Should There Be a Universally Applicable MJO Index?

Abstract:

As documented in Straub (2013), different MJO indices may classify the same MJO-like event as “being” or “not being” an MJO. A good example is the period during late November and early December 1992, during TOGA-COARE, when wavenumber-frequency filtering of OLR identifies a clear MJO, but the RMM (Real-Time Multivariate MJO) index does not. More recently, several additional MJO indices have been developed, including the VPM (Velocity Potential-MJO) index and the OMI (OLR-based MJO) index. Each index was designed to be a universal index, such that it can be applied in all seasons and at all longitudes. However, each index is generally understood by the MJO community to most effectively capture a particular feature or set of features of the MJO, such as its large-scale envelope of cloudiness or its zonal wavenumber 1 dynamical structure, in some cases in a particular geographic region or during a particular season. MJO indices are also typically used in an on/off fashion: either something “is” or “isn’t” an MJO, which is often determined via a specific threshold, e.g. an amplitude of 1 in the RMM index. Straub (2013) and more recent research shows, however, that MJO-like structures do exist at index amplitudes smaller than the on/off thresholds. For example, composites of MJO-like convective events similar those seen during October through December 2011 during DYNAMO, which have longitudinally constrained convective envelopes that dissipate over the Maritime Continent and 20-30 day time scales, illustrate that these smaller-scale MJOs do not project strongly onto the RMM index and thus may be missed in composites using the commonly applied threshold of RMM=1. This research also opens up the larger questions of 1) whether the MJO should continue to be analyzed in this on/off fashion and 2) whether a universal MJO index is in fact the most effective way to analyze the MJO.