Influence of the decoupling degree on the estimation of canopy stomatal conductance for two broadleaf tree species

• The seasonal dynamics of the decoupling coefficient were dominated by stomatal conductance.
• The canopy stomatal conductance was underestimated with a simplified model.
• A function that associated the LAI and decoupling coefficient was derived to estimate the actual stomatal conductance.

Canopy stomatal conductance (GS) estimation is a critical aspect of evapotranspiration research. Selecting an appropriate model for scaling-up is necessary to study plant water regulation. Two monoculture plantations of broadleaf tree species, Eucalyptus urophylla and Schima superba, were selected to study GS via a widely recognized equation simplified by Köstner (GS1) and an inverse Penman–Monteith equation (GS2). The decoupling coefficient (Ω) was estimated to quantify the decoupling extent of stomata from the atmosphere. We found that both species were well coupled with environmental conditions (0.1 ± 0.06 and 0.22 ± 0.09 for E. urophylla and S. superba). Ω increased exponentially with canopy conductance (GC) and was depressed by the increase of Ga, which implied that there was a combined climatic and physiological regulation on Ω. GS1 for both species (63.7 ± 33.9 mmol m−2 s−1 and 48.4 ± 18.1 mmol m−2 s−1 for E. urophylla and S. superba, respectively) was underestimated with the simplified equation compared to GS2 (77.0 ± 52.4 mmol m−2 s−1and 112.0 ± 52.5 mmol m−2 s−1 for E. urophylla and S. superba, respectively). The ratio of GS1/GS2 linearly decreased with Ω by a slope of −1.92 and −1.31 for E. urophylla and S. superba, respectively. The increase of the LAI tended to increase the decoupling extent, which further reduced the accuracy of the estimation of GS1. According to our results, the ratio of GS1/GS2 had a mean of 0.94 for 85.3% and 0.6 for 37.6% of all of the data for E. urophylla and S. superba, respectively, implying a better estimation of GS1 for the stand that had a lower LAI. Based on the relationship between the LAI and the ratios of GS1/GS2 (ratio = 1.0759 × 0.02547(0.053LAI + 0.054) − 0.0078), we can shed some light on the prediction error of stomatal conductance derived from the simplified equation in different forest types.