Remove allometry size thresholds in FATES

[JessicaNeedham]

FATES has some threshold parameters related to allometry, e.g. fates_recruit_seed_dbh_repro_threshold and fates_allom_dbh_maxheight that are not biologically realistic and can cause strange behavior in some contexts. For example, the dbh at max height parameter causes an increase in radial growth at large sizes which is inconsistent with data, and might in part contribute to FATES’s large tree problem. It is also possible that step functions are causing issues with integration as part of daily allocation of carbon.

We discussed in the software meeting today removing these thresholds. While different species do have different max heights in nature, this can be approximated with an asymptotic function e.g. the Martinez-Cano et al. 2019 that is already in FATES. @adamhb ’s TRS (tree recruitment scheme) also relates dbh to reproductive allocation using a continuous function rather than a threshold value which seems more biologically reasonable.

Does anyone see any reason to keep these thresholds or could we remove them in a future PR?

[adamhb]

Hi @JessicaNeedham,

I support removing the step functions for the reasons you mention above. As mentioned in our discussion on PR 1093, there are some fringe cases where it can be useful to cap leaf biomass using fates_allom_dbh_maxheight in this section of code when allom_lmode = 3. But this was really just a stopgap measure to keep the needed number of vai bins below 30 for large, tall, narrow crowns, not a good long term solution.

Another related physiology question that @JessicaNeedham and I were discussing is if leaf biomass should continue to increase with DBH, or if there is a point where this should also asymptote in large trees. If maintenance costs continue to increase then leaf biomass might need to too, but perhaps at some point large trees increase their dbh (just heartwood) without increasing the live (metabolizing) components and therefore leaf biomass doesn't keep growing with DBH...?

[JessicaNeedham]

Here's dbh by leaf biomass from the BAAd database:

It looks like an exponential or almost linear depending on PFT, although you're right that there isn't a huge amount of data for the large sizes.

[JoshuaRady]

@JessicaNeedham, I would not like to see fates_allom_dbh_maxheight deprecated. I am use fates_allom_dbh_maxheight to specify an asymptotic max height for one of my allometries. This might be considered a misuse of this this parameter but with a limited number of parameter sometimes the interpretation needs to vary a bit across routines that uses it. I'm hoping to find time to put in a PR to add my new allometries and was going to bring this issue up in it, possibly suggesting a change in name. In any case I would appreciate it if the parameter stuck around for the moment.

I hear your underlying concern about the unrealistic nature of the way that the max height is implemented in the current allometries. I see that as a code/allometry problem. However, given the other limitations in FATES a max height parameter seems like it may be useful for a while.

[JessicaNeedham]

Hi @JoshuaRady , thanks for the feedback. Could you say more about how you are using the fates_allom_dbh_maxheight parameter? How does your asymptotic max height allometry differ from using the Michaelis-Menten function specified by allom_hmode = 5?