PRECIS support forum

It is currently Thu Mar 28, 2024 10:57 pm

All times are UTC




Post new topic Reply to topic  [ 5 posts ] 
Author Message
PostPosted: Wed Jun 29, 2016 11:12 am 
Offline

Joined: Tue Jun 28, 2016 9:36 am
Posts: 4
Dear all,

Running Precis Model for a simulation period from 1990 to 2010 using ERAEIN reanalysis data set, we encountered with many values of rain rate that have negative values ???

Is that possible, and if yes that is the case or the reason of that??


Top
 Profile  
 
PostPosted: Wed Jun 29, 2016 12:34 pm 
Offline

Joined: Thu Feb 06, 2014 12:29 pm
Posts: 118
Hi Shereef,
You emailed me some sample files separately, and you are right that there are negative values in the precipitation. The table below uses the tool ppfile -s to list information about each field (each file is a monthly mean and contains just one field).

Code:
bash-4.1$ ppfile -s prebba.pmj*           
Input files/directories:
/net/home/h03/hadhd/prebba.pmj0aug.05216.pp
/net/home/h03/hadhd/prebba.pmj1aug.05216.pp
/net/home/h03/hadhd/prebba.pmj2aug.05216.pp
/net/home/h03/hadhd/prebba.pmj3aug.05216.pp
/net/home/h03/hadhd/prebba.pmj4aug.05216.pp
/net/home/h03/hadhd/prebba.pmj5aug.05216.pp
/net/home/h03/hadhd/prebba.pmj6aug.05216.pp
/net/home/h03/hadhd/prebba.pmj7aug.05216.pp

 Start/Valid     End/Data
   Y M D h m    Y M D h m STASH   PROC  EXPT COLS ROWS LEVEL           Min          Mean           Max Valid
199008010000 199009010000  5216    128 prebb  124  102  8888 -1.607407e-14  2.092921e-06  0.0001012639 12648/12648
199108010000 199109010000  5216    128 prebb  124  102  8888 -2.664777e-15  6.302661e-06  0.0001549033 12648/12648
199208010000 199209010000  5216    128 prebb  124  102  8888 -1.989106e-14   9.62439e-06   0.000206182 12648/12648
199308010000 199309010000  5216    128 prebb  124  102  8888 -3.593162e-14  6.222757e-06  0.0001454181 12648/12648
199408010000 199409010000  5216    128 prebb  124  102  8888 -7.612366e-15  6.789779e-06  0.0001797132 12648/12648
199508010000 199509010000  5216    128 prebb  124  102  8888  -1.81756e-14  5.315126e-06  0.0002481709 12648/12648
199608010000 199609010000  5216    128 prebb  124  102  8888 -3.521987e-14  5.900335e-06  0.0001741567 12648/12648
199708010000 199709010000  5216    128 prebb  124  102  8888 -1.624471e-14  7.833532e-06  0.0001713253 12648/12648
   Y M D h m    Y M D h m STASH   PROC  EXPT COLS ROWS LEVEL           Min          Mean           Max Valid


What you see is that although there are negatives, there are extremely tiny values. The reason this happens has to do with how computers solve the complex equations within climate models. I won't get into the details. In any case it's nothing to be concerned about -- it's normal behaviour.

The good news is that there is a way to set all values below zero to be zero. As a matter of fact, it's advisable to do this for the extremely tiny positive values also, because essentially those should be treated as zero as well.

In climate science this is called setting a wet day threshold. One tenth of a millimetre is often used, or half a millimetre.

In PRECIS you can use the tool ppwhere to apply a wet day threshold.

But first, you should carry out at least two necessary processing tasks on your raw PRECIS output data. First to remove the outer 8 point rim (because data in this region is not reliable as this is where the driving GCM and RCM come into agreement) and second to convert the unit (PRECIS uses only SI units, so precipitation is measured in millimetres per second, which is not useful. Millimetres per day is better -- there are 86400 second in day).

Let's say your precipitation data is found in $ARCHIVEDIR/prebb/05216 (05216 is the stash code for precipitation).

Code:
cd $ARCHIVEDIR/prebb/05216

# Remove the Rim
mkdir RIMREMOVE
pprr -r 8  -e RIMREMOVE/   *.pp

# Convert to mm/day
mkdir MMDAY
ppexpr -x "a*86400.0" -e MMDAY a=RIMREMOVE/*.pp

# Set all values below 0.1mm to be zero
mkdir WETDAY
ppwhere -e WETDAY/  -r '0.0'  -w 'x>=0.1'  MMDAY/*.pp



The files found in $ARCHIVEDIR/prebb/05216/WETDAY are rim-removed, in mm/day and have a threshold applied such that all values below 0.1 are set to zero. Try it. :P


Here's the result on the files you sent. See how the numbers have changed, and now the minimum is 0.0?

Code:
bash-4.1$ ppfile -s prebb*.pp
Input files/directories:
/net/home/h03/hadhd/PREBB/WETDAY/prebba.pmj0aug.05216.rr8.expr.whe.pp
/net/home/h03/hadhd/PREBB/WETDAY/prebba.pmj1aug.05216.rr8.expr.whe.pp
/net/home/h03/hadhd/PREBB/WETDAY/prebba.pmj2aug.05216.rr8.expr.whe.pp
/net/home/h03/hadhd/PREBB/WETDAY/prebba.pmj3aug.05216.rr8.expr.whe.pp
/net/home/h03/hadhd/PREBB/WETDAY/prebba.pmj4aug.05216.rr8.expr.whe.pp
/net/home/h03/hadhd/PREBB/WETDAY/prebba.pmj5aug.05216.rr8.expr.whe.pp
/net/home/h03/hadhd/PREBB/WETDAY/prebba.pmj6aug.05216.rr8.expr.whe.pp
/net/home/h03/hadhd/PREBB/WETDAY/prebba.pmj7aug.05216.rr8.expr.whe.pp

 Start/Valid     End/Data
   Y M D h m    Y M D h m STASH   PROC  EXPT COLS ROWS LEVEL           Min          Mean           Max Valid
199008010000 199009010000  5216    128 prebb  108   86  8888             0     0.1185155      7.854496 9288/9288
199108010000 199109010000  5216    128 prebb  108   86  8888             0     0.4522313      11.60172 9288/9288
199208010000 199209010000  5216    128 prebb  108   86  8888             0     0.7149138       17.2667 9288/9288
199308010000 199309010000  5216    128 prebb  108   86  8888             0      0.444399       11.1208 9288/9288
199408010000 199409010000  5216    128 prebb  108   86  8888             0     0.4285355      11.26194 9288/9288
199508010000 199509010000  5216    128 prebb  108   86  8888             0     0.2901314      17.06261 9288/9288
199608010000 199609010000  5216    128 prebb  108   86  8888             0     0.3947067      11.69971 9288/9288
199708010000 199709010000  5216    128 prebb  108   86  8888             0     0.5617172       14.8025 9288/9288
   Y M D h m    Y M D h m STASH   PROC  EXPT COLS ROWS LEVEL           Min          Mean           Max Valid
Fields: 8


Hope this helps
David


Top
 Profile  
 
PostPosted: Sun Jul 03, 2016 7:28 am 
Offline

Joined: Tue Jun 28, 2016 9:36 am
Posts: 4
Thank you very much, this really help


Top
 Profile  
 
PostPosted: Sun Jul 03, 2016 8:15 am 
Offline

Joined: Tue Jun 28, 2016 9:36 am
Posts: 4
another questions please, What is the difference between pa and pm ?? they both seems to be monthly average !?


Top
 Profile  
 
PostPosted: Thu Jul 28, 2016 12:24 pm 
Offline

Joined: Thu Feb 06, 2014 12:29 pm
Posts: 118
Hi Shereef,

*a.pa* are daily mean files. The files contain daily means for all the values in that model month.
*a.pm* are monthly means. There is a single field in the file.
You will see that the *a.pa* are ~30 times larger in size that *a.pm*


David


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 5 posts ] 

All times are UTC


Who is online

Users browsing this forum: No registered users and 49 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group