An “offline” approach to DA is used, where static ensemble samples are drawn from existing CMIP climate‐model simulations to serve as the prior estimate of climate variables. We use linear, univariate forward models (“proxy system models (PSMs)”) that map climate variables to proxy measurements by fitting proxy data to 2 m air temperature from gridded instrumental temperature data; the linear PSMs are then used to predict proxy values from the prior estimate. Results for the LMR are compared against six gridded instrumental temperature data sets and 25% of the proxy records are withheld from assimilation for independent verification. Results show broad agreement with previous reconstructions of Northern Hemisphere mean 2 m air temperature, with millennial‐scale cooling, a multicentennial warm period around 1000 C.E., and a cold period coincident with the Little Ice Age (circa 1450–1800 C.E.). Verification against gridded instrumental data sets during 1880–2000 C.E. reveals greatest skill in the tropics and lowest skill over Northern Hemisphere land areas. Verification against independent proxy records indicates substantial improvement relative to the model (prior) data without proxy assimilation. As an illustrative example, we present multivariate reconstructed fields for a singular event, the 1808/1809 “mystery” volcanic eruption, which reveal global cooling that is strongly enhanced locally due to the presence of the Pacific‐North America wave pattern in the 500 hPa geopotential height field.

The Last Glacial Maximum (LGM), one of the best-studied paleoclimatic intervals, o↵ers a prime opportunity to investigate how the climate system responds to changes in greenhouse gases (GHGs) and the cryosphere. Previous work has sought to constrain the magnitude and pattern of glacial cooling from paleothermometers, but the uneven distribution of the proxies, as well as their uncertainties, has challenged the construction of a full-field view of the LGM climate state. Here, we combine a large collection of geochemical proxies for sea-surface temperature with an isotope-enabled climate model ensemble to produce a field reconstruction of LGM temperatures using data assimilation. The reconstruction is validated with withheld proxies as well as independent ice core and speleothem 18 O measurements. Our assimilated product provides a precise constraint on global mean LGM cooling of 5.9 C (6.3-5.6 C, 95% CI). Given assumptions concerning the radiative forcing of GHGs, ice sheets, and aerosols, this cooling translates to an equilibrium climate sensitivity (ECS) of 3.2 C (2.2-4.3 C, 95% CI), a value that is higher than previous estimates and but consistent with the traditional consensus range of 2-4.5 C.

Abstract. The Last Millennium Reanalysis (LMR) utilizes an ensemble methodology to assimilate paleoclimate data for the production of annually resolved climate field reconstructions of the Common Era. Two key elements are the focus of this work: the set of assimilated proxy records and the forward models that map climate variables to proxy measurements. Results based on an updated proxy database and seasonal regression-based forward models are compared to the LMR prototype, which was based on a smaller set of proxy records and simpler proxy models formulated as univariate linear regressions against annual temperature. Validation against various instrumental-era gridded analyses shows that the new reconstructions of surface air temperature and 500 hPa geopotential height are significantly improved (from 10 % to more than 100 %), while improvements in reconstruction of the Palmer Drought Severity Index are more modest. Additional experiments designed to isolate the sources of improvement reveal the importance of the updated proxy records, including coral records for improving tropical reconstructions, and tree-ring density records for temperature reconstructions, particularly in high northern latitudes. Proxy forward models that account for seasonal responses, and dependence on both temperature and moisture for tree-ring width, also contribute to improvements in reconstructed thermodynamic and hydroclimate variables in midlatitudes. The variability of temperature at multidecadal to centennial scales is also shown to be sensitive to the set of assimilated proxies, especially to the inclusion of primarily moisture-sensitive tree-ring-width records.

No abstract