The role of land cover in the climate of glacial Europe

Abstract

Abstract. Earth system models show wide disagreement when simulating the climate of the continents at the Last Glacial Maximum (LGM). This disagreement may be related to a variety of factors, including model resolution and an incomplete representation of Earth system processes. To assess the importance of resolution and land–atmosphere feedbacks on the climate of Europe, we performed an iterative asynchronously coupled land–atmosphere modelling experiment that combined a global climate model, a regional climate model, and a dynamic vegetation model. The regional climate and land cover models were run at high (18 km) resolution over a domain covering the ice-free regions of Europe. Asynchronous coupling between the regional climate model and the vegetation model showed that the land–atmosphere coupling achieves quasi-equilibrium after four iterations. Modelled climate and land cover agree reasonably well with independent reconstructions based on pollen and other paleoenvironmental proxies. To assess the importance of land cover on the LGM climate of Europe, we performed a sensitivity simulation where we used LGM climate but present-day (PD) land cover. Using LGM climate and land cover leads to colder and drier summer conditions around the Alps and warmer and drier climate in southeastern Europe compared to LGM climate determined by PD land cover. This finding demonstrates that LGM land cover plays an important role in regulating the regional climate. Therefore, realistic glacial land cover estimates are needed to accurately simulate regional glacial climate states in areas with interplays between complex topography, large ice sheets, and diverse land cover, as observed in Europe.

The climate and vegetation of Europe, North Africa and the Middle East during the Last Glacial Maximum (21,000 years BP) based on pollen data

“…Vegetation models driven by LGM climate model simulations generally indicate extensive areas of boreal forest north of the Alps, and a mix of temperate and warm-temperate woodland to the south across southern Europe and much of the Mediterranean. Treeless areas such as steppe are mainly confined to those areas where it is also found today, namely inland Iberia, Ukraine, southern Russia and Turkey, while Tundra is found to the north close to the Scandinavian Ice Sheet (Allen et al, 2010;Cao et al, 2019;Prentice et al, 2011;Velasquez et al, 2021).…”

“…Evaluation of these vegetation-model simulations against data has been largely based on comparison with compilations of pollen-biome reconstructions (Prentice et al, 2011;Allen et al, 2010;Cao et al, 2019;Velasquez et al, 2021). Early studies were based on only a limited number of sites from southern Europe, and showed steppe at all sites in contradiction with model simulations (Elenga et al 2000).…”

“…Further data-model discrepancies have also been highlighted concerning LGM vegetation cover. Earlier pollen synthesis studies, especially those that applied the biomisation method (Elenga et al, 2000) give the impression that non-glaciated areas of LGM Europe were dominated by treeless steppe, while vegetation models driven by climate model simulations indicate large areas of forest and woodlands (Binney et al, 2017;Kaplan et al, 2016;Velasquez et al, 2021). The apparent data-model discrepancy associated with steppe has led to the suggestion that early humans, which are not included in vegetation models, could have reduced the forest cover with only a relatively moderate use of fire because of the cold climate and slow speed of vegetation recovery (Kaplan et al, 2016).…”

“…The apparent data-model discrepancy associated with steppe has led to the suggestion that early humans, which are not included in vegetation models, could have reduced the forest cover with only a relatively moderate use of fire because of the cold climate and slow speed of vegetation recovery (Kaplan et al, 2016). This debate is important because of studies that have shown the sensitivity of the climate system to vegetation boundary conditions during the LGM (Ludwig et al, 2017;Velasquez et al, 2021). This suggests that https://doi.org/10.5194/cp-2022-59 Preprint.…”

Opening up the peer review process

“…3), with the criterion that the manuscript should be a research article. The manuscripts and articles, published with the Creative Commons Attribution 4.0 License, that the analysed text histories relate to are Blanchet et al (2021;acronym used in the examples TH1), Pawlak (2021; TH2) and Velasquez et al (2021;TH3). Analysing three manuscripts' text histories allows for comparison and contrast between the different manuscripts while still being a manageable amount of data to analyse.…”