Blowing up on first time step

Discussion on computers, ROMS installation and compiling

Moderators: arango, robertson

Post Reply
Message
Author
rgawde
Posts: 11
Joined: Sat Oct 10, 2015 1:04 am
Location: UMCES Horn Point Lab

Blowing up on first time step

#1 Unread post by rgawde »

Hello everyone,

I have been running the ROMS version 3.5 on a Linux x86_64 server using ifort as the compiler with no biological modules. But now I have defined the basic BIO_FENNEL and other related modules and the model blows up on the very first time step. I tried to reduce the time step but that doesn't help either. I am attaching the log file I get with the run along with the '.h' and '.in' files below.

Could anyone please help?

Thanks,
Rasika
Attachments
myrun.rtf
(77.35 KiB) Downloaded 241 times
choproms.in
(92.51 KiB) Downloaded 246 times
choproms.h
(10.66 KiB) Downloaded 250 times

User avatar
kate
Posts: 4088
Joined: Wed Jul 02, 2003 5:29 pm
Location: CFOS/UAF, USA

Re: Blowing up on first time step

#2 Unread post by kate »

Well, it's weird. There aren't that many ways for the ecosystem models to feed back to the physics and cause a blowup like that. I would look for compiler options to force a blowup at the first sign of NaN or Inf. Try compiling with USE_DEBUG so you'll get line numbers when it blows up.

Post Reply