Running CLIMBER-X with closed atmosphere-ocean carbon cycle (ctl.l_lnd_co2=F) but including carbon isotopes (just looking at C13 below / ctl.l_c13=T, bgc.l_c13=T), I am getting ncio errors at write time. As far as I looked, the variables that are becoming unrepresentable are usually d13C_lnd, d13Cflx_burial, or d13Cflx_atm_lnd.
d13C_lnd:
I suppose this happens because of this division
./co2/co2_out.f90: ann_ts(y)%d13C_lnd = ((co2%dC13lnd_dt/co2%dClnd_dt)/c13_c12_std - 1._wp) * 1000._wp
although numerator and denominator are zero for l_lnd_co2=F:
./co2/co2_model.f90:
! option to exclude carbon flux from lnd
if (.not.l_lnd_co2) then
co2%dClnd_dt = 0._wp
co2%dC13lnd_dt = 0._wp
co2%dC14lnd_dt = 0._wp
endif
Maybe it's easiest to exclude d13C_lnd (or d13C_ocn) in the co2_diag-subroutine when l_lnd_co2=F (or l_ocn_co2=F)?
d13Cflx_burial and d13Cflx_atm_lnd:
I'm not sure where the large numbers could originate from here, since for lnd.l_burial=F / Cflx_burial=0, the respective lines in lnd_diag with the division should not be reached (line 1100 in ./lnd/lnd_out.f90; or line 1090 for C13flx_atm_lnd).
But, I was wondering if it's okay that the burial fluxes for l_burial=F are moved into C(13)flx_atm_lnd, although ctl.l_lnd_co2=F?
There is way too much C13 in the atmosphere when the model finally stops with that ncio error...
Running CLIMBER-X with closed atmosphere-ocean carbon cycle (ctl.l_lnd_co2=F) but including carbon isotopes (just looking at C13 below / ctl.l_c13=T, bgc.l_c13=T), I am getting ncio errors at write time. As far as I looked, the variables that are becoming unrepresentable are usually d13C_lnd, d13Cflx_burial, or d13Cflx_atm_lnd.
d13C_lnd:
I suppose this happens because of this division
./co2/co2_out.f90: ann_ts(y)%d13C_lnd = ((co2%dC13lnd_dt/co2%dClnd_dt)/c13_c12_std - 1._wp) * 1000._wp
although numerator and denominator are zero for l_lnd_co2=F:
./co2/co2_model.f90:
! option to exclude carbon flux from lnd
if (.not.l_lnd_co2) then
co2%dClnd_dt = 0._wp
co2%dC13lnd_dt = 0._wp
co2%dC14lnd_dt = 0._wp
endif
Maybe it's easiest to exclude d13C_lnd (or d13C_ocn) in the co2_diag-subroutine when l_lnd_co2=F (or l_ocn_co2=F)?
d13Cflx_burial and d13Cflx_atm_lnd:
I'm not sure where the large numbers could originate from here, since for lnd.l_burial=F / Cflx_burial=0, the respective lines in lnd_diag with the division should not be reached (line 1100 in ./lnd/lnd_out.f90; or line 1090 for C13flx_atm_lnd).
But, I was wondering if it's okay that the burial fluxes for l_burial=F are moved into C(13)flx_atm_lnd, although ctl.l_lnd_co2=F?
There is way too much C13 in the atmosphere when the model finally stops with that ncio error...