全球变化知识资源中心

Extremes in climate and weather can pose significant challenges to economy, ecosystems and human health. Changes in land cover are one of the drivers for variability in frequency and magnitude of extreme climate at regional and local levels. In this study, a regional climate model (COSMO-CLM v4.8) is used to simulate effects in climate extremes from two different idealized land cover change scenarios in Europe. These two simulations involve abrupt large-scale conversion of today forestland to herbaceous vegetation (deforestation), and of today cropland to evergreen needle-leave forest (afforestation). A control simulation with today land cover distribution is used to identify differences in extreme climate. We find significant changes in extreme climate in both deforestation and afforestation simulations, with seasonal and spatial differences. Deforestation causes a warmer summer (with higher annual maximum temperature) and a colder winter (with lower annual minimum temperature). Afforestation slightly increases the average intensity of the hot extremes, although with high spatial variability (a reduction is common in several locations), and mitigates cold extremes in winter. Changes in extreme indices show that deforestation increases both the frequency and duration of hot and cold extremes, while afforestation causes a lower frequency of extreme cold climate. The two simulations show opposing results in the number of frozen days, as they increase for deforestation and decrease for afforestation. A drier climate is found after deforestation, whereas a wetter climate is observed after afforestation. In general, deforestation and afforestation increase the frequency of hot extreme climate as they reduce the return period and increase the return level. Overall, our findings show the potential critical effects that land cover changes can have on climate extremes, and the possible synergies that land management strategies and planning can have for climate change mitigation and adaptation at a regional scale.