Probably referring to the Atlantic Multidecadal Oscillation (AMO). This is a mode of N. Atlantic sea surface temperature variability that is thought to be related to northward ocean heat transport; for the past decade or so it's been positive (i.e. relatively warm N. Atlantic) but in 2009 it started shifting to it's negative phase. That could def. influence the NAO state on decadal time scales (as could, say, having extremely reduced summer sea ice cover in the Arctic).