Indications for a potential synchronization between the phase evolution of the Madden-Julian oscillation and the solar 27-day cycle

Abstract. The Madden-Julian oscillation (MJO) is a major source of intraseasonal variability in the troposphere. Recently, studies have indicated that also the solar 27-day variability could cause variability in the troposphere. Furthermore, it has been indicated that both sources could be linked, particularly, that the occurrence of strong MJO events could be modulated by the solar 27-day cycle. In this paper, we analyze whether the temporal evolution of the MJO phases could also be linked to the solar 27-day cycle. We basically count the occurrences of particular MJO phases as a function of time lag after the solar 27-day extrema in about 38 years of MJO data. Furthermore, we develop a quantification approach to measure the strength of such a possible relationship. and use this to compare the behavior for different atmospheric conditions and different datasets, among others. The significance of the results is estimated based on different variants of the Monte Carlo approach, which are also compared. We find indications for a synchronization between the MJO phase evolution and the solar 27-day cycle, which are most notable under certain conditions: MJO events with a strength greater than 0.5, during the easterly phase of the Quasi-biennial oscillation, and during boreal winter. The MJO appears to cycle through its 8 phases within 2 solar 27-day cycles. The phase relation between the MJO and the solar variation appears to be such that the MJO predominantly transitions from phase 8 to 1 or from phase 4 and 5 during the solar 27-day minimum. These results strongly depend on the used MJO index such that the synchronization is most clearly seen when using univariate indices like OMI in the analysis, but can hardly be seen with multivariate indices like RMM. A weaker dependence of the results on the underlying solar proxy is also observed. Although we think that these initial indications are already worth to be noted, we do not claim to unambiguously prove this relationship in the present study; neither in a statistical, nor in a causal sense. Instead, we challenge these initial findings ourselves in detail by varying underlying datasets and methods and critically discuss resulting open questions to lay a solid foundation for further research.