Index Intercomparison for MJO Event Termination

American Meteorological Society
99th Annual Meeting
Seventh Symposium on the Madden-Julian Oscillation and Sub-Seasonal Monsoon Variability

[Session Topic] Observational Analysis of Tropical Convection and Variability

Index Intercomparison for MJO Event Termination

Brett Chrisler¹ and Justin P. Stachnik¹

1. Department of Geography and Atmospheric Science, University of Kansas, Lawrence, KS, USA

The Madden-Julian Oscillation (MJO) is a significant contributor to climate variability in the tropics. Although many studies have been conducted regarding the MJO, few have considered its termination. While multiple indices exist to track and forecast the MJO, relying on different meteorological parameters, the details of each index to MJO decay have not been documented. Understanding the sensitivity of each index for MJO termination is necessary before future work determining precursor physical conditions associated with MJO decay.

This study presents an intercomparison of several MJO indices for a 34-year and select periods (e.g., DYNAMO and YOTC) in order to examine the evolution of MJO events and comprise an index-specific climatology for MJO termination. We present an analysis of common univariate indices including the OLR-based MJO Index (OMI) and filtered OMI (FMO) in addition to the Real-time Multivariate MJO (RMM) and velocity potential MJO (VPM) index. Normalized EOF comparisons are constructed that identify longitudinal shifts in the corresponding convection and circulation anomalies between indices. A time-series analysis is conducted to calculate maximum lag correlations of amplitude and phase. The absolute and fractional daily rate-of-change (ROC) of each index is derived to indicate which indices are subject to greater daily perturbations and extremes. We also calculate the ROC standard deviations and construct probability distribution functions for each index with an emphasis on the analysis of MJO weakening and decay.

Results from the ROC analysis are implemented in an event identification algorithm to study the sensitivity of MJO termination to changes in the buffer (i.e., temporary minimum amplitude) and threshold amplitude used to define MJO events. Furthermore, counts of primary, continuing, and circumnavigating events were recorded with primary events being insensitive to buffer changes, generally less than 5% among indices. Finally, an MJO climatology is created that presents MJO termination as a function of phase for each index. Physical reasons behind differences in the termination climatology are proposed in addition to applications for future work.