Passive crown-fire code development

[slevis-lmwg]

At the 8/16/2019 FATES CA meeting we decided to begin development of the crown-fire code with "passive" crown-fires:

When scorch height >= height of bottom of canopy, fire will now spread upward.

I'm looking at this section of existing code

The simplest implementation involves adding this if statement: if (evergreen) then currentCohort%fraction_crown_burned = 1.0_r8 else currentCohort%fraction_crown_burned = as it was before end if

To add complexity we make the calculation of fraction_crown_burned a function of moisture (btran as proxy) and biomass density.

I'm open to comments and suggestions @jkshuman @lmkueppers @pollybuotte @rgknox @ckoven @xuchongang

[jkshuman]

Thanks for opening this @slevisconsulting Tying the extent of crown fire to moisture as characterized by BTRAN will get us functional until FATES-Hydro is online for CA. The percentage of evergreen biomass may not be necessary for passive crown fire. If the vegetation is a crown fire type, then the implementation plan will address this. Percentage of crown fire trees will be more important with active crown fire (crown to crown fire spread). I think it makes sense to put this update here, but we need to consider how to make this a bit more universal where this regional application will not affect other locations. We cannot add a blanket if=evergreen conditional because we do not want crown fire in the tropics. Perhaps we add a PFT trait of "crown fire" =(1=allow, 0=do not) to the params file? This way the crown fire development will not impact tropical evergreen PFTs. Any other ideas about making this more general, or is a param file trait enough? @lmkueppers @pollybuotte @rgknox @ckoven @xuchongang

[rosiealice]

Seconding that this is a cool idea :)

Wrt. the 'evergreen' criteria, would one never have a crown fire in a larch forest? (and moreover, why wouldn't we -ever- expect- to have crown fire in tropical forest?) Is it to do with needleleaf vs. broadleaf flammability? Or something to do with branch architecture?

I don't know but I also suspect it to be more than just phenology habit...

On Mon, 19 Aug 2019, 19:10 jkshuman, notifications@github.com wrote:

Thanks for opening this @slevisconsulting https://github.com/slevisconsulting Tying the extent of crown fire to moisture as characterized by BTRAN will get us functional until FATES-Hydro is online for CA. The percentage of evergreen biomass may not be necessary for passive crown fire. If the vegetation is a crown fire type, then the implementation plan will address this. Percentage of crown fire trees will be more important with active crown fire (crown to crown fire spread). I think it makes sense to put this update here, but we need to consider how to make this a bit more universal where this regional application will not affect other locations. We cannot add a blanket if=evergreen conditional because we do not want crown fire in the tropics. Perhaps we add a PFT trait of "crown fire" =(1=allow, 0=do not) to the params file? This way the crown fire development will not impact tropical evergreen PFTs. Any other ideas about making this more general, or is a param file trait enough? @lmkueppers https://github.com/lmkueppers @pollybuotte https://github.com/pollybuotte @rgknox https://github.com/rgknox @ckoven https://github.com/ckoven @xuchongang https://github.com/xuchongang

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/NGEET/fates/issues/567?email_source=notifications&email_token=ADC2YQ332RNP4U7HHEFGDVDQFLHYNA5CNFSM4IMQHVNKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD4TUZ3I#issuecomment-522669293, or mute the thread https://github.com/notifications/unsubscribe-auth/ADC2YQ5PAWAJ62ZQRNRAPX3QFLHYNANCNFSM4IMQHVNA .

[jkshuman]

@rosiealice this is a good point. we can just have this conditional based on the parameter "crown_fire" and then it is user defined. In general the conifer vegetation in temperate and boreal regions are more flammable due to the makeup of their needles. Crown fires typically drop when they enter a deciduous broadleaf stand. I should confirm whether or not tropical trees have crown fires, but I know in temperate systems crown fires are a component of conifers and not deciduous trees. The larch is a good question...

[slevis-lmwg]

@jkshuman @rosiealice thank you for your comments. I will start by adding crown_fire to the params file.

[pollybuotte]

Larch can have crown fires. I second the idea of making this a user-defined parameter.

[lmkueppers]

Does adding a user defined crown fire parameter (i.e., crown_fire) mean that for some PFT it could be always and forever 0? This seems unlikely for passive crown fire, but useful for model testing/development.

The fraction of the crown burned seems the more important thing to work out. In addition to a dynamic moisture variable (Btran or LWC), already existing functional traits/parameters (SLA, branch turnover rate, and/or leaf arrangement parameter) may be able to capture the predilection for boreal and temperate conifer canopies to burn. Dead branch retention has been identified as a flammability trait that can favor fire adapted tree taxa. This could also be relevant for shrubs.

[slevis-lmwg]

Does adding a user defined crown fire parameter (i.e., crown_fire) mean that for some PFT it could be always and forever 0? This seems unlikely for passive crown fire, but useful for model testing/development.

The fraction of the crown burned seems the more important thing to work out. In addition to a dynamic moisture variable (Btran or LWC), already existing functional traits/parameters (SLA, branch turnover rate, and/or leaf arrangement parameter) may be able to capture the predilection for boreal and temperate conifer canopies to burn. Dead branch retention has been identified as a flammability trait that can favor fire adapted tree taxa. This could also be relevant for shrubs.

Thank you, @lmkueppers, your comment prompted me to revisit the first attempt at crown fire that I committed yesterday:

For crown_fire = 0, I had kept the existing code that burns a fraction of the canopy when scorch height exceeds bottom-of-canopy. However, now I realize that that canopy fraction should also be the minimum that burns in a (passive) crown fire.

So I'm including that to my changes where now, starting from that minimum, a larger fraction of the canopy may burn if conditions are dry.

I will commit an update soon.

[jkshuman]

@slevisconsulting @lmkueppers when this development is on its own branch you can submit it as a PR, and then open for review. This was I can make in code suggestions that you can accept. Additionally, commits to the branch will become part of the PR. @rgknox suggested we do this on the latest release (tag 1.30 api 8), so I can help you get that setup on Tuesday.

then we can link to that branch from this issue.

link to current development and notes https://github.com/slevisconsulting/slevis_fates_work/commit/8fac62e98b601eecd3fdedd77a0a884f05eab05f#commitcomment-34836317

[jkshuman]

The new development branch for this issue is https://github.com/jkshuman/fates/tree/passive_crown_fire and will be merged to master in PR #572 @adrifoster identified Bessie and Johnson (1995) for this application of passive crown fire, and after consulting with @rgknox and @slevisconsulting I implemented development.

[jkshuman]

Implementation EQ 14 of Bessie and Johnson has been updated per Scott and Reinhardt 2001 to consider stand level canopy fuel density versus tree level fuels for crown fire initiation on a patch #572 and integrated into #584 . edited in PR.