Re: Possible bug in wrf_td fuction

Mary,

The problem appears to be in the calculation of the dew point
temperature using the wrf_usr_getvar NCL function. This NCL function
uses the qvapor and temperature from the WRF-ARW as inputs to calculate
both the dew point temperature and the relative humidty.

As the data and NCL scripts I put on your ftp site show, plots of the
vertical profiles of dew point temp and relative humidty are
contradictory and thereby suggest to me that the problem is with NCL not
the WRF data. If the problem had been in the WRF data, then the
relative humidity plot would show supersaturations near 100 mb, as the
dew point and dry bulb temperature plot indicates. This is not the case
however.

MJ

On Thu, 2011-05-26 at 14:03 -0600, Mary Haley wrote:

> Hi Michael,
>
>
>
> This question is more appropriate for wrfhelp. I've CC'ed them on this
> message.
>
>
> --Mary
>
>
> On May 26, 2011, at 1:02 PM, Michael J Murphy wrote:
>
>
>
> > The Problem:
> >
> > I have been using the wrf_usr_getvar function to calculated
> > and plot dew point temperature from WRF-ARW output (and
> > input) files. When plotting the dew point temp in vertical
> > profiles I have found strange behavior in the upper
> > atmosphere (near 100 mb). This strange behavior includes
> > the same value of dew point temp (approx -80 degrees
> > Celsius) at numerous sigma levels. The relative humidity
> > extracted with the wrf_usr_getvar function does not show the
> > supersaturations implied by these strange dew point temp
> > reading in the upper atmosphere.
> >
> >
> > Error Messages:
> >
> > There is no error message associated with this problem.
> >
> >
> > My Thoughts:
> >
> > The wrf_usr_getvar seems to call the wrf_td function for its
> > dew point temperature calculation. I have been unable to
> > locate the source code for this wrf_td function but suspect
> > this is the source of the strange dew point temp values.
> >
> >
> > Background Info:
> >
> > 1) Name of WRF data file: wrfinput_d01--bug-wrf_td.gz
> > (NOTE: I am giving the wrfinput file because it is smaller
> > and has the same issue as the wrfout file)
> > 2) Name of NCL scripts:
> > plot-dewpt--bug-wrf_td.ncl & plot-rh--bug-wrf_td.ncl
> > 3) NCL version: 5.2.1
> > 4) uname -a: Linux
> > fe-inform 2.6.18-194.32.1.el5 #1 SMP Wed Jan 5 17:52:25 EST
> > 2011 x86_64 x86_64 x86_64 GNU/Linux
> >
> >
> >
> > MJ
> >
> >
> >
> >
> >
> >
> >
> > _______________________________________________
> > 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 Thu May 26 14:21:28 2011