[RegCNET] Trouble in running MIT-Emanuel scheme

mega octaviani gorgeous_egha at yahoo.com
Thu Oct 11 16:58:19 CEST 2007


Dear Dr. Gao and others,

It seems i wrote one more fault in my first email, i.e
the timestep trial, the second trial that i have tried
was reducing dt, and not increasing it, sorry again
for this blunder typing.

And about the possibility of icbc error, i also have
done another trial. Maybe first, i would like to
describe all conditions that i have now:

I have two PCs, one running RegCM3 using Grell's
scheme and another is initially planned to be run for
MIT-Emanuel's scheme simulation.
Both PC have three external hard drives (300GB for
each). The first is for input data, second is for
ICBC, and third is for the output.

Till now, I run the Grell's scheme successfully and it
can continue to the second year. And since both of my
simulation have similar icbc information, so when i
have this problem in the 2nd PC, i tried to copy all
terrain and icbc output from 1st PC (to avoid the
possibility that the first HDD in my 2nd PC might have
trouble in input data storage). Hence, after rerun it,
i still have the same error message.

If it is because the hard disk capacity is not enough
just to run for one year (with 1 hourly output), how
many times larger for the output will it be compare to
the Grell's simulation output (fyi, using Grell with
other setups stay similar, it needs 64 GB for 1 yr
output)?

Thank you again.
Kindly regards,

Mega

--- Gao Xuejie <gaoxj at cma.gov.cn> wrote:

> Hi Mega,
> 
> When the model becomes unstable, you should decrease
> the dt, not increase. Try dt=150, or 120.
> 
> But from you error message of "forrtl: severe (174):
> SIGSEGV", maybe the ICBC file has some problem? Or
> you do not have enough hard disk space?
> 
> Good luck.
> 
> Gao
> 
> ----- Original Message ----- 
> From: "mega octaviani" <gorgeous_egha at yahoo.com>
> To: <regcnet at lists.ictp.it>
> Sent: Thursday, October 11, 2007 11:44 AM
> Subject: Re: [RegCNET] Trouble in running
> MIT-Emanuel scheme
> 
> 
> > Oups, 
> > i am sorry,
> > I have given you the wrong regcm.in
> > That is for my other simulation.
> > the true regcm.in that i run for mit-emanuel
> scheme is
> > below: 
> > 
> > $restartparam
> > ifrest  = .false. ,
> > idate0  = 1997010100,
> > idate1  = 1997010100,
> > idate2  = 1997011100,
> > nslice  =        120,
> > $
> > $timeparam
> > radfrq  =    30.,
> > abemh   =    18.,
> > abatm   =   600.,
> > dt      =   200.,
> > ibdyfrq =          6,
> > $
> > $outparam
> > ifsave  = .true. ,
> >   savfrq  =    24.,
> > iftape  = .true. ,
> >   tapfrq  =     1.,
> > ifrad   = .true. ,
> >   radisp  =     1.,
> > ifbat   = .true. ,
> > ifsub   = .true. ,
> >   batfrq  =     1.,
> > ifprt   = .false.,
> >   prtfrq  =    12.,
> >   kxout   =          6,
> >   jxsex   =         40,
> > iotyp   =          1,
> > ibintyp =          1,
> > ifchem  = .false.,
> >   chemfrq =     6.,
> > $
> > $physicsparam
> > iboudy  =          5,
> > ibltyp  =          1,
> > icup    = 4,
> >   igcc    = 1,
> > ipptls  =          1,
> > iocnflx = 2,
> > ipgf    =          0,
> > lakemod =          0,
> > ichem   =          0,
> > $
> > $subexparam
> > $
> > $grellparam
> > $
> > $emanparam
> > $
> >  $chemparam
> > idirect   =    1,
> > chtrname  = '^@^@^@^@^@',
> > chtrsol   =  0.00,
> > chtrdpv   = 0.00000,0.00000,
> > dustbsiz  =  0.00, 0.00,
> > $
> > Thank you for your respond
> > 
> > Regards,
> > Mega
> > 
> > 
> > --- Shahbaz Mehmood <shahbazmehmood at gmail.com>
> wrote:
> > 
> >> Hi
> >> 
> >> In the regcm.in attached you have set icup=2
> while
> >> for Emanuel scheme you
> >> need to set icup=4.
> >> 
> >> Best regards.
> >> shahbaz
> >> 
> >> Shahbaz Mehmood
> >> Research Fellow (Climate Section)
> >> Global Change Impact Studies Centre
> >> 1st Floor, Saudi Pak Tower
> >> 61/A, Jinnah Avenue
> >> Islamabad, Pakistan
> >> Ph#   +92-51-9219785
> >> Cell#  +92-333-5481767
> >> Fax#  +92-51-9219787
> >> 
> >> On 10/11/07, mega octaviani
> >> <gorgeous_egha at yahoo.com> wrote:
> >> >
> >> > Dear all,
> >> >
> >> > I have tried to run a 60-km domain simulation
> (for
> >> > South East Asia region) using MIT-Emanuel
> scheme
> >> with
> >> > the following set in regcm.in:
> >> >
> >> > $restartparam
> >> > ifrest  = .false. ,
> >> > idate0  = 1997010100,
> >> > idate1  = 1997010100,
> >> > idate2  = 1997011100,
> >> > nslice  =        120,
> >> > $
> >> > $timeparam
> >> > radfrq  =    30.,
> >> > abemh   =    18.,
> >> > abatm   =   600.,
> >> > dt      =   200.,
> >> > ibdyfrq =          6,
> >> > $
> >> > $outparam
> >> > ifsave  = .true. ,
> >> >   savfrq  =    24.,
> >> > iftape  = .true. ,
> >> >   tapfrq  =     1.,
> >> > ifrad   = .true. ,
> >> >   radisp  =     1.,
> >> > ifbat   = .true. ,
> >> > ifsub   = .true. ,
> >> >   batfrq  =     1.,
> >> > ifprt   = .false.,
> >> >   prtfrq  =    12.,
> >> >   kxout   =          6,
> >> >   jxsex   =         40,
> >> > iotyp   =          1,
> >> > ibintyp =          1,
> >> > ifchem  = .false.,
> >> >   chemfrq =     6.,
> >> > $
> >> > $physicsparam
> >> > iboudy  =          5,
> >> > ibltyp  =          1,
> >> > icup    = 2,
> >> >   igcc    = 1,
> >> > ipptls  =          1,
> >> > iocnflx = 2,
> >> > ipgf    =          0,
> >> > lakemod =          0,
> >> > ichem   =          0,
> >> > $
> >> > $subexparam
> >> > $
> >> > $grellparam
> >> > $
> >> > $emanparam
> >> > $
> >> > $chemparam
> >> > idirect   =    1,
> >> > chtrname  = '^@^@^@^@^@',
> >> > chtrsol   =  0.00,
> >> > chtrdpv   = 0.00000,0.00000,
> >> > dustbsiz  =  0.00, 0.00,
> >> > $
> >> >
> >> > I got the program crashed at the 6th day of my
> >> > simulation with following error output:
> >> >
> >> > forrtl: severe (174): SIGSEGV, segmentation
> fault
> >> > occurred
> >> > Image     PC     Routine    Line     Source
> >> > regcm  0805C8B4  Unknown    Unknown  Unknown
> >> > regcm  0805BEF2  Unknown    Unknown  Unknown
> >> > regcm  080ACEF4  Unknown    Unknown  Unknown
> >> > regcm  080A644B  Unknown    Unknown  Unknown
> >> > regcm  0804959A  Unknown    Unknown  Unknown
> >> > libc.so.6  00714E23  Unknown   Unknown  Unknown
> >> > regcm  080494D1  Unknown    Unknown  Unknown
> >> >
> >> > Then, i rerun again by increasing timestep 20%
> 
=== message truncated ===



       
____________________________________________________________________________________
Take the Internet to Go: Yahoo!Go puts the Internet in your pocket: mail, news, photos & more. 
http://mobile.yahoo.com/go?refer=1GNXIC



More information about the RegCNET mailing list