Reorganization of the North Atlantic Oscillation during early Holocene deglaciation

Abstract

The North Atlantic Oscillation is the dominant atmospheric pressure mode in the North Atlantic region and affects winter temperature and precipitation in the Mediterranean, northwest Europe, Greenland, and Asia1. The index1 that describes the sea-level pressure difference between Iceland and the Azores is correlated with a dipole precipitation pattern over northwest Europe and northwest Africa. How the North Atlantic Oscillation will develop as the Greenland ice sheet melts is unclear2. A potential past analogue is the early Holocene, during which melting ice sheets around the North Atlantic3, 4 freshened surface waters, affecting the strength of the meridional overturning circulation5. Here we present a Holocene rainfall record from northwest Africa based on speleothem δ18O and compare it against a speleothem-based rainfall record from Europe6. The two records are positively correlated during the early Holocene, followed by a shift to an anti-correlation, similar to the modern record, during the mid-Holocene. On the basis of our simulations with an Earth system model, we suggest the shift to the anti-correlation reflects a large-scale atmospheric and oceanic reorganization in response to the demise of the Laurentide ice sheet and a strong reduction of meltwater flux to the North Atlantic, pointing to a potential sensitivity of the North Atlantic Oscillation to the melting of ice sheets

Holocene evolution of the North Atlantic subsurface transport

“…Modeling studies (Gladstone et al, 2005) could not confirm a mid-Holocene NAO + mode and a new speleothem dataset reconstructed multi-decadal NAO +/− oscillations that were active over the entire Holocene (Wassenburg et al, 2016). Such a short-term variability is not resolved in our records and cannot account for the long-term subsurface transport changes.…”

“…These changes cannot be driven by insolation changes alone and seem to be strongly overprinted by the extent of the Laurentide Ice Sheet and meltwater surges, as shown in a modeling study using a fully coupled earth system model (Wassenburg et al, 2016).…”

“…In addition to multi-decadal NAO +/− oscillations, Wassenburg et al (2016) reconstructed early to midHolocene changes in the geometry of the Icelandic Low and Azores High pressure cells that were accompanied by a redirection of the westerlies to the southwesterly wind direction during the early Holocene (11-9 ka BP). These changes cannot be driven by insolation changes alone and seem to be strongly overprinted by the extent of the Laurentide Ice Sheet and meltwater surges, as shown in a modeling study using a fully coupled earth system model (Wassenburg et al, 2016).…”

“…Given the strong impact of the ice sheets and meltwater on the atmospheric circulation, as the main driver of the early to mid-Holocene changes (Wassenburg et al, 2016), we examine changes in meltwater and their timing more closely as a potential driver for changes in the subsurface transport.…”

“…In addition to the forcing by increases in atmospheric CO 2, four driving factors may be of importance for the observed Holocene changes in the Azores region: (1) changes in AC transport strength and position due to changes in Mediterranean Outflow Water (MOW) strength, (2) changes in the atmospheric circulation related to the northern hemispheric summer insolation (Renssen, 2005) and changes in North Atlantic Oscillation (NAO) patterns (e.g., Olsen et al, 2012;Wassenburg et al, 2016), (3) changes in meltwater forcing over the North Atlantic region due to melting of continental ice sheets (Mayewski et al, 2004), and (4) changes in solar activity (Bond et al, 2001). The latter will not be discussed due to the relatively low resolution of the late Holocene record, which hampers time series analyses.…”