Ostracod assemblages and their relationship with water depth in surface sediments of Nam Co, Tibetan Plateau

Quantitative reconstruction of lake-level changes at Nam Co is essential for the understanding long-term environmental evolution and climate dynamics of the Tibetan Plateau. Ostracod-based transfer functions for water depth serve as valuable tools for quantitative lake-level reconstructions. However, their reliability at Nam Co has been constrained by limited spatial and depth coverage, insufficient robustness testing, and weak cross-validation with independent regional proxies. To address these limitations, we integrated ostracod assemblages from multiple regions and applied comparative modeling to construct a more robust transfer function for Nam Co. This function was then used to reconstruct lake-level changes since the mid–late Holocene. Our results indicate that ostracod samples collected in different years exhibit no significant compositional differences, allowing their consolidation into a comprehensive lake-wide calibration dataset and improving model representativeness. Reconstructions generated through various statistical approaches revealed broadly consistent lake-level trends, supporting the robustness of the findings. By applying a random forest transfer function based on the integrated dataset, we quantitatively reconstructed the lake-level fluctuations at Nam Co since the mid–late Holocene. The reconstruction reveals a sequence of humid–arid phases followed by a return to humid conditions, closely aligned with other regional paleoclimatic records. These results demonstrate the strong applicability of the model to Nam Co. Further analysis suggests that these fluctuations were likely influenced by the combined effects of the Indian Summer Monsoon, glacial meltwater, and Westerly air masses.