Very low GPP and TLAI in CLM5 (eCLM) BGC point-simulation for a tropical peat site in Borneo — anyone seen this?

[yohanesginting]

Hey everyone,
I'm running a point-based BGC simulation in CLM5 (eCLM) for a tropical peatland sites in Borneo (based on an eddy covariance tower), and I'm getting some really strange results, especially for GPP and TLAI are coming out suspiciously low after the full spin-up sequence (AD → post-AD → production run).
To give some context on how off things look: TLAI for tropical forests at this kind of site should realistically be somewhere in the 3–8 m²/m² range, but mine is coming out almost ~100% lower than that (~0.03–0.08 m²/m²). My gut feeling is that the broken TLAI is what's dragging GPP down too?, but I'm not entirely sure. I've attached some plots so you can see what I mean.

A bit about my forcing data setup

1. I gap-filled the in-situ met data (temp, RH, shortwave radiation, wind speed) using ERA5, and the bias between the two is pretty low for those variables — so I think that part is fine.
2. Precipitation is the one I'm less confident about — ERA5 underestimates my in-situ hourly precip data. That said, when I aggregate to daily and monthly timescales the correlation looks much better (and lower bias), so hopefully it's not a dealbreaker?
3. For longwave radiation and atmospheric pressure I didn't have in-situ data at all, so I just went with ERA5 directly.
4. Everything was packaged into a .nc file following the standard CLM5 forcing format.

What I'm trying to figure out

1. Could the near-zero TLAI point to a spin-up issue — like not enough AD years, or the carbon pools not reaching equilibrium properly?
2. Is the low GPP most likely just a knock-on effect of the bad TLAI, or could there be something else going on independently (PFT settings, however, I checked I am in the right PFT settings, dominated by BET Tropical almost 90%), soil params, etc.)?
3. Are there any known gotchas in CLM5/eCLM specifically for tropical peat ecosystems? I'm wondering if I'm missing something in the soil or hydraulic parameterization.
4. Could the hourly precip underestimation from ERA5 actually matter much for spin-up, even if the monthly totals look reasonable?

Would really appreciate any thoughts, similar experiences, or pointers in the right direction. Thanks in advance!