MEGAN preprocessor - Segmentation fault problem

[Mauro Morichetti]

Hi all,

I am facing problems using MEGAN preprocessor. 

I am working on two polar domains, they are quite big, but with really low resolution (d01 150x150 km, d02 50x50 km grid).

When I run MEGAN, as soon as the preprocessor starts, I have the error message: 

"Program received signal SIGSEGV: Segmentation fault - invalid memory reference."   

This is really wired to me, because I used the same domain (in a different time interval) with no problem.

Could you try for me, if you face the same problem?

In the following link the two wrfinput:

https://drive.google.com/drive/folders/1-KiAm-J4C3hYyia-YbMGR3xWvQqu2b8j?usp=sharing

Thanks

Mauro

[Stacy Walters]

Mauro,

Could you please send me the input file you used with megan_bio_emiss as an email attachment?

Stacy

[Mauro Morichetti]

Hi,

attached the input file.

Thanks

[Stacy Walters]

Mauro,

I have been able to successfully run megan_bio_emis with your bio_emiss.inp file.

I ran on a MacBook Pro.  I used gfortran version 10.1.  The executable had explicit

floating point and memory addressing error trapping turned on.

I assume your segmentation fault error occurred in the first month.  My test case only

processed the first two months.

Stacy

[Mauro Morichetti]

Hi Stacy,

thanks for your answer.

I am working with a supercomputer with CentOS Linux and gcc 8.3.0, but I tried also on Cheyenne, the result doesn't change.

Yes, the segmentation fault occurs as soon as the executable starts (attached the out file).

Do you think could be a compiler version? Do you have any suggest?

Thanks again

Mauro

[Mauro Morichetti]

Hi all again, 

with Stacy help, I was able to run successful megan preprocessor, but now I have another problem.

The wrfbiochemi* files created, have different domain and projection from the original wrfinput*.

I have done some tests on Cheyenne and the folder is the following:

/glade/work/maurom/MEGAN/megan

in the foldere you can find the wrfinput and the wrfbiochemi created.

Could you please check what I have done wrong?

Thanks

Mauro

Il giorno venerdì 1 ottobre 2021 alle 10:09:01 UTC+2 Mauro Morichetti ha scritto:

the attached.

———

Ing. Mauro Morichetti
MRCMRA84P20L366Y 
C/da Vanni n. 13 62010, Mogliano (MC) 
+39 388 4729670
morichet...@gmail.com

[Xiao Yan]

Hi Mauro

I am also facing the same problem.  
I am also working on polar domains (d01 180x180 km, d02 60x60 km grid).

How did you deal with this problem. 

Looking forward to your reply!

[Mauro Morichetti]

Hi Yan,

I think the problem is simple that the pre-processor run out of memory.

You should do:

1) set in your environment the stack size unlimited;

2) run the main domain and than the nest domain (you have to modify or link the wrfinput_d02 in wrfinput_d01)

Hope this help

Mauro