Comparison of uncertainties in land-use change fluxes from bookkeeping model parameterisation

Abstract. Fluxes from deforestation, changes in land cover, use and management practices (FLUC for simplicity) contributed to approximately 14 % of anthropogenic CO2 emissions 2009–2018. Estimating FLUC accurately space time remains, however, challenging, due multiple sources uncertainty the calculation these fluxes. This uncertainty, turn, is propagated global regional carbon budget estimates, hindering compilation a consistent preventing us constraining other terms, such as natural sink. Uncertainties estimates arise many different sources, including differences model structure (e.g. process based vs. bookkeeping) parameterisation. Quantifying uncertainties each source requires controlled simulations separate their effects. Here, we analyse between two bookkeeping models used regularly since 2017: by Hansis et al. (2015) (BLUE) that Houghton Nassikas (2017) (HN2017). The have very similar philosophy, but differ significantly both with respect intensity spatiotemporal variability. land-use forcing also We find larger BLUE compared HN2017 are largely C densities managed vegetation or primary secondary vegetation, higher allocation cleared harvested material fast turnover pools than HN2017. Besides parameterisation forcing, assumptions cause differences: particular represents gross transitions which leads overall losses more quickly realised