Madden-Julian and snow

Also see:  Madden-Julian and snow #2

The Madden-Julian oscillation (MJO) is a fairly regular eastward-propagating pulse of equatorial cloud and rainfall affecting our region, named after the two US meteorologists who identified it in the 1970s. It usually repeats about every 30 to 60 days. While the effect is mostly tropical, it does correlate with our alpine snow in some phases, in ways that may be useful for long-range forecasting.

MJO is described in terms of an intensity (“amplitude”) and location (“phase”). The phase locations look like this (linked images; click for sources):

MJO phase vs cloudiness (negative OLR)

MJO phase vs cloudiness¹
(negative OLR anomalies — blue)

MJO phase regions

Approximate MJO phase regions²

The long record of amplitude looks like this:


There is no obvious long term trend or periodicity. Amplitudes less than 1.0 are treated as “weak”, meaning little MJO effect (cutoff line shown — not a trend line). For amplitudes greater than one, the correlation with winter precipitation in south-eastern Australia is noticeably positive in phases 5 and 6 and negative in phase 2 (barely) and phase 3 (linked BOM image; click for source):

The correlation is relatively strong with the season peak snow depth at Spencers Creek (midway between Perisher Valley and Thredbo, from Snowy Hydro Limited):

Spencers Creek MJO correlation

Spencers Creek MJO correlation

Unfortunately the MJO is not predictable on any sort of moderate time horizon, so that correlation is not useable in my season prediction model. The recent amplitude-phase MJO record is as follows (apologies to the deuteranopics³):


There was a major spike of MJO activity in March — mostly in phases 7 and 8 way out in the Pacific, which may have prodded the developing El Niño. It’s a long while since we’ve seen strong MJO phases 5 or 6 in winter; perhaps we’re overdue.

The Australian Bureau of Meteorology of course tracks MJO closely. They plot the recent MJO record somewhat confusingly on a rotary phase or “path” plot, like this (linked chart will update; click for source):

The position around the chart gives the phase (the numbered sectors) and the distance from the centre gives the amplitude. The centre circle is at amplitude 1.0, so positions within it indicate weak MJO activity. At the moment the MJO is very quiet.

I’ll be keeping a slightly closer watch on the MJO this season. Remember, we want the top-right quadrant.

Notes and references

  1. Cassou, Christophe. “Intraseasonal interaction between the Madden–Julian oscillation and the North Atlantic oscillation.” Nature 455.7212 (2008): 523-527.
  2. Donald, Alexis, et al. “Forecasting with the Madden-Julian Oscillation and the applications for risk management.” Proceedings of the 4th International Crop Science Congress (ICSC 2004). 2004.
  3. That is the commonest form of red-green colour blindness, affecting about 1 in 8 males. They probably won’t be able to distinguish phases 1, 2 and 3 from 4, 5 and 6 in the palette I’ve used.