Abstract The large scale tropical circulation, commonly named the Hadley circulation, is a key element in the global heat and moisture transport. Traditionally it is defined as the meridional circulation of the zonally averaged flow in the tropics, but in recent years studies have shown the importance of looking at the decomposition of the three-dimensional atmospheric flow into local meridional and zonal circulations. These studies gave useful analysis on the regionality and variability of the meridional circulation in different time scales, but were mostly limited to examining the regional strengthening/weakening of the circulation. Here we study the interannual variability of the longitudinally-dependent meridional circulation (LMC), with a focus on its zonal shift. We use hierarchical clustering to objectively determine the 5 main modes of the LMC interannual variability, and apply a Lagrangian air parcel tracking method to reveal the detailed patterns of the circulation. We find that the most prominent interannual variability of the LMC is an east-west shift, which plays a dominant role in the overall interannual variability of the tropical circulation. In addition, the LMC variability is found to be strongly related to other atmospheric variables such as the sea surface temperature, precipitation and air temperature. Using multiple linear regression we analyze these dependencies and discuss their implications for the tropical climate system. We also relate the LMC interannual variability to the Madden-Julian Oscillation (MJO) and find that the 2 La-Nina related modes are significantly correlated with 2 different MJO phases.

In the present-day climate, cold air outbreaks occur when marine air intrudes over high-latitude continental interiors and radiatively cools, producing an abrupt drop in surface air temperature to as low as -40 C. But during the Eocene warm climate period, 55 million years ago, the presence of frost-intolerant species even at high latitudes in the Northern Hemisphere indicates that cold air outbreaks were suppressed. In projected future climate scenarios, relatively high surface temperatures at high latitudes are predicted as part of polar amplification. The lapse rate “feedback”, corresponding to enhanced warming of the lower troposphere, was found to be a major contributor [1]. The suppression of cold air in the Eocene is not well reproduced in global climate models (GCM) and the lapse rate feedback that contributes to polar amplification is still not well understood. Recent work hypothesized that the formation of low clouds as moist air flows from a warm ocean to a cold continental surface could suppress cold air outbreaks in warmer climates. Cronin and Tziperman, 2015, took a one-dimensional Lagrangian column model approach to track cloud formation and surface temperature as an air column migrates from a warm ocean surface to a cold continent [2]. Hu et al, 2018, followed up with an Eulerian analysis of GCM output over a range of cold and warm climates, looking at regional cloudiness, continental interior temperatures, and cold air extremes [3]. But neither approach is complete. The Lagrangian column model does not take into account mixing with surrounding air masses, while the Eulerian analysis does not explicitly follow the formation of clouds and their radiative impact as an air mass moves. In this work, we combine the two perspectives by studying cold air outbreaks in a variety of warm and cold climate scenarios using model output from the Community Atmosphere Model. After identifying cold air outbreaks, we backtrack trajectories for the air parcels that make up the entire cold air column. We then analyze the formation of clouds and the radiative budget to study the effects of clouds along each trajectory. Pithan, F. & Mauritsen, T. (2014). Nat Geo, 7, 181-184. Cronin, T. W. & Tziperman, E. (2015). PNAS, 112(37), 11490-11495. Hu, Z., Cronin, T. W. & Tziperman, E. (2018). JCLI, 31(23), 9625-9640.