Decomposition of the effects on regional climate from recent historical land cover changes in Europe

Abstract:

Land cover changes (LCC) show biophysical effect on regional climate because they modify the land surface albedo, evapotranspiration, and surface roughness. Many previous studies focused on the effects of individual land cover transitions, such as idealized large-scale scenarios of deforestation/afforestation or historical forest clearance, but the combined effects from the detected recent historical land cover changes in Europe have not been explored. In this study, we use a combination of a regional climate model (the Weather Research and Forecasting model, WRF, v3.9.1) with a high resolution land cover data to explore the effects on surface temperature of land cover changes between 1992 and 2015. Previous studies use one unrealistic large-scale simulation for each LCC to estimate its climate effects which present large variations especially in mid-latitudes. Our analysis introduces a new method simultaneously considering the effects of the mix of historical land cover changes in Europe and the individual one contribution. This approach, based on a ridge statistical regression, does not require an explicit consideration of the different components of the surface energy budget, and directly shows the temperature changes from each land transition.

Around 70 Mha of land transitions occurred in Europe from 1992 to 2015. Approximately 25 Mha of agricultural land was left abandoned, which was only partially compensated by cropland expansion (about 20 Mha). Declines in agricultural land mostly occurred in favor of forests (15 Mha) and urban settlements (8 Mha). Compared to 1992, we find that the land covers of 2015 are associated with an average temperature cooling of -0.12±0.20 °C, with seasonal and spatial variations. At a continental level, the mean cooling is mainly driven by agriculture abandonment (cropland-to-forest transitions). Idealized simulations where cropland transitions to other land classes are excluded result in a mean warming of +0.10±0.19 °C, especially during summer. Conversions to urban land always resulted in warming effects, whereas the local temperature response to forest gains and losses shows opposite signs from the western and central part of the domain (where forests have cooling effects) to the eastern part (where forests are associated to warming). Gradients in soil moisture and local climate conditions are the main drivers of these differences. Our findings are a first attempt to quantify the regional climate response to historical LCC in Europe, and our method allows to unmix the temperature signal of a grid cell to the underlying LCCs (i.e., temperature impact per land transition). Further developing biophysical implications from LCCs for their ultimate consideration in land use planning can improve synergies for climate change adaptation and mitigation.

Key words: land use/cover change; regional climate mode; biophysical climate; EURO-CORDEX