Biome reconstruction on the Tibetan Plateau since the Last Glacial Maximum using a machine learning method

Historical biome changes on the Tibetan Plateau provide important information that improves our understanding of the alpine vegetation responses to climate changes. However, a comprehensively quantitative reconstruction of the historical Tibetan Plateau biomes is not possible due to the lack of quantitative methods that enable appropriate classification of alpine biomes based on proxy data such as fossil pollen records. In this study, a pollen-based biome classification model was developed by applying a random forest algorithm (a supervised machine learning method) based on modern pollen assemblages on and around the Tibetan Plateau, and its robustness was assessed by comparing its results with the predictions of the biomisation method. The results indicated that modern biome distributions reconstructed using the random forest model based on modern pollen data generally concurred with the observed zonal vegetation. The random forest model had a significantly higher accuracy than the biomisation method, indicating the former is a more suitable tool for reconstructing alpine biome changes on the Tibetan Plateau. The random forest model was then applied to reconstruct the Tibetan Plateau biome changes from 22 ka BP to the present based on 51 fossil pollen records. The reconstructed biome distribution changes on the Tibetan Plateau generally corresponded to global climate changes and Asian monsoon variations. In the Last Glacial Maximum, the Tibetan Plateau was mainly desert with subtropical forests distributed in the southeast. During the last deglaciation, the alpine steppe began expanding and gradually became zonal vegetation in the central and eastern regions. Alpine meadow occupied the eastern and southeastern areas of the Tibetan Plateau since the early Holocene, and the forest-meadow-steppe-desert pattern running southeast to northwest on the Tibetan Plateau was established afterwards. In the mid-Holocene, subtropical forests extended north, which reflected the optimum condition. During the late Holocene, alpine meadows and alpine steppes expanded south.

Automated summary

This study developed a pollen-based biome classification model using a random forest algorithm, which proved to be more accurate and suitable for reconstructing alpine biome changes on the Tibetan Plateau compared to the biomisation method. The reconstructed biome distribution changes corresponded to global climate changes and Asian monsoon variations.