The Atlantic Multidecadal Oscillation (AMO) is poorly understood -- long after Michael Mann coined the term AMO, he had second thoughts and has since asserted it may not actually be an oscillation. The actual variation caused by unpredictable external forces such as volcanic activity and perhaps man-made aerosols and GHGs.
The answer in what causes AMO is tidal cycles.
In the past, alternating "global warming" and "global cooling" have occurred as a result of glacial cycles, which have been explained as due to well-understood periodic orbital changes. These same orbital cycles control much geophysical behavior such as Earth's rotation rate perturbations (a la LOD) and this has a significant impact on Earth's climate via the inertial response of the ocean's thermocline depth. Currently GFD research is narrowing in on the response by modeling the effects of the long-period lunisolar cycles as a forcing stimulus.
The top-panel in the attached chart shows a cross-validated model of the AMO behavior based on tuning the standing-wave parameters on a training interval spanning the years 1900 to 2000. The regions outside this training interval are the test or validation intervals. So there is both a cross-validation for the years before 1900, and for the years after 2000.
The math analysis is not exactly easy to do as it requires calibration of the dozens of tidal factors to LOD and then iteration to map and scale to the data in the training interval. The assumption is that none of the AMO signal is noise, so no filtering of the time-series is involved. As a result of this modeling approach, it becomes easier to discriminate the natural variations due to orbital forcing and those due to AGW. And consider that this is only the beginning of GFD analysis as Google, NVIDIA and other computational teams are applying machine learning approaches to the data.
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology