The land surface models that provide surface fluxes of energy and mass to the atmosphere in weather forecast and climate models typically represent plant canopies as a homogenous single layer of phytomass without vertical structure (commonly referred to as a big leaf). This modeling paradigm harkens back to a 30–40-year-old debate about whether big-leaf models adequately simulate fluxes for vegetated surfaces compared to more complex and computationally costly multilayer canopy models. This article revisits that scientific debate. We review the early literature to place our findings in context and discuss recent advancements in roughness sublayer theory, observations of canopy structure and leaf traits, and computational methods that facilitate the use of multilayer models. Using a model with variable vertical resolution, we compare a multilayer canopy representation with the equivalent one-layer canopy to ask how well the one-layer canopy replicates the multilayer benchmark and to identify why differences occur. Comparisons with flux tower measurements at several forest sites spanning multiple years show that sensible heat flux, latent heat flux, gross primary production, and friction velocity for the one-layer canopy degrade in comparison to the benchmark multilayer canopy. For the forest sites considered, 5–10 canopy layers sufficiently reproduce the observed fluxes. Vertical variation of within-canopy air temperature, specific humidity, and wind speed in the multilayer canopy alters fluxes compared with the one-layer canopy. The vertical profile of leaf water potential, in which the upper canopy is water-stressed on dry soils, also causes differences between the one-layer and multilayer canopies. The differences between one-layer and multilayer canopies suggest that the land surface modeling community should revisit the big-leaf surface flux parameterizations used in models.
Bonan et al. (2021) studied this question.