Fire Scheme in CLM-FATES

[KumarR]

Dear CESM users,

I am currently using the FATES compset (e.g., I2000Clm50Fates) to run the CLM-FATES model, but I’m a bit confused about which fire scheme is actually being used in the simulation. From what I understand:

• FATES utilizes the SPITFIRE fire scheme.
• CLM5, on the other hand, uses the fire scheme described by Li et al. (2012).

Since these two fire schemes are substantially different, I’m unsure which one is being implemented when using the FATES compset. Could someone clarify which fire scheme is used in CLM-FATES simulations and whether there is any interaction between the two?

Thank you!
Kumar

 

[slevis]

@KumarR by looking in the code I determined that FATES uses the SPITFIRE scheme without interacting with the Li fire model.

 

[KumarR]

@slevis Thank you for the clarification. From my understanding, the fire duration in FATES is set to 240 minutes, while in CLM it is one day. Could you please explain how both CLM and FATES interact when using CLM-FATES for global simulations? Specifically, how do the different time scales for fire duration in both models interact and influence the simulation outcomes?

Thanks,
Kumar

 

[slevis]

I will refer you to the Technical Note:

2. CLM Technical Note — ctsm release-clm5.0 documentation

 

[KumarR]

Thank you.

 

[ajoy]

Dear community member, I have a question about the CLM-FATES variable 'Fire_Closs'. Is carbon loss calculated based on atmospheric processes, or is it calculated based on productivity (GPP before and after)?

Thank you...

 

[afoster]

Ajoy,

FIRE_CLOSS is a diagnostic for how much burned - which would be calculated based on both the fuel and vegetation amount, as well as fire weather conditions.

 

[ajoy]

Ok.
Thank you...

 

[ajoy]

I have another query. If I want to incorporate Real Fire Data (i.e., GFED) into the model, what do I need to do?
The compset is I2000Clm50FatesCruGs.
Thank you in advance...

 

[afoster]

Ajoy, if you mean driving the model with observed burned area that is not possible. If you want to compare to observed burned area I would do this in post-processing.

 

[ajoy]

Thank you for the response. However, I have seen studies where a burnt area dataset is used to drive the model, which is why I’m asking.

 

[ajoy]

I think I am unable to comprehend,
Can you please check the Article
Thank you...

 

[afoster]

I actually had not seen that paper, so thank you for pointing us to it. It looks like the authors archived their code here. They did not submit a PR with their updates to the main CTSM github page. I would suggest contacting the authors to inquire about using their code (or downloading from the link I provided). You could also post an issue or discussion to the CTSM github page linking to the code repository and possibly linking in the authors.

Right now we do not have plans to integrate their code (even though it would be great to do so), but we do not have the resources at the moment.

 

[ajoy]

Thank you for the clarification and the link to the archived code. I will follow up ...