Monday, May 28, 2007

MJO: "I am Back!"

ESRL/PSD reanalysis-1 (R1) data plots (including the AAM plots) have updated through May 23rd as of this writing. Hopefully the Global Wind Oscillation (GWO) phase space plot will be soon.

With the exception of eastern Indonesia, tropical SSTs remain generally quite warm across the Eastern Hemisphere. The most exceptional positive SST anomalies are over the western Indian Ocean having magnitudes of ~2C and totals in excess of 30C. The west central Pacific warmth continues to hang around (along with its positive anomaly horseshoe pattern into the extratropics) with magnitudes of ~1C (down to ~200m) and totals ~30C. Weakening cool anomalies are still present along the equatorial cold tongue east of the date line while the tropical Atlantic remains “luke warm”. Recent surface westerly wind anomalies have led to some moderation of the tropical east Pacific negative anomalies. Whether or not the colder subsurface anomalies (~minus 3-4C ~100m/140W per TAO data) can be upwelled along with any further development of La-Nina depends on the subseasonal events to be discussed.

Now comes some of the fun discussing recent subseasonal atmospheric activities. It is unfortunate we have not been afforded the opportunity to have a dedicated web page effort to illustrate with graphics the appropriate level of science discussions like these requires. The latter is only fair to meteorology and the users. There is much additional diagnostic information that can and must be incorporated into any forecast process for ~weeks 1-4. Kindly stated, recent numerical model performance and official week 1 forecast products issued last week for the current trough digging into the western USA is yet another example where weather-climate linkage information within the GSDM framework would have helped. There was a consensus our current western USA trough would be progressive, which is not going to happen. In order to use numerical models intelligently a detailed and disciplined understanding of the weather-climate situation requiring rigorous daily monitoring is a must, no exceptions!!! The issue of expressing these predictions probabilistically for the benefit of the users is also another matter needing far more attention.

During late April and ~May 11th there was a weak then strong positive global mountain-negative frictional torque index cycle. The latter had a global signal of ~plus 40 Hadleys for the mountain torque, which was roughly 45 days after the big event during early April. It is kind of “funny” how the intraseasonal time scale works into this. What all this means is dynamical processes involving the (tilts of the) midlatitude synoptic eddies interacting with global north-south mountain massifs are adding anomalous zonal mean westerly flow to particularly the subtropical atmospheres of both hemispheres. The May 11th variation contributed to a global AAM time tendency of at least plus 40 Hadleys.

So, where am I going here? I have discussed in past postings ~10-30 day variability involving the tropical convective forcing and mountain-frictional torques. These atmospheric variations have been occurring since at least mid February, within a base state generally fluxing (transporting) AAM out of the subtropics into the midlatitudes (GSDM Stage 1). The transports are the largest contributor to the earth-atmosphere AAM budget. Having both the Indian Ocean and west Pacific Ocean tropical convective forcing occurring “at the same time” has loosely supported a GSDM Stage 4-1 base state, but with at times significant orbits in phase space as seen on the GWO plot. Monitoring these orbits has added some additional week-2 predictability (over the models) to the barrage of western USA troughs seen this spring. The dynamics for this are complicated, but this is where thinking in terms of forcing-response-feedbacks with subsequent interactions is needed. I think we have just gone through another large orbit in GWO phase space, with the above contributing to it.

The additional anomalous subtropical westerly flow (~5-10m/s at 200mb) and interactions with the synoptic eddies forced the dynamical tropical convective signal into the Western Hemisphere about a week-10 days ago. Possibly involving the seasonal cycle including the Eastern Hemisphere monsoon systems, I think this dynamical signal has developed into a true Madden-Julian Oscillation (MJO). The Wheeler phase space plot strongly supports this assertion, as do phase speed computations of recent OLRA anomalies (~3.5m/s). Full disk satellite imagery has shown a rapid increase of intense tropical convection across the warm waters of the Indian Ocean, centered ~5N/60E. There is sporadic convection across the west central Pacific while thunderstorm clusters linger across Africa and roughly northern South America.

I think the weather-climate situation is solidly in GSDM Stage 4. Anomalous zonal mean easterly flow of ~5m/s at 200mb has developed throughout the tropical atmosphere. Twin upper tropospheric subtropical anticyclones remain across the Western Hemisphere (linked to the subtropical jet across the southern USA), but are slowly progressing into Africa and the Indian Ocean. As of May 23rd the AAM time tendency was ~minus 30 Hadleys which included a contribution from a negative East Asian mountain torque. The anomalous Rossby wave train tied to that torque was one of my reasons for thinking late last week that a slower and deeper western USA trough was more probable this week.

Besides the upper tropospheric zonal mean zonal wind anomalies across the tropical and subtropical atmospheres, there are also zonal mean anomalous easterlies ~35N and 30S, with anomalous westerlies poleward. In fact, at 60N these westerly anomalies have been recently ~plus 15m/s at 200mb, and there is evidence that some of this westerly flow goes back to poleward propagation off the equator from the strong December 2006 MJO event. In any case, this gives some reason to the recent observed midlatitude split flows.

This writing is already far too long. I think the global weather-climate situation will be in GSDM Stage 1 by week 2 as the MJO moves slowly east. I am still unclear about the timing and intensity of this new MJO; however, I speculate its centroid will be ~100-120E by the end of week 2. A trade wind surge downstream from this MJO may "revive" La-Nina development. Both the Indian and Southeast Asian monsoon regions should be active, while the west Pacific remains a “wild card”, as it has for several years. Models are finally starting to pick up not only on the slower and deeper western USA trough (developing into a closed low) for week 1, but even the much stronger system starting ~ day 10. The week 1 trough will be more of a heavy rainfall producer for the Plains and Mississippi Valley while the week 2 system may result in widespread severe storms across the central part of the country including tornadoes. Other ramifications should be apparent. During weeks 3-4 the situation evolve into GSDM Stage 2.


An experimental phase space plot of the GSDM (depicting a Global Wind Oscillation (GWO)) utilizing normalized relative AAM time tendency (Y-axis) and normalized relative AAM (X-axis) can be found at

These are probabilistic statements, and work is ongoing to quantify in future posts. The WB (2007) paper on the GSDM has been published in the February issue of MWR. I will not be able to write another posting until next Monday (6/4) due to covering various shifts.

Ed Berry

No comments: