earlier simulation
with the CCM. Since the CCM has been around for such a long time, this
series of input datasets goes back quite a way....
Since the main use of the CCM is for climate studies there hasn't been
a lot of effort put into creating tools to initialize the model from
observed atmospheric states. At least not that I know of. It's possible
other CCM users might have worked on this issue.
Erik Kluzek, (CGD at NCAR)
National Center for Atmospheric Research
Boulder CO, (off) (303)497-1326 (fax) (303)497-1324
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
This archive was generated by hypermail 2b27 : Thu Jun 01 2000 - 09:08:11 MDT