Cloud Radiative Effects on MJO Development in DYNAMO

Abstract

Abstract Observed Madden–Julian oscillation (MJO) events are examined with the aid of regional model simulations to understand the role of cloud radiative effects in the MJO development. The importance of this role is demonstrated by the absence of the MJO in the model simulations that contain no cloud radiative effects. Comparisons of model simulations with and without the cloud radiative effects and observation help identify the major processes arising from those effects. Those processes develop essentially from heating in the upper troposphere due to shortwave absorption within anvil clouds in the upper troposphere and the convergence of longwave radiation in the middle to upper troposphere, with a peak at 300 hPa, during deep convection. First, that heating adds extra buoyancy and accelerates the rising motion in the upper troposphere in deep convection. The vertical acceleration in the upper troposphere creates a vacuum effect and demands for more deep convection to develop. Second, in response to that demand and required by mass balance arises the large-scale horizontal and vertical mass, moisture, and energy convergence. It strengthens deep convection and, with the feedback from continuing cloud radiative effect, creates conditions that can perpetuate deep convection and MJO development. That perpetuation does not occur however because those processes arising from the cloud radiative heating in the upper troposphere stabilize the troposphere until it supports no further deep convection. Weakening deep convection reduces cloud radiative effects. The subsequent reduction of the vacuum effect in the upper troposphere diminishes deep convection completing an MJO cycle. These results advance our understanding of the development of the MJO in the radiative–convective system over warm waters in the tropics. They show that while the embryo of intraseasonal oscillation may exist in the system its growth/development is largely dependent on cloud radiative effects and feedbacks.