[RegCNET] Re: formats

Moetasim mashfaq at purdue.edu
Tue Apr 4 17:24:29 CEST 2006


Hi Afzaal,
You may change the Grads format file to netcdf format. The easiest way is by 
using las4d script in Grads. Few things to remember about last4d:
1) The script doesn't work in gradsc so use it in gradsnc
2) The script doesn't work on sigma level data (RegCM ATM/RAD files). If your 
file is any of these then I suggest you to first convert them in pressure 
level using Sigma2P. Remember that I had sent a Sigma2P code on RegCNET in 
past for RAD file.

Once you have data in netcdf format then you can either use Grads or ferret.In 
Grads, you can use functions such as scorr(), tcorr() to calulate spatial and 
temporal correlation between two variables. In Ferret,look for variance.jnl 
script which calulates correlation/covariance.

I would suggest you that whenever you need to convert data from one format to 
other, the conversion should be 'to Netcdf' not to any other. The data in 
Netcdf format gives user a great ease in manipulation due to available large 
support. 

Quoting Afzaal Karori <afzaalkarori at yahoo.com>:

> Deat Moetasim!
>    
>   I hope it will find you happy and healthy.
>    
>   Dear ! i have two files; one in ctl format and the other is NCEP's NetCDF.
> I need to find correlation between my data (.ctl) and .nc file. Can I convert
> .nc file to gradsc .ctl file, if so then kindly suggest me the way through
> wich I can do it. I will be grateful to.
>    
>   Thanks 
> 
> 
>                           ***************************
>   MUHAMMAD AFZAAL KARORI
>   Meteorologist,
>   Research and Development Division,
>   Pakistan Meteorological Department,
>   Sector H-8/2, P O Box 1214,
>   Islamabad - Pakistan
>   Phone: + 92 - 51 - 9250360,61,62,63
>   Cell: +923345081048 
>   ***************************
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 		
> ---------------------------------
> Yahoo! Messenger with Voice. Make PC-to-Phone Calls to the US (and 30+
> countries) for 2¢/min or less.



More information about the RegCNET mailing list