Research Topics

The Madden-Julian Oscillation (MJO), the dominant mode of intraseasonal variability in the tropics, consists of convective anomalies coupled with planetary-scale atmospheric circulation anomalies. The convection-circulation couplet associated with the MJO tends to propagate eastward over the Indo-Pacific warm pool at a speed of about 5 m s-1. It influences the surface climate not only of the tropics but also of the midlatitudes via teleconnections, serving as a major source of predictability on the sub-seasonal time scales.

Since the influence of the MJO on global weather and climate, studies to simulate the MJO in the general circulation models (GCMs) have been advanced to improve predictability on seasonal to sub-seasonal time scales. From the development of the MJO simulations in the GCM, several challenging questions about the characteristics of the MJO have been raised.

The representative challenging question concerns the propagation, maintenance and initiation of the MJO. In a 'canonical' MJO event, anomalously enhanced convection develops in the western Indian Ocean and propagates eastward across the Maritime Continent (MC) until it reaches the central Pacific, where the MJO weakens. However, not all MJOs are same. The half of the MJO initiated from the Indian Ocean decays in the MC, where the disruption to MJO propagation is known as the 'MC barrier effect'. Furthermore, the MJO propagation characteristics such as the MJO propagation speed and its zonal extension show the interannual variability. The different MJO propagation characteristics have been investigated by dividing the MJO events into two or three types and further extended to the concept of 'MJO diversity'.

The MJO diversity concept explains the MJO events in four different types: i.e. the standing, jumping, slow and fast propagating MJOs. The mechanisms for the different MJO types could be explained by the background conditions, especially the El Nino/Southern Oscillation (ENSO)-like background conditions. The possible relationships between the MJO types and the SST anomalies have been suggested in observational studies: the standing (fast) MJO types may be related to the La Nina (El Nino-like) SST anomalies. However, none of the studies assess the diversity of the MJO in the GCMs. The extent to which ENSO can stochastically modify the MJO types has not yet been investigated.

I am interested in these topics

1) The performance of simulating the MJO diversity in CMIP6 models.

2) How much ENSO can stochastically change the type of MJO?

3) The future changes in MJO types in global warming scenario.

2. Stratosphere-Troposphere Coupling: QBO-MJO connection

Recent studies have revealed that the MJO is modulated by the Quasi-Biennial Oscillation (QBO), which is the dominant interannual variability in the tropical stratosphere. The QBO index, defined as the zonal-mean zonal wind at 50 hPa averaged over 10°S–10°N, shows a significant negative correlation with the MJO index in boreal winter. In general, the MJO convection during the easterly phase of QBO (EQBO) winters is stronger than the one during the westerly phase of QBO (WQBO) winters. Moreover, the MJO persists longer and propagates slower during EQBO winters than WQBO winters.

There are no clear mechanisms for the QBO-MJO connection, but a few hypotheses for the QBO-MJO connection have been suggested. One is the static stability changes in the upper troposphere/lower stratosphere (UTLS). In the zonal-mean sense, the anomalous QBO-related wind creates wind shear and adiabatic temperature changes in the UTLS to satisfy the thermal wind balance. In the localized region, the cold anomalies in the UTLS can be induced by the MJO activity. The other is the cloud-longwave radiation feedback in the whole troposphere. The cloud-longwave radiation feedback is slightly stronger in the EQBO winter than in the WQBO winters. Since the stronger longwave radiation feedback in the EQBO winters, the energy loss from the atmospheric column is smaller during the EQBO winters.

To elucidate the exact mechanism of the QBO-MJO connection, modeling studies are needed. Although the QBO-MJO connection has been well documented in several observational studies, many GCMs fail to capture the QBO-MJO connection. The development of the simulating QBO-MJO connection in the GCM can provide us with knowledge about its mechanisms, which is similar to the MJO development in recent decades. To understand the QBO-MJO connection, we have to know how it can be simulated.

The most challenging question is what the clear mechanisms of the QBO-MJO connection are. First, the modeling of the QBO-MJO connection is required to clarify the mechanisms. I conducted the QBO-nudging experiments in WRF regional models in Back et al. (2020). This study can provide a hint of the QBO-MJO connection in high resolution models without cumulus parameterization. The QBO nudging experiments in GCMs have been investigated, but they fail to reproduce the QBO-MJO connection. From the GCM, we can examine the mechanism denial experiment to investigate why the GCM fails to simulate the QBO-MJO connection.  

I am interested in these topics

1) The mechanism denial experiments for the QBO-MJO connection

2) QBO nudging experiment in the high-resolution models

God created us, he knows everything about us.

We created GCMs, shouldn't we know everything about them?

-Adam Sobel