High latitude of the monsoon trough,

Vortices embedded in monsoon trough,

Large diurnal modulation of convection over Indonesia,

Apparent large influence by the Madden Julian Oscillation (MJO) this season

Hello all,

 In an attempt to restart discussion, I thought I would just give a short description of the current situation.  A copy of this discussion and the  figures I refer to are on my webpage.

Being early January, the Southern Hemisphere summer monsoon is in full swing. On my page I have placed a sequence of satellite images across the Indian Ocean and west Pacific for last night:

The major convective/OLR activity is sitting across Indonesia and Northern Australia.  The Indonesian convection very much follows the land; i.e it lies over Southerrn Sumatra, Java and islands eastward along the Archipelago.  Even though there is a major convective/westerly outbreak at the moment, the diurnal variation is supermiposed over the top of that.  Thus, if you compare the 1200 image with the 0000 image for this morning (12 hours later), the large scale convective outbreak remains, but within it the convection has moved from land (at 1200 UTC) to sea at 00 UTC.

Looking at the low-level winds as depicted by the GASP 850 hPa analysis for the same time,

,

cyclonic relative vorticity (in the Southern Hemisphere) is shaded blue.  Equating the monsoon trough to the shear vorticity between the equatorial westerlies and the higher latitude trade-easterlies, the trough strectches from Madagascar across the Indian Ocean to the waters off northwest Australia and across Austarlia , where it disappears (because the monsoon westerlies finish and give way to Pacific equatorial easterlies, from about 150E).

Thus the monsoon trough currently lies a long way poleward: i.e between 10 and 20 South.  The convection over Indonesia is lying within a stream of fairly strong westerlies, equatorward of the trough.  The monsoon trough itself has a number of embedded discerete vortices: the two tropical cyclones at the western end, and two less-developed vortices at about 110E and over Northern Australia at about 135E.

The Madden Julian Oscillation (MJO) is having a strong effect on Indonesian-Australian monsoon during this current season.  This can be seen on the diagnostics prepared by Matt Wheeler. I have included on my page a link to his current phase-space index of the Oscillation here  The axes of this diagram represent the amplitudes of each of the two leading multivariate EOF's of 200-Hpa zonal wind, 850-hpa zonal wind and OLR averaged between 15Sa nd 15 N.  Each point on the phase-space diagram represents the value of the two EOF's on that day.  The fact the points circle around is because the two EOF's have a very similar structure to each other but are in quadrature (i.e a quarter wavelength out of phase) and so represent a pattern moving eastwards (which Matt and Harry Hendon in a recently submitted paper have identified as being the MJO --- for details see Matt's website: http://www.bom.gov.au/bmrc/clfor/cfstaff/matw/maproom/RMM/index.htm

The composited wind fields for each phase at this time of year are shown here (wind vectors representing 850 hPa anomalies and blue shading representing cold OLR anomalies and yellow warm OLR anomalies).  Referring to this diagram: during phases 4, 5, 6 there is enhanced convection and  westerly winds over northern Australia and Indonesia, which referring back to the current phase diagram was the situation during December.  Conversely in phases 8, 1, 2 there are easterly anomalies and suppressed convection over Indonesia/northern Australia, which was the situation during most of January.

Over the past couple of days we have moved into phase 4 on the diagram, which is consistent with the blow-up of convection and enhanced westerlies over the western end of the Indonesian Archipelago as discussed in the satellite images and 850 hPa chart.

This strong interaction between the MJO and Australina-Indonesian activity occurs in some seasons but not in others as discussed in a recent review article by Matt Wheeler and myself, which is still in draft form.   In that paper we took a 3-day running mean of the daily OLR averaged over the region 15^oS to 5^oS and 120^oE to 140^oE, representing a time series of convection activity over the Indonesian-Australian summer monsoon. We carried out  a multiple regression of this OLR series against the amplitudes of the two EOF's making up Matt's MJO phase space.  In certain years when the MJO activity was strong the percentage of the variance of OLR accounted for by the MJO EOF time series was as high as 50%, whereas in other years the MJO influence was quite weak and accounted for less than 10%

As discussed above, however, so far the current monsoon season has been one where the MJO has a large influence.

Comments and criticisms are encouraged.  Hopefully someone will disagree with some or all of this, so we can have a discussion.

John mcB

Matt Wheeler

Monsoon list,

In relation to John's discussion on the MJO, it is worth noting that while the development of the monsoon over Indonesia and far northern Australia has seen a marked variation with the MJO this year, the rainfall over more southern latitudes in Australia (especially central Queensland) has not. There was a period of heavy rainfalls in central Queensland around the second week of January, as shown here in the average from the 10th to 16th of Jan:

By the MJO RMM Phase-space diagnostic, this occurred during Phase 8 of the MJO. By the composite you were referred to in John's e-mail, this is a phase during which the MJO influence should be of relatively suppressed conditions about more northern parts of Australia and Indonesia.

Thus, this is a good example of a significant rainfall event, even in a "good" MJO year, being what seems to be not MJO-related. Indeed, given the relatively weak signals in the composites, as seen at http://www.bom.gov.au/bmrc/clfor/cfstaff/matw/maproom/RMM/composites/index.htm
over central Queensland, even in summer, this is a location where the MJO influence is either relatively small, or at least different from event to event.

Matt Wheeler

Michio Yanai
Dear John,

Thank you for showing this very interesting set of satellite images and 850 hPa wind.

I have no experience of synoptic analysis in this region, but the picture shown here (such as the explosive development of convection with the  arrival of MJO, and vortices in the monsoon trough suggestive of  dynamic instability) fits quite well with the image of onset of the  Australian summer monsoon we obtained from a climatological study
(Hung and Yanai, 2004, QJ, in press).

In our study, we observed the gradual buildup of thermal contrast  between the heated Australia and Arafura Sea leading to the reversal  of of the sign of dT/dy  (positive to negative) and formation of a  heat low in a shallow layer over the continental Australia.  Can you say something about such pre-conditioning in the thermal field for  this monsoon season?

Michio

Sam Cleland

The recent progression of MJO should go in with the "classics", and could make a good case study.  As Matt pointed out though, one of the more ironic aspects was the drought-breaking (for some) monsoon low that passed through
western Queensland occurred during the inactive phase (mid-January). This  low started up near Darwin and also produced lots of rain over the Top End and Gulf Country before heading to Queensland, but only appears as a small  blip on any OLR time-longitude plot. Hence analysis using station rainfalls over northern Australia will be conflicting with the RMMs.

Does anyone have any thoughts regarding the bad MJO years - that is those  years when the MJO seems to have less effect. Is it typically a case of:
         a) active phases being less active;
         b) inactive phases being less inactive  - perhaps other influences
initiated broad scale convection?
         c) no MJO? (I guess a+b~=c)

It would seem that when "other influences" result in weather during  supposed inactive MJO phases, any technique for diagnosing an MJO will get confused? (As an analysis against rainfall figures will be in this  "classic" (so far)  year)

Cheers

Sam