Re: [gembud] Possible bug in ensemble functionality with precipitation

Keith - Worked!  Even using p06i.  Thanks! Pete



From:   Keith Brill - NOAA Federal <keith.brill@xxxxxxxx>
To:     pmanousos@xxxxxxxxxxxxxxxxxxx
Cc:     gembud@xxxxxxxxxxxxxxxx
Date:   12/10/2012 11:37 AM
Subject:        Re: Possible bug in ensemble functionality with 
precipitation



Hi, Pete,

Try summing as follows:  ens_ssum(mul(gt(p03i,0),p03i)).  This will keep 
any negative values out of the summation.

Keith

On Mon, Dec 10, 2012 at 4:23 PM, <pmanousos@xxxxxxxxxxxxxxxxxxx> wrote:
Greetings.  We believe we may have found a bug when using the cumulative 
precipitation functionality with ensemble functions in GEMPAK gridded 
programs. 

We are wondering if this is a known bug and if not wanted to bring it to 
the user communities attention.  Any thoughts welcome.  Details below. 

Pete 

Reference the Attached Images 
Values plotted are scaled ("scale=2") QPF from the 03Z SREF Dec 10 2012 
run using gdplot3 for "gdfile = {sref | 03}" 

1. 2012121003f24_sref_p03i_ens_ssum.gif 
Sum of all 21 members 3h QPF predictions at f24 using 
"gdpfun=ens_ssum(p03i)".  Note the values of "339" in near Philadelphia 
and "5" near Cincinnati.   

2. 2012121003f27_sref_p03i_ens_ssum.gif 
Same as #1, but for f27.  Note the values of "171" in near Philadelphia 
and "2" near Cincinnati 

3. 2012121003f27_sref_p06i_ens_ssum.gif 
One would expect a 6hr total of 510 at Phili and 7 at Cinci.  However 
using "gdpfun=ens_ssum(p06i)" at f27 yields negative values in Cinci (but 
expected values at Phili). 

Our concern is that with light precipitation events (especially if 
converted via precip type - say like freezing rain) the functionality as 
invoked above will yield erroneous results. 

Additional Findings and Information 
Using GEMPAK 6.4.0 on x64 system running RHEL 6.2 
Confirmed error also occurs with GEMPAK 6.2.0 gempak gridded programs 
Confirmed error occurs for either a p06i or p06m 
Confirmed error occurs for any given SREF run 
Confirmed each member being referenced by the ensemble function "ens_ssum" 
has valid data at each grid point 
Confirmed error does NOT occur when applied to single members (models). 
 Example - no such error occurs when "GDFILE = nam | 12" and "gdpfun = 
p06i" for any given fhr) 
All members were pulled in from NOMADS and processed using NAGRIB2 
By the way we were using the "ens_ssum" to verify the wierdo results we 
found using ens_savg -- which is how we found this bug

----------------------------------------- The information contained in 
this message is intended only for the personal and confidential use of the 
recipient(s) named above. If the reader of this message is not the 
intended recipient or an agent responsible for delivering it to the 
intended recipient, you are hereby notified that you have received this 
document in error and that any review, dissemination, distribution, or 
copying of this message is strictly prohibited. If you have received this 
communication in error, please notify us immediately, and delete the 
original message.