#2537 closed Defect (fixed)
Using salinity array as a passive tracer
Reported by: | jchanut | Owned by: | systeam |
---|---|---|---|
Priority: | low | Milestone: | |
Component: | TRA | Version: | trunk |
Severity: | minor | Keywords: | |
Cc: |
Description
Context
Test cases often require the use of a passive tracer to ease diagnostics. Since most of the time a simplified (temperature dependent) equation of state is used, it would be convenient to use salinity as a passive tracer (that could be zero, and eventually slightly negative where monotone schemes fail).
Analysis
There are several instances in the code that check or correct if salinity is <=0:
- A source term is added in trazdf to ensure S>=0.1. That's brutal.
- The model stops in stpctl if S<=0
Recommendation
Remove those constrains if a temperature dependent, simplified equation of state is used, i.e.:
ln_seos.AND.(rn_lambda2==0._wp)
Commit History (4)
Changeset | Author | Time | ChangeLog |
---|---|---|---|
14135 | jchanut | 2020-12-08T15:03:43+01:00 | report change #2537 in ISOMIP+ (eosbn2 is clearly outdated) - SETTE tests all ok on my side now |
14131 | jchanut | 2020-12-08T13:09:14+01:00 | #2537, repais rcompilation issue induced by previous commit |
14128 | jchanut | 2020-12-08T09:09:09+01:00 | #2537, remove local changes in /tests/*/MY_SRC directories if salinity is used as a passive tracer |
14127 | jchanut | 2020-12-08T09:03:03+01:00 | #2537, remove checks/corrections if salinity is a passive tracer |
Change History (6)
comment:1 Changed 3 years ago by smasson
comment:2 Changed 3 years ago by jchanut
In 14127:
comment:3 Changed 3 years ago by jchanut
In 14128:
comment:4 Changed 3 years ago by jchanut
- Resolution set to fixed
- Status changed from new to closed
- Version set to trunk
comment:5 Changed 3 years ago by jchanut
In 14131:
comment:6 Changed 3 years ago by jchanut
In 14135:
Yes of course, great idea!
I face this problem in CANAL since years, copying and maintaining a stpctl in CANAL/MY_SRC...
I should have thought about it... ;-)