Accounting for forest carbon pool dynamics in product carbon footprints: Challenges and opportunities Joshua P. Newell a, ⁎, Robert O. Vos b, 1 a School of Natural Resources and Environment, University of Michigan, Ann Arbor, USA b Spatial Sciences Institute, University of Southern California, USA abstract article info Available online xxxx Keywords: Forest carbon Land use change Land use modification Product carbon footprint protocols Biogenic carbon Wood life cycle inventories Modification and loss of forests due to natural and anthropogenic disturbance contribute an estimated 20% of annual greenhouse gas (GHG) emissions worldwide. Although forest carbon pool modeling rarely suggests a ‘carbon neutral’ flux profile, the life cycle assessment community and associated product carbon footprint protocols have struggled to account for the GHG emissions associated with forestry, specifically, and land use generally. Principally, this is due to underdeveloped linkages between life cycle inventory (LCI) modeling for wood and forest carbon modeling for a full range of forest types and harvest practices, as well as a lack of transparency in globalized forest supply chains. In this paper, through a comparative study of U.S. and Chinese coated freesheet paper, we develop the initial foundations for a methodology that rescales IPCC methods from the national to the product level, with reference to the approaches in three international prod- uct carbon footprint protocols. Due to differences in geographic origin of the wood fiber, the results for two scenarios are highly divergent. This suggests that both wood LCI models and the protocols need further development to capture the range of spatial and temporal dimensions for supply chains (and the associated land use change and modification) for specific product systems. The paper concludes by outlining opportuni- ties to measure and reduce uncertainty in accounting for net emissions of biogenic carbon from forestland, where timber is harvested for consumer products. © 2012 Elsevier Inc. All rights reserved. 1. Introduction Forests cover approximately 65% of the total land surface and play a vital, yet complex role in the global carbon cycle. Holding 90% of the plant biomass carbon and 80% of soil carbon found in all terrestrial eco- systems, they also assimilate 67% of the total carbon dioxide removed from the atmosphere by these ecosystems (Landsberg and Gower, 1997). Annual loss of forests due to disturbance (harvesting, conversion, fire, insects, pathogens, and wind) contributes as much as 20% of total global greenhouse gas (GHG) emissions each year—rivaling emissions from the global transportation sector (Denman et al., 2007). Gower (2003) divides the carbon cycle into two interrelated phases: 1) initial disturbance effects on carbon pools and 2) changes in carbon cycle processes during forest ecosystem recovery or succes- sion. These can be ‘natural’ disturbance events (e.g., fire, pest outbreak, etc.) or anthropogenic events (e.g., timber harvest, road construction, mining, etc.). The life cycle assessment (LCA) community and associated guidance methodologies, such as product carbon label protocols, have struggled to adequately account for GHG emissions associated with direct and indi- rect land use change in consumer products made from forest resources. A major reason for this is the spatial complexity of land use, as geography and industrial practices vary widely. Advances in terrestrial carbon modeling reveal a complex dynamic of the carbon pools in forests, but rarely a ‘carbon neutral’ flux profile as has been modeled traditionally in LCA (Johnson, 2009). The work by Searchinger et al. (2008) re- minds us of the potential for ‘accounting’ error if biogenic carbon emissions related to land use are ignored. Starting in 2008, procedures and methodologies to account for di- rect land use change were developed in several major product carbon footprint protocols. Within these methodologies, land use modifica- tion, where the category of the land (e.g., forest) remains the same, but the carbon pools in that land have been fundamentally disturbed, is also at issue (viz., ‘old-growth’ forest converted into a working forest, with short-term harvest rotations and intensive cultivation practices, including management of plantation seedlings). The IPCC has issued a well-developed methodology for calculating emissions in the same land use category (e.g., forestland remaining forestland) and for shifts from one land use category to another (e.g. forestland to cropland), but this guidance is designed for GHG in- ventories at the national level, not the product level (IPCC, 2006). Environmental Impact Assessment Review xxx (2012) xxx–xxx ⁎ Corresponding author at: 440 Church Street, Ann Arbor, MI 48109, USA. E-mail addresses: jpnewell@umich.edu (J.P. Newell), vos@usc.edu (R.O. Vos). 1 Present address: 3616 Trousdale Parkway, AHF B55, Los Angeles CA 90089-0374, USA. EIR-05778; No of Pages 14 0195-9255/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.eiar.2012.03.005 Contents lists available at SciVerse ScienceDirect Environmental Impact Assessment Review journal homepage: www.elsevier.com/locate/eiar Please cite this article as: Newell JP, Vos RO, Accounting for forest carbon pool dynamics in product carbon footprints: Challenges and oppor- tunities, Environ Impact Asses Rev (2012), doi:10.1016/j.eiar.2012.03.005