Recommendations for using FABM with vertically resolved sediments?

[BeatrizArellano]

Hello,

We are working on developing a host model that solves transport processes both in a 1D water column and in a vertically resolved sediment column. At the moment, our approach is to merge the water-column layers and sediment layers into a single vertical grid and then pass that combined column to FABM.

Before going too far down this path, I’d really appreciate some guidance on how FABM is intended to handle situations where the domain extends below the seabed. In particular:

• How should bottom-bound variables be interpreted when sediments are represented explicitly?
  Do “bottom” variables in FABM refer to the true sediment–water interface, or would they implicitly attach to the bottom of my entire (water+sediment) column?

• Is there a recommended way to tell FABM where the seabed is located within the combined vertical grid?
  For example, can the host specify which layer index corresponds to the physical seabed so that FABM can treat sediment layers differently from the pelagic ones?

I would appreciate your guidance, any design tips or special recommendations to handle this particular scenario.

Thanks very much!

[jornbr]

Hi! From a FABM point of view, the bottom and surface are the lower and upper interfaces of a depth-resolved column. Thus, if your column represents a continuum of water and sediment, the FABM "bottom" becomes the lower interface of your sediment column, located at the point below which you no longer resolve processes.

In principle, any interior process that you model with FABM will then apply in every cell of your column, both water and sediment. Sometimes this is desired - the concept of a continuous water-sediment column is often introduced exactly because the same process takes place in both realms - but other times not, e.g., when processes are linked to either water-dwelling or sediment-dwelling biota.

To be able to distinguish water from sediment in this scenario, you'd need to introduce an additional environmental input, which needs to be provided to FABM by the host, and then picked up (as dependency) by those biogeochemistry modules that care whether they are in water or sediment. You could for instance use a field with 1 for water, 0 for sediment. Since any input in FABM is real-valued, such a field could also be seen as a metric of porosity - you could potentially name it such, and also allow values in between 0 and 1 within the sediment to represent actual porosity, if useful for other purposes. In that case, 1.0 means water, anything lower means sediment. In any case, you may want to define your own standard variable for this new input (no need to change FABM for this - just declare your own standard variable in code and use the same definition in both host and BGC module).

Here are a couple of examples that did something similar with FABM:
https://doi.org/10.5194/gmd-10-453-2017 (@e-yakushev, anything you'd want to add?)
https://doi.org/10.1029/2023JC019651 (same BGC in water and sediment, but separate transport; @lumlauf, anything to add?)

Hope this helps!

[BeatrizArellano]

Hi @jornbr,

Thanks so much for the clear and helpful explanation. So, in this particular scenario, biogeochemical models with bottom-only variables would need to be adapted so they can distinguish where they actually are in the column through requesting that specific new environmental variable.

Thanks again, very clarifying!