search

ESCOMP / CAM-SIMA

Community Atmosphere Model - System for Integrated Modeling of the Atmosphere
3 stars 12 forks source link

Eliminate requirement of always having specific humidity #208

Open nusbaume opened 1 year ago

nusbaume commented 1 year ago

What is the feature/what would you like to discuss?

There is a requirement in CAM, and for the moment CAMDEN, that specific humidity always be present as an advected constituent. However, for certain scenarios (such as a dry physics suite or an exoplanet simulation) there is no reason to have specific humidity, and thus requiring the tracer increases the computational cost and complexity of those simulations.

Thus it would be beneficial, and improve model flexibility, if specific humidity was treated like all other advected constituents, and only included when a physics scheme required it.

Is there anyone in particular you want to be part of this conversation?

No response

Will this change (regression test) answers?

I Don't Know

Will you be implementing this enhancement yourself?

Any Software Engineer can do this

gold2718 commented 1 year ago

I may be missing the point of this issue but has anyone checked to see if our dycores work with pcnst == 0? What happens if you try to allocate an array where one of the dimensions is pcnst?

nusbaume commented 1 year ago

@gold2718 Those are all great points that we are planning to examine whenever we get to this issue (which might be awhile as it isn't that high of a priority). However, even if in the end we can only partially remove it that still might be an improvement in terms of memory savings, or at least that's the hope!