Pollen-Based Climate Reconstructions in the Tibetan Plateau

Pollen-based climate reconstruction on the Tibetan Plateau uses fossil pollen assemblages preserved in sediments to infer past temperature, precipitation and vegetation dynamics. By comparing fossil spectra with modern calibration datasets, researchers establish transfer functions that translate pollen percentages into quantitative climate variables. These reconstructions reveal the timing and magnitude of Holocene monsoon fluctuations, the onset and retreat of glacial stages, and mid-Holocene moisture optima across different plateau sectors. Advances in high-resolution coring, extended modern pollen surveys and multivariate calibration techniques have refined spatial reconstructions from small alpine lakes to large inland basins. The resulting palaeoclimate records highlight the interplay of the Westerlies and Asian summer monsoon, document abrupt shifts linked to global teleconnections, and provide context for recent warming and water-resource management in this climatically sensitive region.

Research from Nature Portfolio

Recent methodological studies have tested machine-learning algorithms on fossil pollen assemblages to improve the fidelity of primary and secondary climate reconstructions. Ensemble approaches such as boosted regression trees demonstrate superior predictive power in reconstructing variables like winter temperature and water balance, while rigorous spatial cross-validation schemes minimise autocorrelation biases. Separate work on modern surface samples along altitudinal transects in eastern Tibet has clarified vertical pollen transport processes, showing that upslope wind dispersal can introduce lower-elevation taxa into high-altitude records. These insights underscore the need to account for taphonomic and transport biases when interpreting long-term pollen records across complex mountain landscapes.

Pollen-Based Climate Reconstructions in the Tibetan Plateau publication trend

The graph below shows the total number of articles in pollen-based climate reconstructions in the tibetan plateau across all publications each year (not limited to Nature Index journals).

Technical terms

Pollen assemblage: The composition and relative abundance of pollen types in a sediment sample, reflecting past vegetation communities.

Transfer function: A statistical model that relates modern pollen data to measured climate variables, used to estimate past climate from fossil pollen.

Weighted averaging partial least squares (WA-PLS): A multivariate calibration technique combining weighted averaging and partial least squares regression to develop robust pollen–climate relationships.

Pollen source area: The geographic region contributing pollen to a given sedimentary site, determined by dispersal and depositional processes.

Boosted regression trees: An ensemble learning method that builds and combines multiple decision trees sequentially to enhance predictive accuracy in climate-reconstruction models.

References

1. A modern pollen dataset from lake surface sediments on the central and western Tibetan Plateau. Earth System Science Data (2024).
2. Machine-learning based reconstructions of primary and secondary climate variables from North American and European fossil pollen data. Scientific Reports (2019).
3. Surface Pollen Distribution from Alpine Vegetation in Eastern Tibet, China. Scientific Reports (2017).
4. Modern pollen distribution in moss samples along an elevational gradient in southeast Tibet. Ecological Indicators (2023).
5. Palynological evidence reveals an arid early Holocene for the northeast Tibetan Plateau. Climate of the Past (2022).