search

tilbud / rCMEM

GNU General Public License v3.0
2 stars 4 forks source link

Make model inputs and outputs consistent with PEcAn #13

Closed HolmquistJ closed 4 years ago

HolmquistJ commented 4 years ago

PEcAn (the predictive ecosystem analyzer) is a great platform for model-data integration and has a lot of functionality for doing things like parameter-data-assimilation, sensitivity analysis, and forecast uncertainty propagation.

To get a model into PEcAn involves writing some wrapper functions to translate model inputs and outputs to standardized vocabulary and units. The maintainers of PEcAn recommended that if we're still coding up the model we can adopt the units and vocab at this stage and not need the wrapper scripts later.

Here's their manual and the part with details about the controlled varaibles

ktoddbrown commented 4 years ago

In general, I like this idea! Devil is in the details however.

The respiration and carbon stock units are per area (kg m-2 s-1 for respiration and kg m-2 for carbon stocks) which are problematic for a depth explicit model like this. Is there anyway we can push back and get a density flux instead?

Also while I'm happy to work in seconds this is not quite what most experimentalist will measure things in :D Not sure if we want to push back on this as well.

At this stage my vote would be for a wrapper over direct adoption.

HolmquistJ commented 4 years ago

Ok. I agree. Let's go for the wrapper solution.