Low-frequency cloud-radiation interactions

Abstract

The observed 30 to 60 day oscillation of the tropical winds, often referred to as the Madden Julian (MJ) oscillation, has excited lots of interest in recent years, because its existence suggests the possibility that the behavior of the atmosphere is at least partially predictable on such relatively long time scales. Most theories of the MJ oscillation have been based on the wave instability theories, such as wave-CISK (conditional instability of second kind). As discussed by Hu and Stevens (HS), these theories have not yet satisfactorily explained the observations. Recently, HS have suggested that the MJ oscillation is actually a forced response to periodic heating. They suggest that the forced oscillations originates in the hydrologic cycle, without the active participation of large scale dynamics. According to their model, the oscillation originates in a progressive build up of atmospheric water vapor, which continues until a (prescribed) threshold is reached, after which precipitation begins and rapidly dries the atmosphere. The drying due to a precipitation episode is followed by renewed gradual moistening, and the cycle continues indefinitely in this way. Preliminary results are presented from a 1-D version of the Randall et al. GCM, which lends further support to the ideas of HS. The present 1-D results also differ in important ways from those of HS, however. In particular, cloud radiation effects are essential for the oscillatory behavior of the model, although they are not essential in the model of HS.