Paleoclimate data assimilation: Principles and prospects

Reconstructing climate states during geological periods is a key focus in studying Earth system evolution. Substantial progress has been achieved in reconstructing paleoclimate and paleoenvironment using both Earth system models and paleoclimate proxies. However, current paleoclimate reconstructions face several challenges: the accuracy of Earth system model simulations relies on model parameter settings. Paleoclimate proxy data exhibit significant variability across different periods and regions, and proxy data are often sparse, hindering the accuracy and global relevance of proxy-based reconstructions. Addressing the pros and cons of these methods, paleoclimate data assimilation can effectively integrate Earth system models and paleoclimate proxy data, enhancing the precision and global relevance of reconstructions. Using approaches such as the ensemble Kalman filter as an example, this paper introduces the principles of paleoclimate data assimilation and reviews recent advancements in reconstructing paleoclimate states using these techniques. Paleoclimate data assimilation offers new insights and advanced techniques for paleoclimate reconstruction, with potential applications extending to the entire Cenozoic, Mesozoic, and even Paleozoic eras. These applications could deepen our understanding of the past climatic backgrounds of extreme climate events such as glacial-interglacial cycles, hyperthermals, and oceanic anoxic events, providing a reference for predicting future climate change.