MARBL ERROR: bounding bracket for pH solution not found

[Jay Zhu]

Hi all,

I am running a CESM2.1.3 cases with BSSP585cmip6_BPRP. The program runs normally until June 2090 when an ERROR occurs, which shows a 'MARBL ERROR: bounding bracket for pH solution not found' in cesm log file. Can someone tell me about this error and how to fix it？

Any help would be greatly appreciated!

Jay

 

[dbailey]

Moving this to the BGC forum.

 

[mlevy]

This is a very common error, and is typically indicative of a CFL violation -- you need to take a smaller time step. Notice in your error message that temperature is being reported as -42.9 degrees; in future simulations, sea ice melts off the coast of Greenland exposing ocean in portions of the gx1v7 with very small grid cells. See this post for slightly more detail, but increasing dt_count will force POP to take smaller time steps and hopefully prevent this instability from cropping up.

 

[Jay Zhu]

This is a very common error, and is typically indicative of a CFL violation -- you need to take a smaller time step. Notice in your error message that temperature is being reported as -42.9 degrees; in future simulations, sea ice melts off the coast of Greenland exposing ocean in portions of the gx1v7 with very small grid cells. See this post for slightly more detail, but increasing dt_count will force POP to take smaller time steps and hopefully prevent this instability from cropping up.

Thank you for your answer. I would like to know whether the change of dt_count will bias the previous results？

 

[mlevy]

It shouldn't -- as long as you restart from a period where the smaller dt_count value was stable, changing to a bigger value should not drastically change the simulation (we consider it answer-changing but not climate-changing).