The Link Between the Mixed Layer Depth in the Seychelles-Chagos Thermocline Ridge and the Initiation of the Madden-Julian Oscillation

The mechanism behind the initiation of the Madden-Julian Oscillation (MJO) remains one of the most challenging and least understood phenomena. There are many proposed mechanisms to explain its behavior, one of which is anomalously high sea-surface temperatures (SSTs) over the thermocline dome of the southern Indian Ocean, known as the Seychelles-Chagos Thermocline Ridge (SCTR). The SCTR is a hot spot for air-sea interactions on MJO time scales because of its high SSTs, shallow thermocline and shallow mixed layer. An analysis of an SCTR index, Real time Multivariate MJO (RMM) index, and lag correlations for high and low resolution Community Earth System Model (CESM) ocean and atmospheric data reveals inconsistencies between model and observational values. The model produces cooler SSTs, underestimates net surface heat flux, and produces a mixed layer depth (MLD) that is about two times deeper than observed values. The SCTR lies on the edge of the atmospheric convergence zone so small changes in oceanic components can lead to large changes in convection. This is seen in the model's RMM index composites of outgoing longwave radiation (OLR) as a proxy for convection. The spatial structure and strength of convection in the model is inconsistent with that of observations. Since MLD determines the thermodynamic heat budget for the ocean, and since the MLD and SSTs are connected by a feedback mechanism, it is hypothesized that the inconsistencies in MLD could contribute to the inconsistencies in the MJO. An analysis of lag correlations between the atmospheric and oceanic components associated with the MJO highlight how entrainment and heat flux tie into convection, illustrating how an error in one can propagate back onto itself, further propagating the initial error. A previous study using a slab ocean model (SOM) coupled to the Community Atmosphere Model (CAM) showed that intraseasonal SST and heat flux variance differ greatly depending on MLD. This project will perform experiments in which the observed MLD values are imposed over the SCTR to the SOM coupled to CAM in order to understand the impact that MLD spatial variability in the SCTR has on intraseasonal atmospheric variability, specifically MJO initiation.