Land Use as a Climate Forcing in Europe Over the Last 7,000 Years: A Review

The biogeophysical forcing resulting from the change in land-surface properties (such as albedo) is of particular interest as it can increase, decrease, or offset the biogeochemical forcing resulting from the changes in atmospheric greenhouse gas concentrations due to deforestation, forest management, or reforestation/afforestation (i.e., land-use/land-cover change, LULCC). Moreover, the biogeophysical effects operate locally and can affect ecosystems and humans directly, while biogeochemical effects influence the entire globe over the long term. Climate models need descriptions of the past vegetation on Earth to take account of the feedbacks and forcings on past climate due to both climate-induced and human-induced changes in the vegetation cover. Therefore, in order to study land-use forcing in the past, climates simulations have used hypothetical total afforestation or deforestation of Europe, or Holocene deforestation estimated using models of population growth, estimates of per capita land use and historical-archaeological data (e.g., KK10 and HYDE data sets), or pollen-based reconstructions of Holocene land-cover change (using various methods and models). Pollen-based reconstructions have the advantage to provide datasets of the actual vegetation in the past, i.e. both climate-induced (natural) vegetation and human-made vegetation via land use. Such pollen-based reconstructions were produced in a format appropriate for climate-model simulations. Dynamic vegetation models are useful to simulate potential (natural) vegetation of the past and can help to disentangle natural vegetation from human-made vegetation in pollen-based reconstructions. Moreover, land-use harmonization schemes have been developed specifically for use in dynamic vegetation modelling in order to account for human-induced vegetation. Global circulation model simulations of past climates on millennial scales to study the land-use biogeophysical forcing are rare and have used mainly the KK10 and HYDE data sets to represent past LULCC. Regional climate-model simulations of past climates are more common, however still few and have mainly focused on Europe. Although climate-modelling studies to study land-use as a climate forcing in the past are still few and uncertainties are large, there is already a relatively large body of evidence from global and regional climate-model studies suggesting that land use is a significant climate forcing that should be further quantified, both in terms of response magnitude and spatiotemporal scale, to answer unresolved questions related to past climate change and to inform mitigation strategies of current and future anthropogenic climate change via land management. Future studies should focus on ensembles of global and regional climate simulations using multiple LULCC descriptions to better quantify uncertainties and on transient Holocene climate simulations with alternative LULCC descriptions. Improved data sets of past LULCC and an increase in their number, and implementation of these data sets in, for example, future phases of the Paleoclimate Modeling Intercomparison Project would increase researchers ability to employ lessons from the past and achieve more reliable climate predictions using alternative land-management scenarios to mitigate global warming.