PP44A-05
Straightening out Earth System Sensitivity Across the Late Pleistocene and Pliocene

Thursday, 17 December 2015: 17:00
2012 (Moscone West)
Peter J Huybers, Harvard University, Cambridge, MA, United States
Abstract:
There are good reasons to expect feedback strengths to alter with changes in background climate, but empirically constrainting such changes is difficult. Earth System Sensitivity (ESS) --- which accounts for both fast and slow feedbacks in the climate system, including those associated with ice sheets and vegetation --- is relatively well constrained for the Late Pleistocene because of accurate and well-resolved records of temperature and greenhouse gas concentration. The Pliocene offers the geologically most recent epoch during which to examine ESS in warmer climates, but empirical estimates have come to conflicting results. Estimates relying on mean changes in Pliocene temperature and CO$_2$ concentration find ESS values consistent with those from the Late Pleistocene, whereas more recent estimates from variability within the Pliocene find ESS values of half the magnitude, implying that ESS bends downward with increased global temperature. This set of observations might be physically self-consistent, for example, if within-Pliocene ESS reflects a smaller ice-albedo feedback, whereas average Pliocene temperature reflects warmer conditions brought about by a singular ocean-atmospheric transition resulting from higher greenhouse gas concentrations. Another possibility is that discrepancies result from errors in some or all of the ESS estimates. ESS may be overestimated if temperature changes are misattributed to greenhouse gases, when instead, for example, they are driven by changes in orbital variations or opening of oceanic passages. Alternatively, ESS may be underestimated because errors in forcing and misalignment of age models generally result in a statistical bias termed regression dilution. Here a method for optimally aligning age models is coupled with a spectral regression technique in order to account for age model uncertainty and delineate contributions from orbital variations. The method is applied to within-Pleistocene and within-Pliocene temperature and greenhouse gas variations, and preliminary results indicate lower Pleistocene ESS, higher Pliocene ESS, and greater uncertainties than previously indicated, all of which make ESS estimates across colder and warmer climates less inconsistent.