CO2 emissions from land-use change affected more by nitrogen cycle, than by the choice of land-cover data.

The high uncertainty in land-based CO2 fluxes estimates is thought to be mainly due to uncertainty in not only quantifying historical changes among forests, croplands, and grassland, but also due to different processes included in calculation methods. Inclusion of a nitrogen (N) cycle in models is fairly recent and strongly affects carbon (C) fluxes. In this study, for the first time, we use a model with C and N dynamics with three distinct historical reconstructions of land-use and land-use change (LULUC) to quantify LULUC emissions and uncertainty that includes the integrated effects of not only climate and CO2 but also N. The modeled global average emissions including N dynamics for the 1980s, 1990s, and 2000-2005 were 1.8 ± 0.2, 1.7 ± 0.2, and 1.4 ± 0.2 GtC yr(-1) , respectively, (mean and range across LULUC data sets). The emissions from tropics were 0.8 ± 0.2, 0.8 ± 0.2, and 0.7 ± 0.3 GtC yr(-1) , and the non tropics were 1.1 ± 0.5, 0.9 ± 0.2, and 0.7 ± 0.1 GtC yr(-1) . Compared to previous studies that did not include N dynamics, modeled net LULUC emissions were higher, particularly in the non tropics. In the model, N limitation reduces regrowth rates of vegetation in temperate areas resulting in higher net emissions. Our results indicate that exclusion of N dynamics leads to an underestimation of LULUC emissions by around 70% in the non tropics, 10% in the tropics, and 40% globally in the 1990s. The differences due to inclusion/exclusion of the N cycle of 0.1 GtC yr(-1) in the tropics, 0.6 GtC yr(-1) in the non tropics, and 0.7 GtC yr(-1) globally (mean across land-cover data sets) in the 1990s were greater than differences due to the land-cover data in the non tropics and globally (0.2 GtC yr(-1) ). While land-cover information is improving with satellite and inventory data, this study indicates the importance of accounting for different processes, in particular the N cycle.