[gradsusr] NAM and APCPSFC data

Fri Oct 14 15:40:16 EDT 2016

Mark,
Below is the bit of code I use to ’normalize' precip accumulations from the GFS for the meteograms. (I write out a new file.) The records alternate between 3hr and 6hr accumulations, so you may have to adapt for your data and your particular needs. The full code source is ftp://cola.gmu.edu/pub/jma/meteograms/meteogram_subset_GDS.gs
—Jennifer
p.s. I also use t=1 as the initialization time and p is alwasy undefined at t=1, but I write it out anyway.

* Write out the precip variables
'set fwrite mysubsetp.dat'
'set t 1'
'd p'
t=2
while (t<=tmax-1)
* these are 3-hourly accumulations
  'set t 't
  'd p'
* these are 6-hourly accumulations
* so we must subtract the previous 3hr totals
  'set t 't+1
  'd p-p(t-1)  '
  t=t+2
endwhile
'disable fwrite’

On Oct 14, 2016, at 1:49 PM, Mark Sponsler <msponsler at COMCAST.NET<mailto:msponsler at COMCAST.NET>> wrote:

Thanks Jeff!  Perfect - that is what I needed.
Let me go play with my script. Bet I made a logic error somewhere.
Thanks again!
Thanks,
Mark
----- Original Message -----
From: Jeff Duda <jeffduda319 at gmail.com<mailto:jeffduda319 at gmail.com>>
To: GrADS Users Forum <gradsusr at gradsusr.org<mailto:gradsusr at gradsusr.org>>
Sent: Fri, 14 Oct 2016 17:32:12 -0000 (UTC)
Subject: Re: [gradsusr] NAM and APCPSFC data
'set t 2'
'd apcpsfc'
should display 1-hr precip accumulation between f00 and f01
'set t 3'
'd apcpsfc'
should display 2-hr precip accumulation between f00 and f02
'set t 4'
'd apcpsfc'
should display 3-hr precip accumulation between f00 and f03
'set t 5'
'd apcpsfc'
should display 1-hr precip accumulation between f03 and f04
...
...
...
and so on.
If you want to display 1-hr precip at "off" hours (not f01, f04, f07, ...) then yes, you need to display the difference over consecutive hours.
f02: 'set t 3' ; 'd apcpsfc - apcpsfc(t-1)'
f03: 'set t 4' ; 'd apcpsfc - apcpsfc(t-1)'
Notice how the pattern is the same despite the particular time. The only exception is at f01, f04, f07, ... when you just display apcpsfc without any subtraction.
You appear to have this idea. Is this not getting you the results you expect?
Jeff

On Fri, Oct 14, 2016 at 12:26 PM, Mark Sponsler <msponsler at comcast.net<mailto:msponsler at comcast.net>> wrote:
Hi Jeff,
Thanks for the info.
Yeah, I use the first time step which of course has no precip data in it. So the cycle would start with t = 2.
Just to confirm, for either the 1 hr NAM or the 3 hr NAM, I simply should use the below logic and all should be good?
In other words, the 1 hr NAM accumulates in 3 time step sequential bundles (timesteps 2,3,4 refresh then 5,6,7, refresh etc) and the 3 hr NAM accumulates the same.
Display Time = 2
Time 3 = Time 3 - Time 2
Display Time 3
Time 4 = Time 4 - Time 3
Display Time 4
Display Time = 5
Time 6 = Time 6 - Time 5
Display Time 6
Time 7 = Time 7 - Time 6
Display Time 7
etc...
Thanks,
Mark
----- Original Message -----
From: Jeff Duda <jeffduda319 at gmail.com<mailto:jeffduda319 at gmail.com>>
To: GrADS Users Forum <gradsusr at gradsusr.org<mailto:gradsusr at gradsusr.org>>
Sent: Fri, 14 Oct 2016 16:59:14 -0000 (UTC)
Subject: Re: [gradsusr] NAM and APCPSFC data
Mark,

Why do you not think it works the way you described in (1)? That's the way I've always used it, and I've never had a problem. The particular t-index value you use in grads depends on how you setup the control file. If you didn't start the time index at forecast hour 0, then there will be a phase difference between what you think and what actually works.

Jeff Duda

On Fri, Oct 14, 2016 at 10:55 AM, Mark Sponsler <msponsler at comcast.net<mailto:msponsler at comcast.net>> wrote:
Hi,
I'm trying to display variable 'apcpsfc' from the 3 hr and 1 hr NAM.  The resulting images appear to pulsate/accumulate precip every few images then reset only to accumulate again.

I've seen the threads in the archive on this issue with HD GFS data, but I'm trying to understand 2 things:

1) What is the accumulation period for NAM 1 hr and 3 hr data.  I was thinking it was every 3 time steps (i.e. excluding time step 1 / 00 hr), it is 2, 3, 4 then reset and start accumulating again for the next 3 time steps 5, 6 and 7).  But that doesnt seem to be the case.

2) What does it mean to 'normalize' the data? Westly did something in wgrib2 for HD GFS data to normalize it and I was wondering if that was also applicable to NAM data.  What is done to mathmatically to get the data to be valid for just a single timestep?.  Can I write something in GRDADS to perform the normalization rather than having to use WGRIB2.

Assuming the accumulation theory is correct (that apcpsfc accumulates every three time steps), I've been trying to write code to remove the accumulation by:

Display Time = 2
Time 3 = Time 3 - Time 2
Display Time 3
Time 4 = Time 4 - Time 3
Display Time 4

But it doesnt work.

Any help would be appreciated.

Thanks,
Mark
Thanks,
Mark

_______________________________________________

gradsusr mailing list
gradsusr at gradsusr.org<mailto:gradsusr at gradsusr.org>
http://gradsusr.org/mailman/listinfo/gradsusr

--
Jeff Duda
Post-doctoral research associate
University of Oklahoma School of Meteorology

_______________________________________________
gradsusr mailing list
gradsusr at gradsusr.org<mailto:gradsusr at gradsusr.org>
http://gradsusr.org/mailman/listinfo/gradsusr

--
Jeff Duda
Post-doctoral research associate
University of Oklahoma School of Meteorology
_______________________________________________
gradsusr mailing list
gradsusr at gradsusr.org<mailto:gradsusr at gradsusr.org>
http://gradsusr.org/mailman/listinfo/gradsusr

--
Jennifer Miletta Adams
Center for Ocean-Land-Atmosphere Studies (COLA)
George Mason University

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://gradsusr.org/pipermail/gradsusr/attachments/20161014/56acb51f/attachment-0001.html 