Skip to content

Integration of the litter model needs to be improved #979

@jacobcook1995

Description

@jacobcook1995

Is your feature request related to a problem? Please describe.
I currently integrate the litter model using the simplest possible method, i.e. I find the total decay of each pool by

$$D = k P\Delta t ,$$

where $k$ is the decay rate, $P$ is the pool size and $\Delta t$ is the time step. This approximation only works when $\Delta t$ is small relative to the rate at which the pool sizes are changing, which given we use a month time step as default is only true when the pools are close to equilibrium (or when the forest floor is basically frozen 🫠)

Describe the solution you'd like
I'm unsure what I was thinking with the above as the dynamics are simple enough that they can be integrated to obtain an analytical solution (which I even provided to @hrlai here). I think I should try and extend this analytical solution to also include litter input, which should stop the weird behaviour we are seeing

Describe alternatives you've considered
Alternatively, I could convert the whole litter model into rates, and integrate the model using scipy. The big problem with doing this is that this require animal consumption to be expressed as a rate rather than a total amount, which would introduce a new way for litter pools to reach negative mass. So, I don't think I should go down this route unless I can't obtain an analytic solution

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions