Annual and Biennial Modulations in the Tropical Hadley Cell Circulation

Abstract

THE temperature in the vicinity of the tropical tropopause is lower in January than in July (ref. 1 and our unpublished results), the difference ranging to about 10° C in the 100–50 mbar layer (about 16–21 km). The low temperature depends on the calendar year rather than on mean zenith angle of the Sun, because it occurs in either January or February at tropical stations in both hemispheres (our unpublished results). Concomitantly the tropopause itself is higher at these times. The air near the tropopause is kept cold chiefly by the adiabatic cooling produced by rising motions in the tropical Hadley cell circulation², the contribution by horizontal motions carrying heat away from tropical regions being relatively small³. The adiabatic cooling is balanced by radiative heating, and Kennedy⁴ has shown that the main component here is a convergence of radiative flux in the 15 micron carbon dioxide band. At the tropopause itself this infrared component produces heating because the carbon dioxide both above and below is warmer. From either Kennedy's work or more recent radiative heating calculations by Rodgers⁵ the heating rate is about 0.25° C/day which, together with an observed horizontal heat flux divergence of about 0.05° C/day, leads to a rising motion necessary for balance³ of about 0.02 cm s⁻¹. Similar conclusions about the physical factors governing the tropopause temperature have recently been reached from a numerical model study by Manabe and Hunt⁶. An important point from Rodgers's results is that the total heating rate at the tropopause is essentially the same in January and July within the limits of accuracy of the calculations.