How does the air-sea coupling frequency affect the convection during the MJO passage?

The Madden Julian Oscillation (MJO) is a dominant variability in the tropics on the intra-seasonal scale, which is characterized by an eastward-propagating large-scale convective envelope. Nowadays, many operational atmosphere-ocean coupled models could successfully simulate/predict the MJO, while some studies also suggest that the coupling frequency affect the simulated MJO and its intensity. In this study, we intend to understand how the coupling frequency affects the convection during the passage of MJO and the possible mechanisms lying behind by case study of YMC-Sumatra 2017 campaign.

We performed numerical simulations using the COAWST Modeling System [4], which consists of the ocean (ROMS) and atmosphere (WRF) models at a 30-min, 1-hour, 3-hour, 6-hour, 12-hour, and 1-day coupling frequencies. The atmosphere-only (WRF-alone) run was also conducted for comparison. For model input and validation, we used the data from OISSTv2, GSMaP, GOFS 3.0, NCEP-FNL, NCEP-DOE reanalysis II, and the observations obtained from the campaign. The simulation period is Nov. 21st to Dec. 6th, 2017, when an MJO event develops and migrate over the Maritime Continent.

The figure shows the frequency–altitude distributions of the simulated relative humidity in the Sumatra region (90-110E, 10S-10N) during the active phase of MJO (Nov. 26^th - Dec. 6th). The lower troposphere (1000–800 hPa) was characterized by frequent occurrences of 80%–100% RH in all runs. At mid-troposphere, the atmosphere became much drier (the domination of 70% RH) in the runs with coupling frequency lower than twice per day (panel e-g), suggesting the convections were weakened. But, that’s not the case when the diurnal signal appeared. Vigorous convections enhanced the upward extension of the moisture, inducing the occurrences of high RH at mid-troposphere (700-400 hPa, panel d). On the other hand, it’s worthy to notice that when the coupling intervals was shorter than 3 hours, the distributions of RH did not change obviously (panel a-c).

As a next step, we would investigate how does the diurnal signal help the moistening of mid-troposphere and why the moistening processes were not further enhanced after 3-hourly coupling.