Trcrad: Bug in ocean.output regarding tracer conservation

Good morning,

We noticed a small bug regarding the ocean.output lines written by the trcrad routine (src/TOP/TRP/trcrad.F90 · main · NEMO Workspace / Nemo · GitLab). I’m reporting it here before opening an issue on the gitlab as requested.

In line 105:
IF( ln_trcrad ) THEN ; WRITE(numout,*) ' ===>> ensure the global tracer conservation',

the model writes that global tracer conservation is ensured when using trcrad. However, this is only the case if the PISCES biogeochemical model is used, because in

SUBROUTINE trc_rad_sms( kt, Kbb, Kmm, ptr, jp_sms0, jp_sms1, cpreserv )',

cpreserv, which activates the conservative correction is only used with PISCES.
IF( ln_pisces ) CALL trc_rad_sms( kt, Kbb, Kmm, ptr, jp_pcs0, jp_pcs1, cpreserv='Y' ) ! PISCES model
IF( ln_my_trc ) CALL trc_rad_sms( kt, Kbb, Kmm, ptr, jp_myt0, jp_myt1 ) ! MY_TRC model

Note also, that setting negative concentrations exactly to zero as done by the default trcrad routine causes our biogeochemical model to produce NaNs, which is why I use a slightly modified version with a small number instead (e-12). Thanks!