psi_max or psi_min in leaf phenology

[Lo-hatch]

when IPHEN_SCHEME=4, phenology is drought deciduous, the material to decide whether it is a low-psi day is :
cpatch%dmax_leaf_psi(ico) < leaf_psi_tlp(ipft)
however, leaf_psi_tlp always lower than -100m, while dmax_leaf_psi barely lower than -100m.
maybe using dmin_leaf_psi here is more reasonable?

https://github.com/EDmodel/ED2/blob/master/ED/src/dynamics/phenology_driv.f90#:~:text=cpatch%25dmax_leaf_psi(ico)%20%3C%20leaf_psi_tlp(ipft)

[xiangtaoxu]

I tried dmin_leaf_psi before, which resulted in too much deciduousness in Costa Rica dry forests because leaf_psi can be very low during midday (for an hour or so). Ecologically, it might be an overreacted to drop your leaves just because leaf water potential was low for an hour or so during midday. Physiologically, TLP can become lower during midday due to higher sugar loading, which is not considered in ED2 right now.

Therefore, the philosophy to use dmax_leaf_psi is that plants shed leaves only when it is dry enough that they cannot recharge daytime water loss during the night. I usually tune belowground parameters (which are much more uncertain) or leaf/sapwood allometry (also uncertain) if you need to get desired phenology.

Brett Wolfe has a paper on drought-deciduousness in tropical seedlings, where he used the average of min/max water potential to predict leaf shedding. You can try that as well (https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.14087)