Hadley circulation (HC) is the zonally symmetric part of the tropical circulation. It consists of two zonally-averaged cells, which are positioned symmetrically about the Equator. A single cell consists of the ascending motion in the inter-tropical convergence zone (ITCZ), poleward flow in the upper troposphere, descending motion in the subtropical region [20°, 35°] and equatorward flow in the lower troposphere. Strong convective precipitation is present in the ITCZ, while descending motion in the subtropics is associated with precipitation deficit. Therefore, HC has a significant impact on the distribution of vegetation and population density and any changes of it could have significant damaging impacts on the biosphere. Studies involving climate model projections and reanalyses (retrospective analysis of past global weather) have already shown that the northern Hadley cell is widening poleward. On the other hand, climate model projections show a robust weakening of the northern Hadley cell in the 21st century, while reanalyses indicate its strengthening in the past 40 years.
In this thesis, the zonally-averaged Hadley cell is described in the vertical-meridional plane by a stream function and its strength is evaluated by taking average stream function value within a cell. Changes in HC strength are studied using modern ECMWF reanalyses. ERA5 reanalysis exhibits strengthening of northern and southern Hadley cell in the past 40 years, mostly in the winter hemisphere. Trends in ERA-Interim reanalysis are not statistically significant. The extended Kuo-Eliassen equation is derived and solved numerically to identify how different physical processes drive the changes in HC. Diabatic heating (condensational and radiative heating) and zonal friction are found to be contributing the most to the observed changes. The HC variability and trends are then compared to GPCP precipitation time series. Contrary to some recent studies involving reanalyses, we find that the trends of HC strength in ERA5 reanalysis are not an artifact of the misrepresentation of latent heating. Thus, HC changes are most likely a consequence of internal multidecadal climate variability.
|