Because of the orientation of the Earth with respect to the Sun, there is significantly more solar heat flux at the Equator than at the poles. There exists a temperature gradient between the Equator and the poles; however, the gradient would be much larger if there were no meridional heat transport. In order to reach the temperature gradient observed today, there needs to be a pole-ward flux of about 6 petawatts, which includes both atmospheric and oceanic heat transport (Marshal & Plumb 2008).
Figure 1 reveals the general circulation of the atmosphere. Near the equator there is uprising of hot air that moves to the low-pressure poles. The majority of this air sets at approximately 30°. When the descending air reaches the surface, some goes northward, and some goes southward. Because of the Coriolis force, air in the North Hemisphere is deflected to the right. As a result, pole-ward air starts to move eastward (known as the westeries in meteorological terms, because the air is moving from the west), while the equator-ward air moves westward (easterlies).
Particles moving pole-ward experience an increase in velocity because of the conservation of angular momentum. In Figure 2, there is a strip of air with velocity u and radius r. If this strip of air were to move upward, the radius would decrease and, in order to maintain angular momentum, the velocity would have to increase.
When air moving pole-ward gets to about 30°, the speed of particles gets to be too fast, and the strip of air becomes unstable, leading to a wavering strip of air as shown in Figure 1. Wind beyond this point to about 60° is very chaotic and unpredictable. The overturning circulation from the equator to around 30° is Hadley circulation, and the chaotic motion is eddy circulation. From this point on, the focus is on Hadley circulation.
Atmospheric Data of Hadley Circulation
In order to show Hadley circulation in the atmosphere, it is important to show several different atmospheric variables: vertical pressure velocity, potential temperature, and zonal and meridional wind. Following (Figures 3-10) are zonally averaged profiles of these variables averaged from 1948 to 2012 (ESRL). Examples will be given from both January and July. Firstly, Figure 3 and 4 are of vertical pressure velocity, measured in Pascals/second. The uprising occurs where heat is greatest, which results in greater uprising of particles on the side of the equator that is experiencing summer (January – Southern Hemisphere, July – Northern Hemisphere). The air then sets at around 30° on the winter hemisphere, and around 45° on the summer side. As will be seen in all the following atmospheric figures, Hadley circulation is stronger in the hemisphere experiencing winter.
Figures 5 and 6 are of the zonally average meridional wind in January and July respectively. Meridional wind is measured from north to south. Positive numbers represent northward motion. In Figure 5, the greatest winds are observed in the Northern Hemisphere, while it is experiencing a winter. Similarly, in Figure 6, the greatest winds are in the Southern Hemisphere.
As particles move pole-ward in the upper atmosphere, they get deflected (right in the Northern Hemisphere and left in the Southern). As discussed previously, particles gain zonal speed upon moving pole-ward. Figures 7 and 8 are of this zonal wind speed. The figures also show the trade winds between the equator and 30°, from the return flow at the surface being deflected westward as a result of the Coriolis force.
From the thermal wind relation, the greatest wind speeds are associated with the largest temperature gradient. In Figures 9 and 10, the steepest slopes are on the winter hemisphere. In July, there is a much larger difference between the zonal wind speeds of the two hemispheres than in January. This difference is likewise apparent in the July potential temperature data, for which the slope is very clearly steeper on one side.
The two key characteristics of the Hadley cell (from the equator to about 30°) are that there is large-scale overturning circulation, and that wind speeds increase in the pole-ward direction. The next goal was to simulate this in a laboratory environment to study the heat transport associated with Hadley circulation.
