how to modify the settings to increase the NPLANT of pft 3 ? ?

[Lo-hatch]

I run a single site with PFT 2,3,4 in IED_INI_MODE=6 using repeated cycle of ERA5 data (1980-2007). However, the nplant of pft3 is small. I have tried to modify some parameters (e.g. CARBON_MORTALITY_SCHEME, TRAIT_PLASTICITY_SCHEME, REPRO_SCHEME, rho of pft3, leave_turnover_rate of pft3), but they all don't work. Besides, the nplant may change dramatically in some years. Anybody knows the possible reasons for these problems (small density and its fluctuation) and the solutions?
Below is the figure of monthly series of NPLANT of PFT3 and PFT4 from the start year. The subfigures represent different dbh classes (ALL; 1-5cm, 5-10, 10-20, 20-30, 30-40, 40-50, >60). The lines of different colors represent different simulation under different settings. The grey dotted line represent the expected nplant. The black dots represent 5yr-interval (because my inventory data is of 5yr resolution).
Also attached with the ED2IN of "carbonM2" simulation (blue line).


ED2IN.txt

[xiangtaoxu]

Some quick thoughts:

1. PFT3 is usually small in my tropical simulations as well. My understanding is that the 'intermediate' trait combinations cannot compete with 'late successional' in the deep understory while they cannot compete with 'early successional' in young and open patches. So there is not a strong competitive advantage of the PFT at least in the current setup. Enabling trait plasticity (TRAIT_PLASTICITY_SCHEME) helped a little from some recent simulations at Harvard if I remember correctly. If you really need to force PFT3 to be abundant, maybe tune seedling_mortality_rate and seed_rain? You can either increase seedling mortality rate/reduce seed_rain for PFT2 or reduce seedling_mortality_rate/increase_seed_rain for PFT3.
2. For NPLANT spikes, they are common in ED and other demography models because when a cohort with large NPLANT go across the predefined DBH threshold, NPLANT of the smaller DBH bin will decline and the bigger DBH bin will increase. This can be especially obvious after major disturbances (e.g. drought) that led to mass mortality. If you feel your NPLANT varies too much, you can either increase the # of cohorts (so transitions would be smoother) and reduce tree sensitivity to major disturbances in your simulations

[Lo-hatch]

@xiangtaoxu Thanks! I have tried with smaller seedling_mortality for PFT3 (0.8) while keep 0.95 for PFT2 and 4. However, the effect is not very obvious (light yellow line in the figure) compared with control group (the red line), maybe I should try to decrease it a bit more?
But I found that if I turn the PLANT_HYDRO_SCHEME=1 and HYDRAULIC_MORTALITY_SCHEME=1 (before they were both 0), the plant density of small trees of three PFT all increase a lot, at the cost of the decrease of NPLANT of large trees (blue line). I have no idea why this could happen because 1) with the HYDRAULIC_MORTALITY "on" the mortality rate increase. 2) the small NPLANT of large trees means that the BSEEDS is not enough to reproduce new cohorts