ENSO-driven hydroclimate changes in central Tibetan Plateau since middle Holocene: Evidence from Zhari Namco’s lake sediments

Abstract

The Tibetan Plateau, often referred to as the “Asian Water Tower”, holds immense significance as a critical water source for billions of people in surrounding regions. Its unique location and extreme environmental conditions contribute to the formation of some of the world largest lakes, crucial components of the regional water cycle. These lakes not only store vast freshwater resources but also play a vital role in regulating the flow and quality of rivers, maintaining a delicate balance in the local ecosystem. Recent decades have witnessed rapid expansion of many plateau lakes, primarily attributed to glacial melt. However, evaluating the contribution of climate forcings to these changes remains challenging. In this study, we reconstructed Zhari Namco's lake-level changes between 5.3 ka and 1.3 ka based on optically stimulated luminescence dating, a period with relatively stable glacial amount on the Tibetan Plateau. Our analysis employed sediment grain size distribution, organic matter contents and their stable carbon isotopes. In contrast to recent expansions, our findings reveal a declining trend in Zhari Namco's lake level since the middle Holocene, consistent with intermittent paleo-shoreline records from major central Tibetan Plateau lakes. This regional decline is likely the result of a weakening of the Indian Summer Monsoon. In addition to the long-term drying trend, we identified high-frequency fluctuations in Zhari Namco's lake level on centennial timescale. Further comparison with hydroclimate records from the tropics suggests that centennial high lake levels correlate with increased monsoonal precipitation during periods when ENSO activities were reduced. This study sheds lights on the historical dynamics and driving mechanisms of Zhari Namco's lake levels, providing valuable insights for predicting the impacts of future climate change on the Tibetan Plateau's water cycle.