Model outputs for the estimated ocean states, link to files in NetCDF format@en

The global ocean state for the modern age and for the Last Glacial Maximum (LGM) was dynamically reconstructed with a sophisticated data assimilation technique. A substantial amount of data including global seawater temperature, salinity (only for the modern estimate), and the isotopic composition of oxygen and carbon (only in the Atlantic for the LGM) were integrated into an ocean general circulation model with the help of the adjoint method, thereby the model was optimized to reconstruct plausible continuous fields of tracers, overturning circulation and water mass distribution. The adjoint-based LGM state estimation of this study represents the state of the art in terms of the length of forward model runs, the number of observations assimilated, and the model domain. Compared to the modern state, the reconstructed continuous sea-surface temperature field for the LGM shows a global-mean cooling of 2.2 K, and the reconstructed LGM ocean has a more vigorous Atlantic meridional overturning circulation, shallower North Atlantic Deep Water (NADW) equivalent, stronger stratification, and more saline deep water.本研究采用先进的数据同化技术，对现代以及末次冰盛期（Last Glacial Maximum，LGM）的全球海洋状态进行了动态重建。研究整合了海量观测数据，包括全球海水温度、盐度（仅用于现代海洋状态估算），以及氧、碳同位素组成（末次冰盛期仅采用大西洋海域数据），结合伴随方法将上述数据同化至海洋环流模式中，通过模式优化重建出合理的示踪物连续场、翻转环流及水团分布。本研究基于伴随方法的末次冰盛期状态估算，在正演模式运行时长、同化观测样本量以及模式覆盖区域维度上均达到当前国际领先水平。相较于现代海洋状态，重建得到的末次冰盛期海表温度连续场呈现出2.2 K的全球平均降温；重建的末次冰盛期海洋还具备更强盛的大西洋经向翻转环流、等效北大西洋深层水（North Atlantic Deep Water，NADW）埋深更浅、层结更强，且深层水盐度更高。