A recent paper by Chikira (2014) has changed the way I think about moisture mode theory and the Madden Julian Oscillation (MJO). If you’re not familiar with the MJO or moisture mode theory, then this post is gonna be over your head (sorry). But if you’ve been following the decades long search to explain what the MJO is and how it works, then I encourage you to read this paper.
One particularly interesting result of this paper has to do with the effect of a radiative heating anomaly on atmospheric moisture. These results require a special dynamical balance known as the weak temperature gradient approximation (WTG; Charney 1963; Sobel et al. 2001; Romps 2012). In short, the Tropics are known to have weak gradients of temperature and pressure, which is a consequence of the small Coriolis parameter near the equator.
To put this idea in context, let’s first take a look at some of Chikira’s analysis. Here is a composite of MJO specific humidity anomalies from ERA interim re-analysis data.
This next figure shows the moisture tendency by horizontal advection (top) and everything else (bottom), which Chikira refers to as the “column process” because it includes everything that works vertically in the column.
Notice that the advective tendency is always drying in the troposphere. In an anomalous sense, advection can moisten ~10 days prior to the “wet phase” of the MJO, which has been noted by several studies (ex. Maloney 2009). On the other hand, the column process is always moistening. Interestingly, the largest moistening by the column process occurs with the largest moisture anomalies. This suggests that the net moistening by the column process does not account for the buildup of moisture, but instead only acts to sustain a large-scale moisture anomaly. This way of thinking is very different from the “recharge-discharge” theory of the MJO (Blade and Hartmann 1993), and is more consistent with the theory of “moisture modes” (Raymond and Fuchs 2009; Sugiyama 2009).
The main goal of Chikira’s paper is to understand the processes that make up this moistening by the column process. His results show that part of the moistening can be explained by vertical circulations “induced” by radiative anomalies. This idea has actually been explored before, most notably by Sherwood (1999), who explored how the presence of cirrus clouds can result in an upward “pumping” of water vapor (see figure below). The essence of Sherwood’s pumping mechanism relies on the anomalous radiative heating that results from the reduction of longwave cooling in the upper atmosphere by the cirrus clouds.
Dr. Chikira was not aware of Sherwood (1999) when writing his paper (personal communication), but by taking advantage of WTG he was able to come up with an interesting parameter that describes the efficiency of this mechanism:
This parameter can tell us a lot about various mechanisms that induce large scale vertical velocity can impart a moistening tendency. But this idea can only work on timescales that WTG is relevant, so it is probably not that important for individual cumulus clouds.
When this alpha parameter is applied to the radiative heating anomalies associated with the active phase of the MJO (yellow line below) Chikira shows that the radiation helps contribute moisten the lower troposphere.
This idea might seem very counterintuitive, but people have long recognized the importance of radiative feedbacks in tropical convectively coupled disturbances (Bony and Emanuel 2005). With respect to moisture mode theory, it’s nice to have a clean way to relate radiative process and moisture tendencies. In my opinion this paper has a lot of great ideas and analysis, and I think it’s going to be a well-cited paper going forward, so it’s worth a careful read!