Light attenuation

[Ulf Gräwe]

Hi all,
I have written a light attenuation model for the Baltic Sea and North
Sea that correlates salinity and coloured dissolved organic matter
(CDOM). The model is called iow_light and is stand alone.
If I run a npzd model, I would like to take the CDOM light attenuation
into account, in addition to selfshading.

If I set in iow_light get_light_extinction
_SET_EXTINCTION_(k_w)
is that enough or do I have to create dependencies to gotm_npzd?

If yes, how to do that?

Best, ulf


--
Ulf Gräwe

Leibniz Institute for Baltic Sea Research Warnemuende
Dept. for Physical Oceanography and Instrumentation
Seestrasse 15
D-18119 Rostock-Warnemuende
Germany

Tel. : +49 (381) 5197-358
Fax : +49 (381) 5197-440

[Jorn Bruggeman]

Hi Ulf,

_SET_EXTINCTION_(k_w) should suffice - let me know if you find that it does not. FABM will then aggregate all extinction values, and present them to the host model to compute the light field.

Note that there is currently a light attenuation variable in the output that does not include contributions from _SET_EXTINCTION_ yet (to be addressed). Nevertheless, total light extinction sent to the host will be correct.

Cheers

Jorn

> --
> You received this message because you are subscribed to the Google Groups
> "FABM-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to fabm-devel+...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[Ulf Gräwe]

>
> If I set in iow_light get_light_extinction
> _SET_EXTINCTION_(k_w)
> is that enough or do I have to create dependencies to gotm_npzd?

Hi Ulf,

_SET_EXTINCTION_(k_w) should suffice - let me know if you find that it
does not. FABM will then aggregate all extinction values, and present
them to the host model to compute the light field.
>

Hi Jorn,
the total light attenuation, as sent to the host model, is fine.
Im still wondering if I have a spm model and a npzd model: is the light
field, as seen by the npzd model, influenced by the spm models
attenuation?

[Jorn Bruggeman]

> Hi Jorn,
> the total light attenuation, as sent to the host model, is fine.
> Im still wondering if I have a spm model and a npzd model: is the light field, as
> seen by the npzd model, influenced by the spm models attenuation?

The way it normally works: just before the host (gotm) computes the light field, it call FABM's get_light_extinction to get current attenuation values. The returned attenuation is based on FABM models that provide get_light_extinction, and on any variables with specific_light_extinction set. The host should then apply the returned attenuation when computing the light field (e.g., shortwave and photosynthetically active radiation). After this light update, do_* is called. At that point, all biogeochemical models should thus see the light field with all biogeochemical attenuation accounted for.

Cheers

Jorn

[Ulf Gräwe]

Ok,
thats fine. Thanks for the explanation.

Since the light field is an aggregated variable from all individual
light models, I need to be careful, if I do a parameter study and run
for instance 10 npzd models.
Good to know, that I either have to switch of bioshade_feedback or
change the light routine in the npzd model.

Best, ulf


> Jorn

[Jorn Bruggeman]

Hi Ulf,

> Since the light field is an aggregated variable from all individual light models, I
> need to be careful, if I do a parameter study and run for instance 10 npzd
> models.
> Good to know, that I either have to switch of bioshade_feedback or change
> the light routine in the npzd model.

Yes, that's right. For the ensemble functionality, we need a mechanism to assign subsets of biogeochemical models to isolated groups, which do not share aggregate variables. Each group would then have its own attenuation value, and by bringing the light model into FABM (infrastructure for that is already there), each could also have its own light field. Finally, only one of the isolated groups should be selected to feedback light absorption (i.e., heating) to the host.

The main feature that's currently missing in FABM is the ability to define those isolated groups. That'll require some thought.

Cheers

Jorn