Multidecadal variability in the North Atlantic
The North Atlantic (NA) region plays a key role in the global climate system and exhibits pronounced multidecadal climate variability. There is a controversy about the nature of the multidecadal climate variability in the NA region. This thesis provides an enhanced understanding about the multidecadal variability in the NA sector by applying and analyzing climate models and investigating observations, especially about the mechanisms for sea surface temperature (SST) and Atlantic meridional overturning circulation (AMOC) variability. Firstly, different driving factors operating on the extratropical and tropical NA SST on different timescales are investigated by using observations and model simulations. Secondly, a coupled air-sea multidecadal mode is discovered in the NA region by analyzing the fully coupled Kiel Climate Model (KCM). Thirdly, possible AMOC slowing is discussed by analyzing observational datasets and historical simulations with climate models. In the first part of this thesis, the different influences on the NA SST variability are examined based on observations and climate models. This analysis is conducted by using the basin-averaged NA SST index (NASST) and the low-pass filtered version, termed Atlantic multidecadal oscillation (AMO) index. In particular, the relationships of the two indices with some of its mechanistic drivers including Atlantic meridional overturning circulation (AMOC), North Atlantic Oscillation (NAO), subpolar gyre (SPG) and El Niño–Southern Oscillation (ENSO) are investigated. The results show that the NASST index lumps together SST variability driven by different mechanisms and operating on different timescales. Meanwhile, the AMO index emphasizes the SST variability over the extratropical NA, which is connected to the AMOC in climate models. In addition, models with a large cold bias over the NA exhibit a relatively weak linkage between the AMOC and AMO. The second part discusses the roles of ocean circulation and atmosphere-ocean coupling in the multidecadal climate variability in the NA region. A multidecadal mode is found in KCM, where both ocean circulation and atmosphere-ocean coupling are essential. A fast positive feedback and a delayed negative feedback are crucial for the multidecadal mode. The positive ocean-atmosphere feedback is between the SST and low-level atmospheric circulation over the Southern Greenland area. SPG and AMOC mutually influence each other and together provide the delayed negative feedback necessary for maintaining the oscillation. The stochastic heat-flux variability associated with the NAO keeps exciting the mode. A number of climate models predict that the AMOC will slow if the anthropogenic greenhouse gas emissions continue to rise unabatedly. However, there are debates about as to whether the AMOC is already slowing. The NA warming hole, which is a region where the SST cooled despite the global surface ocean warmed, has been suggested to be an indicator of anthropogenic AMOC slowing. The cooling may reflect diminishing AMOC-related northward upper-ocean heat transport. In the last part of the thesis, by using observational datasets since the beginning of the 20th century, the Atlantic SST variability linked to the net radiative forcing is identified and this pattern hardly accounts for any variance in the NA warming hole. In the NA warming hole, the so-called interhemispheric SST dipole pattern dominates that well records long-term internal AMOC variability in climate models. Furthermore, historical simulations with climate models in the ensemble mean only predict minor AMOC slowing. This part demonstrates the importance of natural internal AMOC variability in recent AMOC slowing and the need for systematic and sustained direct observations of the AMOC.