[Met_help] [rt.rap.ucar.edu #62410] History for Support for MODE

[Tressa Fowler via RT]

----------------------------------------------------------------
  Initial Request
----------------------------------------------------------------

Hello,

I am trying to compare precipitation skill from two different models using MODE, grid_stat, and wavelet_stat.  My models, however, are of different grid resolutions (one is WRF and one is ECMWF).  I was wondering if you had any ideas and/or comments on the quality of the results when evaluating two models of different resolutions?  In grid_stat and wavelet_stat, the skill scores typically take the scale into account, so it is not so much of an issue.  However, in MODE, I am struggling with the issue of interpreting results given the fact that the 'observation' objects are different between the models (due to the fact that conv_radius is dependent on the grid_resolution).  Of course, I could manually set this parameter to a fixed value, but I am not sure if that makes sense.  And if I did set it to a fixed value, if this value should be at the resolution of the ECMWF or the WRF.  Would you have any comments or ideas about how to achieve meaningful
 results while using MODE to compare models of different resolutions?

Thank you!

- Andrew



________________________________
 Von: John Halley Gotway via RT <met_help at ucar.edu>
An: andrewwx at yahoo.com 
Gesendet: 21:52 Freitag, 14.Juni 2013
Betreff: [rt.rap.ucar.edu #61828] Resolved: Support for PCP_Combine
 

According to our records, your request has been resolved. If you have any
further questions or concerns, please respond to this message.

----------------------------------------------------------------
  Complete Ticket History
----------------------------------------------------------------

Subject: Support for MODE
From: Tressa Fowler
Time: Tue Jul 30 10:18:34 2013

Hi Andrew,

Comparison across resolutions is tricky and there are no answers, only
suggestions. I would suggest setting MODE to use a convolution radius
that turns out to be about equal distance in both models. You can also
set up MODE so that observed object identification is independent of
model objects, thus ensuring the same object field for comparison of
both models. (Review the MODE configuration file documentation for
details.)


Please let us know if you need further assistance.

Tressa

On Mon Jul 29 02:58:21 2013, andrewwx at yahoo.com wrote:
> Hello,
>
> I am trying to compare precipitation skill from two different models
>    using MODE, grid_stat, and wavelet_stat.  My models, however, are
>    of different grid resolutions (one is WRF and one is ECMWF).  I
was
>    wondering if you had any ideas and/or comments on the quality of
>    the results when evaluating two models of different resolutions? 
>    In grid_stat and wavelet_stat, the skill scores typically take
the
>    scale into account, so it is not so much of an issue.  However,
in
>    MODE, I am struggling with the issue of interpreting results
given
>    the fact that the 'observation' objects are different between the
>    models (due to the fact that conv_radius is dependent on the
>    grid_resolution).  Of course, I could manually set this parameter
>    to a fixed value, but I am not sure if that makes sense.  And if
I
>    did set it to a fixed value, if this value should be at the
>    resolution of the ECMWF or the WRF.  Would you have any comments
or
>    ideas about how to achieve meaningful
>  results while using MODE to compare models of different
resolutions?
>
> Thank you!
>
> - Andrew
>
>
>
> ________________________________
>  Von: John Halley Gotway via RT <met_help at ucar.edu>
> An: andrewwx at yahoo.com
> Gesendet: 21:52 Freitag, 14.Juni 2013
> Betreff: [rt.rap.ucar.edu #61828] Resolved: Support for PCP_Combine
>
>
> According to our records, your request has been resolved. If you
have
>    any
> further questions or concerns, please respond to this message.



------------------------------------------------