+Missing Pointers for Benthic Fluxes in Direct Sinking of Phytoplankton P

[yichengt900]

As @charliestock suggested, now the pointers for the benthic fluxes for the direct sinking of phytoplankton P are set correctly in generic_COBALT_update_from_bottom subroutine. This PR is to solve #18.

@charliestock , While reviewing the generic_COBALT_update_from_bottom subroutine, I noticed that pointers for the benthic fluxes for the direct sinking of phytoplankton Fe are also turned off. Below is an example for diazotroph Fe:

ocean_BGC/generic_tracers/generic_COBALT.F90

Lines 7707 to 7708 in 01f7a0c

	!call g_tracer_get_pointer(tracer_list,'fedi_btf','field',temp_field)
	!temp_field(:,:,1) = phyto(DIAZO)%ffe_btm(:,:)

Could you confirm if this was done intentionally?

[charliestock]

Thanks for implementing this Yi-Cheng. It looks good to me. If we could run just confirm conservation in the 1D case, that would be great.

[charliestock]

Thanks Yi-Cheng, In answer to your question about the iron, this was indeed intentional. The iron flux to the sediment is removed. The flux from the sediment is then set using a published relationship:

ocean_BGC/generic_tracers/generic_COBALT.F90

Lines 9903 to 9906 in bfa0211

	! iron from sediment (Dale)
	cobalt%ffe_sed(i,j) = cobalt%ffe_sed_max * tanh( (cobalt%fntot_btm(i,j)*cobalt%c_2_n*sperd*1.0e3)/ &
	max(cobalt%btm_o2(i,j)*1.0e6,epsln) )
	cobalt%ffe_geotherm(i,j) = cobalt%ffe_geotherm_ratio*internal_heat(i,j)*4184.0/dt

[yichengt900]

Thanks Yi-Cheng, In answer to your question about the iron, this was indeed intentional. The iron flux to the sediment is removed. The flux from the sediment is then set using a published relationship:

ocean_BGC/generic_tracers/generic_COBALT.F90

Lines 9903 to 9906 in bfa0211

	! iron from sediment (Dale)
	cobalt%ffe_sed(i,j) = cobalt%ffe_sed_max * tanh( (cobalt%fntot_btm(i,j)*cobalt%c_2_n*sperd*1.0e3)/ &
	max(cobalt%btm_o2(i,j)*1.0e6,epsln) )
	cobalt%ffe_geotherm(i,j) = cobalt%ffe_geotherm_ratio*internal_heat(i,j)*4184.0/dt

Thanks @charliestock . I believe the relationship is from Dale et al. 2015, is that correct? I will add more detailed inline documentation to highlight this.

[yichengt900]

Thanks for implementing this Yi-Cheng. It looks good to me. If we could run just confirm conservation in the 1D case, that would be great.

@charliestock . Thanks. I can confirm with this fix, the total P is conservative in the 1D case by checking wc_vert_int_p:

[yichengt900]

This PR passed the CI test and has proven mass conservation for P in the 1D case. Merge now.