Search for a command to run...
Climate variability over the last two millennia (Past2k) reflects the combined influence of internal climate dynamics, including modes such as ENSO, external forcings such as volcanic and solar activity, and, in recent centuries, increasing anthropogenic forcing. Studying the Past2k period provides critical context for understanding current and future climate change in a warming world. In the pre-instrumental era, the absence of direct climate observations necessitated reliance on climate models and paleoclimate archives to infer past climate conditions.Paleoclimate data assimilation (PDA) is an effective approach to integrate the information from both climate models and natural proxies. Here, we employ an offline PDA framework to generate a new spatiotemporally resolved global reconstruction of the Past2k at annual and seasonal (summer and winter) resolution, including key climate fields (e.g., temperature, pressure, precipitation), drought indices (scPDSI, SPEI), and climate indices (e.g., ENSO, PDO, AO).Proxy observations are compiled from the Iso2k, PAGES2k, CoralHydro2k, and SISALv3 databases, together with additional tree-ring width datasets. The model priors are derived from isotope-enabled Earth system model simulations using MPI-ESM-wiso and iCESM.Relative to previous global reconstructions, this assimilation framework advances PDA by explicitly assimilating isotopic proxy records, employing physically based proxy system models, accounting for proxy seasonality and climate sensitivity, and incorporating low-resolution records across multiple timescales. We conducted comprehensive evaluations of the reconstructed fields and indices against instrumental/reanalysis datasets, as well as existing global PDA-based reconstructions, and also assessed the reconstruction skill in earlier centuries. These evaluations show that our reconstructions perform well and yield reliable results. We expect that our datasets will provide a millennial-scale climate context to support further studies of past climate variability and to inform analyses of ongoing and future climate change.