Re-examining inferences from Hadley cell theory on tropical expansion under global warming throughout the seasonal cycle

Simulations of global warming in numerical models ranging from full-complexity atmosphere-ocean global climate models (GCMs) to highly idealized, dry, atmospheric GCMs almost invariably feature poleward expansion of the annual-mean Hadley cell extent.  The attendant widening of the subtropical dry zones underlying the Hadley cell descending branches makes understanding this response of the large-scale circulation to climate change of paramount societal and ecological importance.  Two theories, one that neglects the role of large-scale eddy process and one that does not, yield similar but ultimately distinct dependencies of the Hadley cell width on planetary parameters, including those such as the equator-to-pole temperature gradient that also robustly change under global warming.  A common approach, therefore, is to use the responses of these parameters diagnosed from GCM simulations to make arguments about their influence on the Hadley cell widening.  This talk offers a critical examination of that approach.

The approach's key flaw is that the quantities such as the equator-to-pole temperature gradient that appear in the theoretical scalings refer to their values in the *absence* of any large-scale overturning circulation, Hadley cells or eddies, i.e. in the hypothetical state of latitude-by-latitude radiative convective equilibrium (RCE).  This RCE state is what "forces" the Hadley cells, and once the Hadley cells emerge they modify (among others) the equator-to-pole temperature gradient.  Using these theories to understand the Hadley cell response to increased CO2 therefore requires analyzing the responses of the hypothetical RCE state to the increased CO2, which we do via single column model simulations.  In addition, we present a new scaling for the Hadley cell extent applicable to the solsticial seasons that, unlike the existing scalings, does not depend sensitively on the presence or absence of large-scale eddies, which we use in conjunction with solsticial RCE simulations to clarify arguments regarding tropical expansion over the course of the annual cycle in addition to the annual mean.  The implications for these refined theoretical arguments on results from prior studies and on constraining future Hadley cell expansion are discussed.