Re: Domain Averaged Precipitaion

You do not have to "reproject" (regrid) to a rectilinear grid.

You can use:
   http://www.ncl.ucar.edu/Document/Functions/Built-in/wgt_areaave2.shtml

   function wgt_areaave2 (q, w[*][*], opt)

Ideally, the argument 'w' would contain the area represented
by each grid point.

For example, the CESM POP model outputs the following variable:
           float TAREA(nlat, nlon) ;
                 TAREA:long_name = "area of T cells" ;
                 TAREA:units = "centimeter^2" ;

The weighted areal average of some variable 'x' is
         xAreaAvg = wgt_areaave2 (x, TAREA, opt)

==
I am not that familiar with WRF output but I don't think this type of
information is available. However as a proxy you could use the following:

     lat2d = f->XLAT(0,:,:)
     printVarSummary(lat2d)

     clat2d = cos(0.01745329*lat2d)
     xAreaAvg = wgt_areaave2 (x, clat2d, opt)

Good Luck

On 11/3/11 1:50 PM, Lisi Pei wrote:
> Hello,
>
> I am doing the WRF model output validation and may need to calculate a
> domain averaged variables like precipitation etc. to compare with the
> observations.
> My question is that the WRF simulation I conducted is with a lambert
> projection so the XLAT and XLONG variable are both two-dimensional.But
> in the NCL function wgt_areaave
> (http://www.ncl.ucar.edu/Document/Functions/Built-in/wgt_areaave.shtml)
> all the examples give me the impression that we can only give it a 1
> dimension range of the lat and lon to specify the domain you want to
> be averaged.So in this way,this function can not work for the wrfout*
> variables,right?
> What I am thinking about is that seems I must reproject the wrfout*
> variables into a native lat and lon which each coordinate is 1
> dimension.Is that what people do with wrf output data to get a domain
> average?Or can anyone share with me how you do it in this case?
>
> Thank you for any coming help in advance.
>
> Best Regards,
> Lisi
> _______________________________________________
> 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 Fri Nov 4 07:27:03 2011