— Ch. 1 · Defining The Oscillation —
Atlantic multidecadal oscillation.
~4 min read · Ch. 1 of 6
In 1980, researchers Folland and colleagues began noticing patterns in sea surface temperature data that suggested a decades-long cycle. Schlesinger and Ramankutty focused entirely on this oscillation by 1994. Michael Mann coined the term Atlantic Multidecadal Oscillation during a telephone interview with Richard Kerr in 2000. He later recounted this moment in his book The Hockey Stick and the Climate Wars published in 2012. Scientists define the AMO signal by removing linear trends from North Atlantic sea surface temperature variability. This detrending process attempts to isolate natural cycles from human-induced global warming. If the global warming trend is not perfectly smooth, it can leak into the AMO definition. Such leakage may cause correlations with the index to mask actual effects of climate change.
Measuring Climate Variability
Trenberth and Shea subtracted the global mean sea surface temperature from North Atlantic values to create an early version of the AMO index. They assumed global forcing affected the North Atlantic similarly to the rest of the ocean. Ting and colleagues argued that forced patterns were not uniform globally. They used signal-to-noise maximizing EOF analysis to separate internal variability from external forces. Van Oldenborgh and team averaged temperatures over the extra-tropical North Atlantic while regressing out global mean temperature changes. Guan and Nigam removed non-stationary global trends before applying their own EOF analysis to residual data. Linearly detrended indices suggest the late twentieth-century anomaly was split evenly between forced components and internal variability. Other methodologies indicate most of that anomaly resulted from external forcing rather than natural oscillations. Frajka-Williams noted in 2017 that recent cooling in the subpolar gyre created spatial gradients missed by standard AMO calculations.