drnimbusrain
commented
10 months ago @quaz115 I would like to add Maggie here as a contributor here as well. This does seems an important simplification in canopy-app biogenics currently, but not one that is easy to remedy.
As discussed at today's UFS-CATChem development meeting, there may be need to expand the current capabilities of canopy-app biogenics to have more granularity and flexibility across chemical complexity and more PFT complexity.
We could also bring in more complexity of the many PFTs as in MEGANv3 input files mapped across fractional gridded growth forms and ecotypes. Since different tree species can have very different BVOC emission rates, this can be very important for accurate canopy-app predictions. Currently, the compound class EFs are based on the simplified, mapped MODIS/VIIRS input vegetation types, but likely we should bring in this data/methods in some way to have more granularity across different PFTs. The global, lat/lon gridded input ecotypes and fractional growth forms (in NetCDF CF-convention) used in MEGANv3 are found by example on Hopper at: _/groups/ESS/pcampbe8/MEGAN/MEGAN3inputs/.
MEGANv3 uses the python scripts (e.g., on Hopper: _/groups/ESS/pcampbe8/MEGAN/MEGAN3_work_GFSv16_NACC_G1/02.MEGEFP32/MEGANEFP.py) to integrate the input gridded ecotype and fractional growth forms maps across plant species composition and species-specific emission factor CSV file data to get the final EFs across 19 compound classes. The input PFT species-specific emissions factor data and associated fractional growth form mapping/speciation CSV files are found on Hopper at: _/groups/ESS/pcampbe8/MEGAN/MEGAN3_work_GFSv16_NACCG1/02.MEGEFP32/inputs/EFP/. This is ultimately more complex than canopy-app (i.e. uses MEGANv2 with simple vegtype dependent EFs from Guenther et al., 2012), and could be tricky, so can we include these processes in some simpler way?