[Met_help] [rt.rap.ucar.edu #71901] History for SAL query

[Tressa Fowler via RT]

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

Hi John
I was came across SAL (Structure, amplitude n location) idea of evaluating the forecasts. pertaining to this, there is confusion in my understanding. 
1. i read that It answers the question "is domain averaged PPTN correctly FCST"??. this is what is called as the AMPLITUDE. So what is the difference between upscaling and calculation of amplitude (A) in SAL???2. Secondly, cam MET compute the A??? (Amplitude) of SAL???. 
3. what actually is computed after computing the OBS and FCST areal averages. is is the correlation between the two values????Pls clarifyGeeta 		 	   		  

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

Subject: SAL query
From: Tressa Fowler
Time: Thu May 14 13:14:09 2015

Hi Geeta,

MET does not compute the SAL components. We have a variable name that
looks similar, but it contains the L1 and L2 norm values for the
scalar anomaly (SAL1L2).

The Amplitude is really just the magnitude bias scaled by the average
climatology of the two domains. If this value is your primary
interest, simply look at the overall magnitude bias value produced by
MET.

The Wernli paper assumes the two domains are on the same grid and SAL
is an object based way to assess the differences between the two
fields. My understanding of upscaling is that it is primarily a way to
compare data from different grids or to create a grid from point
observation data. Perhaps there is a specific case where they are the
same, but I am not experienced with use of SAL or with upscaling, so
this may not be correct.

I hope my answer was helpful. Let me know if you have further
questions.

Tressa

On Tue May 12 10:51:31 2015, geeta124 at hotmail.com wrote:
> Hi John
> I was came across SAL (Structure, amplitude n location) idea of
>    evaluating the forecasts. pertaining to this, there is confusion
in
>    my understanding.
> 1. i read that It answers the question "is domain averaged PPTN
>    correctly FCST"??. this is what is called as the AMPLITUDE. So
what
>    is the difference between upscaling and calculation of amplitude
>    (A) in SAL???2. Secondly, cam MET compute the A??? (Amplitude) of
>    SAL???.
> 3. what actually is computed after computing the OBS and FCST areal
>    averages. is is the correlation between the two values????Pls
>    clarifyGeeta


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