Formation of rain layers in the Indian Ocean and their feedbacks to atmospheric convection
December 05, 2022
Kyle Shackelford
Committee: Peter Jan van Leeuwen (Adviser); Charlotte DeMott (Co-adviser); Eric Maloney; Karan Venayagamoorthy (Civil and Environmental Engineering)
Abstract
Rainfall over the tropical warm pool spanning the Indian and West Pacific Oceans is relatively colder, fresher, and less dense than the near-surface ocean. Thus, under low-to-moderate winds, rainfall can act to stably stratify the upper ocean, forming a rain layer (RL). RLs cool and freshen the ocean surface and shoal ocean mixed layer depth, confining air-sea interaction to a thin, near-surface ocean layer. The shallow, transient nature of RLs has limited their observation, and RL impact on air-sea interaction is not well understood. This two-part thesis aims to address knowledge gaps surrounding 1) RL formation and characteristic traits, and 2) RL feedbacks to the atmosphere.In the first part of this thesis, we examine Indian Ocean RLs and their potential feedbacks to the atmosphere using a 1D ocean model. Initial experiments focus on model validation and demonstrate that the model is able to effectively replicate upper ocean response to precipitation as revealed by in situ measurements. Following model validation, Indian Ocean RL characteristics are studied by forcing a 2D array of 1D model columns with atmospheric output from an existing convection-permitting simulation. Results from this experiment demonstrate that SST reduction within RLs persists on time scales longer than those of the parent rain event. To evaluate RL feedbacks to the atmosphere, a second 2D array experiment is conducted over the same domain with identical atmospheric forcing except rainfall is set to zero at every time step. Comparison between simulations with and without rain forcing demonstrate that RLs reduce SST through cold rain input to the ocean surface, and maintain and enhance SST reductions through a stable salinity stratification. Through prolonged SST reduction, RLs also enhance spatial SST gradients that have previously been shown to excite atmospheric convection.
In the second part of this thesis, RL feedbacks to the Madden-Julian Oscillation (MJO) are studied by conducting regional ocean-atmosphere coupled simulations. Output from two convection-permitting coupled simulations of the November 2011 MJO event, one with rain coupling to the ocean surface and a second without rain coupling, is used to evaluate two potential RL feedback mechanisms. The first feedback is the “SST gradient effect,” which refers to RL-enhanced SST gradients imposing low-level ii patterns of convergence/divergence in the atmospheric boundary layer. The second is the “SST effect,” which refers to RL-induced SST perturbations altering turbulent heat fluxes. During the MJO transition from suppressed to enhanced convection, the SST gradient effect and SST effect have opposing feedbacks to convection, as RL-enhanced SST gradients favor convective initiation, while RL-induced SST reduction hinders convection. Comparison of coupled simulations with and without rain coupling to the ocean demonstrates that RL-induced SST reduction has a more substantial impact than enhanced SST gradients during this transitory phase. A delayed pathway in which RLs feedback to the MJO through the SST effect arises from frequent RL presence during the disturbed phase, which isolates subsurface ocean heat from the atmosphere. At the onset of the MJO active phase, westerly wind bursts erode near-surface RLs and release previously trapped subsurface ocean heat to the atmosphere, amplifying the intensity of MJO convection. Between the direct and delayed SST effect, RLs are shown to modify intraseasonal tropical variability.