Embedding a one-column ocean model in the Community Atmosphere Model 5.3 to improve Madden–Julian Oscillation simulation in boreal winter

Abstract

Abstract. The effect of the air–sea interaction on the Madden–Julian Oscillation (MJO) was investigated using the one-column ocean model Snow–Ice–Thermocline (SIT 1.06) embedded in the Community Atmosphere Model 5.3 (CAM5.3; hereafter CAM5–SIT v1.0). The SIT model with 41 vertical layers was developed to simulate sea surface temperature (SST) and upper-ocean temperature variations with a high vertical resolution that resolves the cool skin and diurnal warm layer and the upper oceanic temperature gradient. A series of 30-year sensitivity experiments were conducted in which various model configurations (e.g., coupled versus uncoupled, vertical resolution and depth of the SIT model, coupling domains, and absence of the diurnal cycle) were considered to evaluate the effect of air–sea coupling on MJO simulation. Most of the CAM5–SIT experiments exhibit higher fidelity than the CAM5-alone experiment in characterizing the basic features of the MJO such as spatiotemporal variability and the eastward propagation in boreal winter. The overall MJO simulation performance of CAM5–SIT benefits from (1) better resolving the fine vertical structure of upper-ocean temperature and therefore the air–sea interaction that results in more realistic intraseasonal variability in both SST and atmospheric circulation and (2) the adequate thickness of a vertically gridded ocean layer. The sensitivity experiments demonstrate the necessity of coupling the tropical eastern Pacific in addition to the tropical Indian Ocean and the tropical western Pacific. Coupling is more essential in the south than north of the Equator in the tropical western Pacific. Enhanced MJO could be obtained without considering the diurnal cycle in coupling.