Purpose This paper aims to clarify the application of a landuse baseline in attributional life cycle assessment (ALCA) for product systems involving land use, through consideration of the fundamental purpose of ALCA. Currently, there is no clear view in the literature whether a baseline should be used when accounting for environmentally relevant physical flows related to land use. Methods An extensive search of literature was carried out using the key terms 'attributional life cycle assessment' and 'attributional LCA' in the Google Scholar web search engine. Approximately 700 publications were reviewed and summarised according to their type and scope, relevance of land use, key statements and references given for ALCA, and arguments for and against using a baseline in ALCA. Based on the literature review and supplementary literature references, a critical discussion on the use of a baseline and determination of the most appropriate land-use baseline in ALCA is provided. Results and discussion A few studies clearly argued that only absolute (observable) flows without a baseline are to be inventoried in ALCA, while the majority of the studies did not make any clear statement for or against. On the other hand, a land-use baseline was explicitly applied or proposed in a minority of the studies only, despite the fact that we classified land use as highly relevant for the majority of the studies reviewed. Furthermore, the LCA guidelines reviewed give contradictory recommendations. The most cited studies for the definition of ALCA provide general rules for selecting processes based on observable flows but do not argue that observable flows necessarily describe the environmentally relevant physical flows. Conclusions We conclude that a baseline is required to separate the studied parts of the technosphere from natural processes and to describe the impact of land use on ecosystem quality, such as carbon sequestration and biodiversity. The most coherent baseline for human-induced land-use in ALCA is natural regeneration. As the natural-regeneration baseline has typically been excluded, may vary bio-geographically and temporally, and is subject to uncertainties, case studies applying it should be performed so that implications can be studied and evaluated. This is particularly important for agricultural and forestry systems, such as food, feed, fibre, timber and biofuels.
Responsible editor: Matthias FinkbeinerElectronic supplementary material The online version of this article (