Global biome changes over the last 21,000 years inferred from model-data comparisons

• Main Authors: Li, Chenzhi, Dallmeyer, Anne, Ni, Jian, Chevalier, Manuel, Willeit, Matteo, Andreev, Andrei A., Cao, Xianyong, Schild, Laura, Heim, Birgit, Herzschuh, Ulrike
• Format: Text
• Language: English
• Published: 2024
• Subjects: Ice Sheet; Tundra
• Online Access: https://doi.org/10.5194/egusphere-2024-1862; https://egusphere.copernicus.org/preprints/2024/egusphere-2024-1862/

We present a global megabiome reconstruction for 43 timeslices at 500-year intervals throughout the last 21,000 years based on an updated and thus currently most extensive global taxonomically and temporally standardized fossil pollen dataset of 3,691 records. The evaluation with modern potential natural vegetation distributions yields an agreement of ~80 %, suggesting a high degree of reliability of the pollen-based megabiome reconstruction. With its high temporal and spatial resolution, this reconstruction is ideally suited for the evaluation of paleo-simulations from Earth System Models (ESMs). As an example, we compare the reconstruction with an ensemble of six different biomized simulations based on transient vegetation simulations performed by ESMs. The global spatiotemporal patterns of megabiomes estimated by the simulation ensemble and reconstructions are generally consistent, i.e., from glacial non-forest megabiomes to Holocene forest megabiomes, in line with the general climate warming trend and continental ice-sheet retreat. The shift to a global spatial megabiome distribution similar to today’s took place during the early Holocene. At a global scale over the last 21,000 years, the deviations between the reconstruction and the simulation ensemble are (a) largest during the Last Glacial Maximum and early deglaciation periods, mainly due to different estimates of tundra in the circum-Arctic areas and the Tibetan Plateau; and (b) moderate during the Holocene, mainly due to different estimates of non-forest megabiomes in relatively semi-arid zones such as North Africa and the Mediterranean that increases over time. To some extent, these mismatches could be attributed to systematic model biases in the simulated climate, as well as to the different plant representations and low taxonomic resolution of pollen in the reconstructions.