The Holocene Thermal Maximum as an Analog for Future Warming: Insights from Paleoclimate Data Assimilation

The Holocene thermal maximum (HTM; 9-5 kyr BP) in Greenland has been estimated to be 2-3°C warmer than present, similar to the 1-4°C of warming projected over the next 100 years. Though the forcings contributing to this warming differ for each time period, the HTM may provide valuable insight into how the Greenland ice sheet will respond to future climate changes. Here we reconstruct the spatial pattern and timing of the HTM over Greenland with the intent to: 1) determine the reliability of the HTM as an analog for future warming, and 2) provide skillful climate reconstructions to the modeling community, with a focus on ice-sheet modeling. We use paleoclimate data assimilation, a novel framework developed under the Last Millennium Reanalysis Project (Hakim et al., 2016), which reconstructs past climate by optimally combining information from paleoclimate proxy records and climate model simulations. This data assimilation approach has several advantages over other paleoclimate reconstructions commonly used in ice-sheet modeling, including 1) allowance for non-stationary teleconnections through time, 2) unlimited incorporation of proxy records, 3) easily-calculated uncertainty estimates, 4) physically-consistent reconstruction of multiple climate variables, and 5) minimal computational cost. Our inputs include oxygen-isotope and accumulation data from long Greenland ice cores and two different transient Community Climate System Model simulations. Our outputs are spatially-complete reconstructions of two-meter air temperature, accumulation, and sea surface temperature for 11-0 kyr BP. Results show agreement with independent sources, including a recent lake sediment core in northwest Greenland that indicates summer HTM temperatures 4-7 °C warmer than present (McFarlin et al., 2018). The reconstructed HTM warming has a distinct North-South gradient that is found in some modern warming trends derived from reanalysis datasets, including Berkeley Earth and Last Millennium Reanalysis. This finding lends support to the idea that the HTM is a useful analog for future climate forcing of the Greenland ice sheet.