[MITgcm-support] flux limited advection scheme

[Samar Khatiwala]

Hi

I am running an ecosystem model that generates large spatial  
gradients. To avoid negative values, I use the flux limited DST3
scheme (#33). Unfortunately, I still get negative values, and fairly  
largish ones at that. I am puzzled by this because I thought
this was a positive scheme (although, this is apparently true only if  
some conditions on the courant number are met). One additional
complication is that I am using implicit vertical advection. (In the  
non flux limited case (scheme #30), this does help with spurious
negatives). Is there some strange interaction between scheme #33 (or  
any of the other flux limited advection schemes, e.g., 77) and
implicit vertical advection I should be aware of? Can the two be used  
together?

Thanks, Samar