Incorporate the impact of O₂ concentration on fuel flammability

[samsrabin]

incorporate the impact of O₂ concentration on fuel flammability to address the fire overestimation in Tibet:

(1) Add:

forc_po2 => atm2lnd_inst%forc_po2_grc, & ! Input: [real(r8) (:)]  O2 partial pressure (Pa)

(2) fire_m should be changed to:

fire_m = ((afuel*arh30+(1._r8-afuel)*arh)**1.5_r8) &
     *((1._r8-btran_col(c)/wtlf(c))**0.5_r8) &
     *min(1._r8, max(0._r8, (forc_po2(g)-15000)/2500._r8))

Originally posted by @lifang0209 in https://github.com/ESCOMP/CTSM/issues/2726#issuecomment-2307841940

[lifang0209]

@samsrabin @ekluzek @wwieder Sam requested a figure demonstrating the impact of the O₂ limit. The attached figure displays the spatial distribution of the annual burned area fraction. Observations (obs) represent the average of three satellite-based burned area products (GFED5s, FireCCI, and MODIS C6). Simulations show results for the new fire schemes both with and without an O₂ limit. The figure includes global burned area totals and spatial correlation coefficients (R) between observations and simulations. As illustrated, the O₂ limit reduces the overestimation of the burned area fraction in Tibet, and improves the spatial pattern simulation performance (global spatial correlation with observations increases from 0.76 to 0.79).

[samsrabin]

Thanks, Fang! I'm assuming that's for 2001-2020 or so?

[wwieder]

Thanks, Fang.

[lifang0209]

Thanks, Fang! I'm assuming that's for 2001-2020 or so?

Hi Sam, this is for 2001-2014 and CTSM5.2 with fire scheme updates forced by CRUJRA. I'll check the CTSM5.3 outputs soon to make sure everything about the fire scheme goes correctly.

[samsrabin]

@lifang0209 We discussed it this morning, and we think it is possible to get this in for CTSM6/CESM3. We were wondering, though: Where did you get this parameterization? It would be good to have a source cited in the code and docs.

[wwieder]

The other comment from our meeting was that in coupled simulations we're not yet running with interactive fires, so bringing in the change won't have a big impact on climate. It was also discussed if this region is burning so much because of a high bias in fuel loads, the lack of O2 limitation at high altitudes, or (more likely) a combination of these factors?

[lifang0209]

@samsrabin Source: Li et al. (2024, in prep.) will describe this. The parameters are calibrated based on the observed relationship between the burned area and o2 partial pressure. In CNFireLi2024Mod, "revised in May, 2015, according to Li et al. (2015, in prep.) ! Fire-related parameters were calibrated or tuned in May, 2015 based on the ! 20th Century transient simulations at f19_g16 with a CLM4.5 version ! (clm50fire), CRUNCEPv5, and climatological lightning data." should be replaced by "revised in Jun 2024, according to Li et al. (2024, in prep.). "

[lifang0209]

@wwieder Thanks for the information. I agree that the modification mainly affects fire simulations, and has limited influence on the climate. I also agree that both the limitation of oxygen at high altitudes and the possibly overestimated biomass of C3 Arctic could contribute to an overestimation of fires in the Tibet center. CESM2 much overestimated both the LAI and biomass in the region. Will CLM6 address the issue or not?

[samsrabin]

In an email thread, @lifang0209 says she now suspects there are issues with too much biomass and a dry bias in Tibet. I'll close this issue as wontfix, but I've noted the problem with too much Tibetan fire in the Known Issues discussion.