Moist convection and the large scale tropical circulation

The interaction of moist convection and the large scale tropical circulation is explored through a combination of observational analysis, simplified quasi-analytic modeling and numerical modeling. These three strategies separate this thesis into three distinct parts, unified by their insights into the coupling of diabatic forcing, convection, and dynamics.; In the first part, the relationship between water vapor path W and surface precipitation rate P over the tropical oceans is analyzed using 4 years of gridded daily SSM/I satellite microwave radiometer data. A tight monthly mean relationship for all tropical oceans and seasons is found between P and a column relative humidity r. A similar relation also holds for daily time scales and can be interpreted as a moisture adjustment time scale of 12 hours for connective rainfall to affect humidity anomalies on 300-km space scales. Cross spectral analysis shows statistically significant co-variability of actual and r-predicted precipitation at all frequencies, with negligible phase lag.; In the second part, cloud-water-radiative feedbacks between precipitation, atmospheric temperature, ocean temperature and circulation strength are studied with an idealized model of the steady Walker circulation in a non-rotating atmosphere coupled to an ocean mixed layer. Due to their strong top-of-atmosphere radiative cancellation, high clouds are found to have little overall effect on the circulation strength and connective area fraction. Low clouds are found to cool both the ocean and atmosphere and to shrink the convective area significantly. The climate is found to be less sensitive to doubled greenhouse gas experiments with low clouds than without.; In the third part, this analysis is extended to a fully three dimensional, rotational framework. A diabatically accelerated cloud resolving model is coupled to an aqua planet beta plane over an ocean mixed layer. It is found that the classical view underlying the simplified model of net export of moist static energy due to vertical motions and small horizontal advection may need to be re-evaluated. In addition, the second baroclinic mode is found to be of leading order importance for the time-mean circulation suggesting that its inclusion into refined conceptual models is necessary.