Active crown fire (Scott & Reinhardt 2001) #857

slevis-lmwg · 2022-04-19T23:16:30Z

Description:

This PR follows in the footsteps of
#573
#572
jkshuman#5
https://github.com/jkshuman/fates/tree/active_crown_Scott_2001
in getting the FATES fire module closer to having an active crown fire parameterization.

Collaborators:

@jkshuman @lmkueppers @pollybuotte @ckoven @glemieux

Expectation of Answer Changes:

At least when fates_fire_active_crown_fire = 1 and active crown fires do occur.

Checklist:

My change requires a change to the documentation.
I have updated the in-code documentation .AND. (the technical note .OR. the wiki) accordingly.
I have read the CONTRIBUTING document.
FATES PASS/FAIL regression tests were run
If answers were expected to change, evaluation was performed and provided

Test Results:

CTSM (or) E3SM (specify which) test hash-tag:

CTSM (or) E3SM (specify which) baseline hash-tag:

FATES baseline hash-tag:

Test Output:

Add EQ to evaluate potential for crown fire Add EQ for passive crown fire igntion Set fraction of crown burnt based on this conditional Add crown fire threshold, crown fire flag, and crown fire PFT param

@jkshuman

Following @jkshuman's introduction of the Bessie and Johnson (1995) formulation for determining the presence of passive crown fire (NGEET#572), I now add the same paper's formulation for determining the presence of active crown fire. The corresponding issue is NGEET#573

Co-Authored-By: Samuel Levis <[email protected]>

…active_crown_fire Resolved conflicts in fire/SFMainMod.F90

…active_crown_fire Resolved Conflicts: main/EDTypesMod.F90

slevis-lmwg · 2022-04-22T22:58:38Z

fire/SFMainMod.F90

+       max_height = 0._r8
+       biom_matrix(:) = 0._r8
+
+!      if (currentPatch%active_crown_fire == 1) then


@jkshuman I see your comment here. I'm not putting the if-statement here because it will prevent passive_crown_FI from becoming greater than 0 when we run without active crown fire.

I am adding the if-statement later, after the comment
! check threshold intensity and rate of spread
That's the only place where I think this if-statement is required.

slevis-lmwg · 2022-04-22T23:05:31Z

main/EDParamsMod.F90

@@ -62,7 +62,7 @@ module EDParamsMod
   real(r8),protected, public :: ED_val_canopy_closure_thresh ! site-level canopy closure point where trees take on forest (narrow) versus savannah (wide) crown allometry
   integer,protected, public  :: stomatal_model  !switch for choosing between stomatal conductance models, 1 for Ball-Berry, 2 for Medlyn

-   logical,protected, public :: active_crown_fire        ! flag, 1=active crown fire 0=no active crown fire
+   logical,protected, public :: active_crown_fire_switch  ! flag, 1=active crown fire 0=no active crown fire


Added suffix _switch to prevent compiler error that complained about same name as subroutine active_crown_fire.

slevis-lmwg · 2022-04-28T23:40:24Z

I haven't worked on this PR this week due to other priorities. I expect to get back to it next week.

As far as I can tell these variables need to be currentPatch% variables: canopy_fuel_load passive_crown_FI heat_per_area ros_torch lb

slevis-lmwg · 2022-05-03T00:31:35Z

fire/SFMainMod.F90

+    !evaluates if there will be an active crown fire based on canopy fuel and rate of spread
+    !returns final rate of spread and fire intensity in patch with added fuel from active crown fire.
+    !currentCohort%fraction_crown_burned is the proportion of crown affected by fire
+    ! TODO slevis: Jackie, you had ROS_torch coming in here but not used


This far, the test
SMS_Lm12_D_Mmpi-serial.1x1_brazil.I2000Clm50FatesCruRsGs.izumi_intel.clm-FatesFireLightningPopDens
returns identical answers when I set
fates_fire_active_crown_fire = 1 in the params file, as when I set
fates_fire_active_crown_fire = 0

@jkshuman before I begin investigating (site too wet? no lightning? maybe it's time to switch to the site that @pollybuotte is running), I wanted to ask you about two variables:

You had ROS_torch entering subroutine active_crown_fire but not used.

You also had canopy_bulk_denisty calculated but not used.

Had you intended to use these variables? If so, where? If the answer is in S&R (2001), feel free to refer me there.

By the way, I stopped passing ROS_torch and four other variables as subroutine arguments, because my sense was that they wouldn't work as scalars and needed to be currentPatch% variables.

@slevisconsulting indeed you are correct those should be Patch variables. I thought that I had a mistake in implementation there, but didn't get around to fixing it. also, yes ROS_torch is not necessary. Might be an informative calculation, but indeed it is not used. Let me remind myself what is going on with canopy_bulk_density.

@slevisconsulting canopy_bulk_density can be removed. it is no longer used.

slevis-lmwg · 2022-05-05T23:02:49Z

fire/SFMainMod.F90

+                   phi_wind_ROS_SA = (ROS_active_min * fuel_bd * eps * q_ig - 3.34 * ir * xi) / (3.34 * ir * xi)
+                else
+                   phi_wind_ROS_SA = 0._r8
+                end if


@jkshuman running this by you:
I was getting phi_wind_ROS_SA = infinity due to ir = 0, which was due to moist_damp = 0. I added this if statement to confirm that ir > 0 .and. xi > 0.
Question 1: Did I make the right choice for the else?

This change caused canopy_frac_burnt < 0 a few lines down due to ROS_SA = 0. So I added max(0... to the calculation of canopy_frac_burnt.
Question 2: Does that seem reasonable?

I have now completed a 5 yr simulation at the CZ2 site. I will share some figures here on github to initiate discussion.

Just saw your message about other updates @jkshuman
I will try those out before posting figures. Thanks.

@jkshuman
in the context of your latest modifications, I'm keeping the if statement and the max(0... that I talked about above, because they still seem relevant when moist_damp = 0 and ROS_SA = 0.

@slevisconsulting normally I would say yes, sounds good, but there must be something more going on in here. you are in this section because the patchFI >= passive_crown_FI indicating there is a fire. And then we are calculating the pieces to get the reaction intensity "ir" for the active crown fire using a given fuel loading for the FM 10 model. It would seem that I have a mistake in this section prior to calculating "ir". I am looking it over now. I belive it is a conversion error. Let me make an adjustment.

slevis-lmwg · 2022-05-06T00:56:06Z

I still get identical results in the runs with and without active_crown_fire = .true.
I think this means that
ROS_active never becomes greater than or equal to ROS_active_min
or
moist_damp never exceeds 0
(or both)

I will investigate further tomorrow. Meanwhile here are plots from the 5-yr run:
$burned_frac_and_drought_status$

slevis-lmwg · 2022-05-06T16:32:21Z

I have confirmed my previous statement. More precisely:
moist_damp for the crown doesn't exceed 0 and, as a result, ROS_active doesn't exceed 0 in the 5-yr run.
The latter means no active crown fire in the 5-yr run.
The former means also no passive crown fire in the 5-yr run.

Sample values from year 2 (see plots above) during the fire season:

Surface fuel

moist_damp	mw_weight	fuel_eff_moist	fuel_mef
0.55	0.48	0.25	0.52

Crown fuel

moist_damp	mw_weight	fuel_eff_moist	fuel_mef
0	1.92	0.48	0.25

where mw_weight = fuel_eff_moist / fuel_mef
and moist_damp > 0 for mw_weight < 1

@jkshuman I'm sharing all this in case something stands out as strange. To me things look reasonable for now.

jkshuman · 2022-05-06T20:00:27Z

@slevisconsulting I pushed a set of conversions for the FM10 fuels to your branch. Can you test again to see if you still get ir = 0 due to moist damp = 0?

pollybuotte · 2022-05-06T20:39:55Z

fire/SFMainMod.F90

+             ROS_active = 3.34_r8 * ((ir*xi*(1.0_r8+phi_wind)) / (fuel_bd * eps * q_ig))
+
+             ! critical min rate of spread (m/min) for active crowning
+             ROS_active_min = (critical_mass_flow_rate / fuel_bd) * 60.0_r8


@jkshuman is this using the correct equation? To me it looks like it's supposed to be calculating the critical minimum rate of spread for active crowning, which is Eq 14 in Scott & Reinhardt. But this line corresponds to Eq 12, which is for calculating the rate of spread for initiation.

Sorry @jkshuman I see I was getting variable names mixed up and that ROS_active_min is the Scott and Reinhardt R'initiation. You can ignore my previous comment.

slevis-lmwg · 2022-05-07T00:19:35Z

Can you test again to see if you still get ir = 0 due to moist damp = 0?

Good news @jkshuman
I now see canopy_frac_burnt > 0 though always < 1 (in the two yrs of simulation). This means that we do not cross our active crown fire threshold in these two yrs, which seems fine, do you agree?

jkshuman · 2022-05-07T02:04:28Z

@slevisconsulting that's great news! so now the race to active crown fire begins.
With these where canopy_frac_burnt > 0 but < 1 there is some canopy burnt, and it alters the FI. Check Patch%FI before canopy burning, FI_final which updates Patch%FI after adjusting for potential canopy fuel consumed. A similar update should happen to patch%ROS_front, and that should be checked as well to confirm it actually changes with canopy consumption.

slevis-lmwg · 2022-05-09T17:09:20Z

fire/SFMainMod.F90

+  currentPatch%canopy_fuel_load * canopy_ignite_energy * canopy_frac_burnt) * &
+  currentPatch%ROS_front / 60._r8
+             ! update patch FI to adjust according to potential canopy fuel consumed (passive and active)
+             currentPatch%FI = FI_final


[...] there is some canopy burnt, and it alters the FI. Check Patch%FI before canopy burning, FI_final which updates Patch%FI after adjusting for potential canopy fuel consumed. A similar update should happen to patch%ROS_front, and that should be checked as well to confirm it actually changes with canopy consumption.

@jkshuman thank you for the suggestion. I checked and, before the update of ROS_front and FI, I sometimes find ROS_final < ROS_front and FI_final < FI. Would you agree that we shouldn't update to reduce ROS_front and FI in those cases? If so, how about:

Suggested change

currentPatch%FI = FI_final

if (FI_final > currentPatch%FI) then

currentPatch%FI = FI_final

end if

...and similar for ROS_front a few lines up.

@slevisconsulting sounds good.

slevis-lmwg · 2022-05-09T20:28:28Z

@jkshuman anything additional that you recommend I check in my quick 2-yr tests at CZ2? Otherwise I may submit a multi-yr simulation to see if conditions cross the active crown fire threshold at some point, i.e.:
ROS_active >= ROS_active_min

slevis-lmwg · 2022-05-10T17:16:02Z

Two 20-yr simulations ran on izumi, one with fates_fire_active_crown_fire = 1 and the other with fates_fire_active_crown_fire = 0. See history output here:

/scratch/cluster/slevis/archive/CZ2_acf_on/lnd/hist
/scratch/cluster/slevis/archive/CZ2_acf_off/lnd/hist

Case directories here (always on izumi):

/home/slevis/cases_FATES/CZ2_acf_on
/home/slevis/cases_FATES/CZ2_acf_off

The first simulation (CZ2_acf_on) satisfies the active crown fire requirement at times, so the outcomes from the two simulations differ. I am happy to perform additional testing or to hand over to @jkshuman and @pollybuotte .

jkshuman · 2022-05-11T22:23:31Z

Thanks for your work on this @slevisconsulting Will keep you posted if there are other things that come up.

FATES_FUEL_AMOUNT_APFC and FATES_MORTALITY_FIRE_SZPF Other fields that Polly requested were active already

Jacquelyn Shuman and others added 30 commits September 4, 2019 16:06

Passive Crown Fire development

7193bc6

Add EQ to evaluate potential for crown fire Add EQ for passive crown fire igntion Set fraction of crown burnt based on this conditional Add crown fire threshold, crown fire flag, and crown fire PFT param

Correct EQ for crown_FI

bad24e4

Update and correct variable names for passive crown fire

093db93

Update name for crown_fire flag

f3077df

Remove redundant details within evaluation for crown fire ignition

f28f546

Add q_dry as available to entire fire module

79a8f5e

Change name for passive_crown fire intensity

ae0a5e7

Revert q_dry to be local within fire ROS subroutine

cafdac5

Update fire/SFMainMod.F90

5558381

Co-Authored-By: Samuel Levis <[email protected]>

Clean up calculation of fraction crown burnt

1d23671

Re-order for clearer calculation of clear branch bole height

764d6c4

Clean up else statements for scorch height fraction crown burnt

a1ebec9

Add Cohort variables for crown damage to CohortDynamics

a0e59a2

Define variable name for crown fire damage

1365bf4

Zero new cohort values for crown damage

45377c9

Add missing parentheses crown damage conditional

73c8762

Clean up crown damage subroutine

c288d76

Remove unnecssary line character crown damage routine

0fc0c14

Define new crown damage variables in EDTypes

fa1973f

Update to only "nan" crown damage variables in EDCohortDynamics

ee2e576

Remove extra parentheses crown damage

39a39f3

Add crown_fire_threshold read from EDTypes

703d87f

Update crown damage for clarity with surface fires

7e691e0

Update crown damage variables to be local to subroutine

52b0734

Remove local crown damage variables from EDTypes

cce9213

Remove local crown damage variables from EDTypes, another location

eb6e6d1

Merge remote-tracking branch 'jkshuman_repo/passive_crown_fire' into …

8ac2053

…active_crown_fire Resolved conflicts in fire/SFMainMod.F90

Zero cambial_mort inside cambial_damage_kill not zero_cohort

dc1ba20

Merge remote-tracking branch 'jkshuman_repo/passive_crown_fire' into …

8eeb2c7

…active_crown_fire Resolved Conflicts: main/EDTypesMod.F90

slevis-lmwg commented Apr 22, 2022

View reviewed changes

Changed 5 variables to currentPatch% variables

03f7992

As far as I can tell these variables need to be currentPatch% variables: canopy_fuel_load passive_crown_FI heat_per_area ros_torch lb

slevis-lmwg commented May 3, 2022

View reviewed changes

slevis-lmwg added 3 commits May 3, 2022 17:40

Some clean-up

594db97

Added checks to avoid div by zero and negative canopy_frac_burnt

e9e282c

Reduction of duplicated code by moving ROS_final to after if-block

82f0df1

slevis-lmwg commented May 5, 2022

View reviewed changes

Bring in Jackie's mods pertaining to canopy_bulk_density

80c6c5a

pollybuotte reviewed May 6, 2022

View reviewed changes

Bring in Jackie's mods: 5cb79fe

4d43cf1

slevis-lmwg commented May 9, 2022

View reviewed changes

slevis-lmwg added 2 commits May 9, 2022 12:25

Ensure that ROS_front & FI only increase due to fire in the upper canopy

cd6ca3f

Add the decimal in the declaration of a real parameter

a87fa2e

Correction of if statement now allows active crown fire occurrence

7109173

Changed two history fields to active

afaf136

FATES_FUEL_AMOUNT_APFC and FATES_MORTALITY_FIRE_SZPF Other fields that Polly requested were active already

jkshuman mentioned this pull request May 24, 2022

Active crown fire development discussion #573

Open

rgknox added the PR status: Not Ready The author is signaling that this PR is a work in progress and not ready for integration. label Nov 14, 2022

adrifoster mentioned this pull request Oct 26, 2023

Active crown scott 2001 #1108

Draft

4 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Active crown fire (Scott & Reinhardt 2001) #857

Active crown fire (Scott & Reinhardt 2001) #857

slevis-lmwg commented Apr 19, 2022

slevis-lmwg Apr 22, 2022

slevis-lmwg Apr 22, 2022

slevis-lmwg commented Apr 28, 2022

slevis-lmwg May 3, 2022

jkshuman May 3, 2022 •

edited

Loading

jkshuman May 3, 2022

slevis-lmwg May 5, 2022

slevis-lmwg May 5, 2022

slevis-lmwg May 6, 2022

jkshuman May 6, 2022

slevis-lmwg commented May 6, 2022

slevis-lmwg commented May 6, 2022 •

edited

Loading

jkshuman commented May 6, 2022

pollybuotte May 6, 2022

pollybuotte May 6, 2022

slevis-lmwg commented May 7, 2022

jkshuman commented May 7, 2022

slevis-lmwg May 9, 2022 •

edited

Loading

jkshuman May 9, 2022

slevis-lmwg commented May 9, 2022

slevis-lmwg commented May 10, 2022

jkshuman commented May 11, 2022

Active crown fire (Scott & Reinhardt 2001) #857

Are you sure you want to change the base?

Active crown fire (Scott & Reinhardt 2001) #857

Conversation

slevis-lmwg commented Apr 19, 2022

Description:

Collaborators:

Expectation of Answer Changes:

Checklist:

Test Results:

Choose a reason for hiding this comment

Choose a reason for hiding this comment

slevis-lmwg commented Apr 28, 2022

Choose a reason for hiding this comment

jkshuman May 3, 2022 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

slevis-lmwg commented May 6, 2022

slevis-lmwg commented May 6, 2022 • edited Loading

jkshuman commented May 6, 2022

Choose a reason for hiding this comment

Choose a reason for hiding this comment

slevis-lmwg commented May 7, 2022

jkshuman commented May 7, 2022

slevis-lmwg May 9, 2022 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

slevis-lmwg commented May 9, 2022

slevis-lmwg commented May 10, 2022

jkshuman commented May 11, 2022

jkshuman May 3, 2022 •

edited

Loading

slevis-lmwg commented May 6, 2022 •

edited

Loading

slevis-lmwg May 9, 2022 •

edited

Loading