CLM5.4 currently underestimates Arctic-boreal fires

[lifang0209]

The PPE improved the spatial pattern of burned area across many regions outside the Arctic boreal zone (ABZ), and decreased the overestimation of fire carbon emissions due to better fuel load simulations. However, it led to an underestimation of ABZ fires. This underestimation is already a major issue in CMIP6 coupling simulations, and PPE worsens it. To address this, fsr_pft and fd_pft for the relevant PFTs need to be tuned higher, or simple functions of NA canopy fire developed from FireAtlas should be added.
Plot is observed and simulated annual burned area. Numbers in parentheses are global total and global spatial pattern of simulations with different observations.

[wwieder]

Thanks for creating this issue, @lifang0209.

It sounds like the dead arctic vegetation in 103-107 is likely (partially) responsible for the low ABFs. Let's see what fires lllok like in 108 (assuming they're alive).

After that, can you please propose new values for fsr_pft and fd_pft that we should consider in our next set of land-only runs to increase our Arctic-boreal burned area?

[samsrabin]

Removing "next" while we wait for @lifang0209's response to @wwieder.