PAR field from gotm/light upside down

[janko-om]

Hi!

I use gotm/light to calculate the PAR field for my ecosystem model (nersc/ecosmo). Unfortunately first tests show phytoplankton growth at the seafloor and no growth at the surface, leading to the conclusion that the PAR field must be supplied to ECOSMO upside down. The host is SCHISM v5.11 (with FABM_VERTICAL_BOTTOM_TO_SURFACE) and I use FABMv2.1.3. The PAR field written to the output NetCDF file strangely has the correct vertical orientation.

Any suggestions where to start looking for the cause?

[jornbr]

Hi, that's is certainly surprising. Does this occur with the latest, unmodified FABM code? Specifically, is src/drivers/schism/fabm_driver.h identical to https://github.com/fabm-model/fabm/blob/master/src/drivers/schism/fabm_driver.h?

If yes, is this at all familiar to other SCHISM users? Have there been previous SCHISM-FABM simulations with the same ECOSMO/light models where this issue did not occur? And if so, is this associated with specific SCHISM or FABM versions, for instance, did it occur only after updating one or the other?

[janko-om]

Yes, this occurs with the latest FABM code. I however use a ecosmo.F90 that currently isn't included in the FABM code. The src/drivers/schism/fabm_driver.h is identical to the current master branch.

I previously used SCHISMv5.9 with FABMv0, but did not invoke gotm/light because the light conditions where calculated by an internal light model using get_light and get_light_extinction. The differences now are:

• SCHISM code is updated to v5.11
• fabm_schism.F90 is slightly modified because of some work in progress, but this isn't anything major I think: fabm_schism.F90
• FABMv2.1.3 cloned in September 2024, with no modification of the core code
• gotm/light is now used to provide PAR to ECOSMO
• I adapted ecosmo.F90 for use with FABMv2 and implemented light extinction using self%add_to_aggregate_variable(standard_variables%attenuation_coefficient_of_photosynthetic_radiative_flux)

I'm currently the only user of SCHISM-FABM with ECOSMO+gotm/light. But @platipodium mentioned they were also having problems with the light field?