Quantifying Convective Aggregation using the Moist Tropical Margin’s Length

A peculiar feature of the tropical atmosphere is its tendency to be either rather moist or very dry. It has

been shown that the two states are separated by a sharp margin, well identified by the anti-mode of

the bimodal tropical column water vapor distribution (CWV). Despite recent progress in understanding

physical processes governing the spatio-temporal variability of CWV near the Equator, the behavior of

this meander remains elusive, and we lack a simple framework to understand the bimodality of tropical

CWV in observations. A tendency of the atmosphere to be either moist or dry can also be found in

idealized simulations of radiative-convective equilibrium (RCE). A number of studies assessing self-

aggregation of convection in RCE simulations have indicated that positive feedbacks between

radiation and moisture lead to spatially concentrated moisture in a small area of the computational

domain. As a result, RCE simulations also exhibit a bimodal distribution of CVW and a distinct moist

margin.

Motivated by theory originating from statistical physics, we argue that, in steady-state RCE, the length

of the margin separating moist and dry regions should minimize. Based on these ideas, we design a

diagnostic relating the moist margin length to the circumference of a defined equilibrium shape. Using

that Moist Margin Length Index (MMLi), we assess the evolution of self-aggregation in idealized

convection-resolving RCE simulations and contrast it to the oscillations of the Atlantic Inter-Tropical

Convergence Zone in the ERA5 reanalysis product.

We find that MMLi successfully describes aspects of convective organization not captured by more

traditional metrics of convective organization (e.g., the CVW's interquartile range) while offering a new

way to characterize the seasonal cycle of convective organization in the Atlantic ITCZ. Overall, our

new framework uses the moist meander margin to connect the observationally-motivated, object-

oriented view of convective organization with the processes driving the spatio-temporal evolution of

the CWV field. It can be broadly deployed across models and observations to lead further insight into

the bimodal nature of CWV while offering an efficient and visual way to quantify convective

organization.