CAM3.1 fails to add a new aerosol interface

[dckim@mit_edu]

Hi,

Somehow CAM3.1 fails with the aerosol interface. My experiment was simple. First I tested CAM3.1 with AERO_CARBON flag OFF (false) which was successful. To test carbon interface module, I set AERO_CARBON flag ON (true) and removed all "call endrun" lines in the carbon_intr.F90 which is a way worked with CAM3.0

The model stops without an error message but with a large core file. The error occured at the subroutine shr_sys_flush under 'cam1/models/csm_share/shr/shr_sys_mod.F90'.

Have you met this problem before? I am using 64 bit Linux with single processor and pgf90 compiler.

Dongchul

 

[mvr]

hi...

as you probably already know, the use of prognostic carbon
aerosols in cam is in an experimental stage...using endrun
calls in carbon_intr.F90 in cam3.0 and cam3.1 was meant to
remind users who set the namelist variable aero_carbon to true,
that this code was under development and not available with the
release...

when you say you removed all endrun calls from carbon_intr.F90,
can i assume these were replaced with your own code for
prognostic carbon aerosols?...if so, your problem may be the
result of a change to the parameter list for the subroutine
carbon_wet_intr which occurred between versions 3.0 and 3.1...

i hope that helps,
-mathew