Significant inconsistency of vegetation carbon density in CMIP5 Earth system models against observational data

Abstract

Earth system models (ESMs) have been widely used for projecting global vegetation carbon dynamics, yet how well ESMs performed for simulating vegetation carbon density remains untested. We compiled observational data of vegetation carbon density from literature and existing data sets to evaluate nine ESMs at site, biome, latitude, and global scales. Three variables—root (including fine and coarse roots), total vegetation carbon density, and the root:total vegetation carbon ratios (R/T ratios), were chosen for ESM evaluation. ESM models performed well in simulating the spatial distribution of carbon densities in root (r = 0.71) and total vegetation (r = 0.62). However, ESM models had significant biases in simulating absolute carbon densities in root and total vegetation biomass across the majority of land ecosystems, especially in tropical and arctic ecosystems. Particularly, ESMs significantly overestimated carbon density in root (183%) and total vegetation biomass (167%) in climate zones of 10°S–10°N. Substantial discrepancies between modeled and observed R/T ratios were found: the R/T ratios from ESMs were relatively constant, approximately 0.2 across all ecosystems, along latitudinal gradients, and in tropic, temperate, and arctic climatic zones, which was significantly different from the observed large variations in the R/T ratios (0.1–0.8). There were substantial inconsistencies between ESM‐derived carbon density in root and total vegetation biomass and the R/T ratio at multiple scales, indicating urgent needs for model improvements on carbon allocation algorithms and more intensive field campaigns targeting carbon density in all key vegetation components.

Predicted Land Carbon Dynamics Are Strongly Dependent on the Numerical Coupling of Nitrogen Mobilizing and Immobilizing Processes: A Demonstration with the E3SM Land Model

“…Differences between simulations are larger for slow-response variables (Figures 3f-i), particularly in soil (Figures 3f,g); these accumulate differences from the fast variables (note the units of those slow variables are petagrams, of which a small change in number could result in large changes in land-atmosphere carbon exchange). However, the differences are well within the range reported in multimodel comparisons (e.g., Beer et al 2010;Ghimire et al 2016;Song et al 2017). The most notable differences between simulations are for cumulative net ecosystem exchange (NEE; which, like other slow-response variables, accumulates the effect from fast processes; Figure 3a;), where the MNL and NUL simulations are quite close to each other, with a difference of 28 PgC by year 2000.…”

Improved Estimates of Biomass Expansion Factors for Russian Forests

“…However, the limited amount of relevant measurements may lead to substantial biases in the estimates of biomass structure. For example, substantial underestimation of root biomass by Earth system models has been reported previously by Song et al [6].…”

Root traits as drivers of plant and ecosystem functioning: current understanding, pitfalls and future research needs

“…All rights reserved global biosphere and climate studies, simulation models can also aid the prioritization of research through sensitivity analyses, for example by identifying key traits whose variation have large consequences for the function of interest (McCormack et al, 2015b). But, most importantly, when tested against empirical data, the results of simulations can discriminate between diverse theoretical models, or reveal (structural or context-dependent) gaps in our mechanistic representation of trait-functioning relationships (Song et al, 2017).…”