| Summary: | Many past warm periods exhibited greatly reduced latitudinal temperature gradients as a result of amplified Arctic surface temperatures as well as more seasonably equable temperatures. The Pliocene is a period of particular interest because CO2 forcing was comparable to today and yet Arctic temperatures were significantly warmer than today. Here we describe an atmospheric general circulation model experiment assessing the response of terrestrial temperatures in the mid-Pliocene (3.02 to 3.26 Ma) to an ice-free Arctic, and we compare the simulation with a compilation of proxy-based Pliocene paleotemperature reconstructions. Our experiments indicate that the amplification of Arctic surface temperatures is much more sensitive to the extent of sea ice than continental ice. The removal of Arctic sea ice results in simulated mean annual surface temperatures that better match terrestrial proxy data (RMSE = 2.9 °C) than experimental conditions that included seasonal sea ice (RMSE = 4.5 °C). Our simulations also show a decrease in the seasonal amplitude of temperatures in the absence of sea-ice, which is consistent with theory predicting more equable climates in the Arctic during warmer intervals in Earth's history. Our results demonstrate that once sea-ice is removed, latent heat is lost from the ocean to the atmosphere as water vapor that can be circulated by the atmosphere,which results in warming of continental interiors. Although our sensitivity experiment does not help to identify the full array of feedback mechanisms responsible for the amplification of Arctic surface temperatures during the Pliocene, it does demonstrate that Arctic terrestrial surface temperatures are extremely sensitive to the spatial and seasonal extent of sea-ice.
|
|---|