RESUMO
Determining wood carbon (C) fractions (CFs)-or the concentration of elemental C in wood on a per unit mass basis-in harvested wood products (HWP) is vital for accurately accounting embodied C in the built environment. Most estimates of embodied C assume that all wood-based building material is comprised of 50 % C on a per mass basis: an erroneous assumption that emerges from the literature on tree- and forest-scale C estimation, which has been shown to lead to substantial errors in C accounting. Here, we use published wood CF data from live trees, alongside laboratory analyses of sawn lumber, to quantify generalizable wood CFs for HWPs. Wood CFs in lumber average 51.7 %, deviating significantly from a 50 % default wood CF, as well as from CFs in live wood globally (which average 47.6 % across all species, and 47.1 % in tree species not typically employed in construction). Additionally, the volatile CF in lumber-i.e., the quantity of C lost upon heating of wood samples, but often overlooked in C accounting-is lower than the volatile CF in live wood, but significantly >0 % suggesting that industrial lumber drying processes remove some, but not all, of volatile C-based compounds. Our results demonstrate that empirically-supported wood CFs for construction material can correct meaningful systematic biases when estimating C storage in the built environment.