There is little change in the spatial distribution of global tropical SSTs from a week ago. SST anomalies along the equatorial cold tongue from ~165E to the west coast of South America vary from ~minus 1-3C with the coldest around 120W, extending to depths of at least 200m. SST totals of 29C and warmer are well entrenched across the Tropical Northwest Pacific (TNWP) into portions of the South Pacific Ocean loosely centered ~10N/140E. Recall that we use SST totals of 29C as a threshold for supporting persistent deep tropical convective forcing. Recent enhanced convection has slightly reduced the magnitudes of the warm anomalies from a week ago while this anomalously expanded warm pool slowly shifts south with the seasonal cycle. The equatorial Indian Ocean SSTs remain roughly .5-1.5C above normal as do portions of the Caribbean into the deep tropical Atlantic Ocean.
Per time-longitude section plots derived from the TAO buoy data, the evolution of this La-Nina appears not only ahead of the 2005-06 event, but slightly stronger. For instance, SST anomalies of minus 2C and lower have more persistently covered a larger region of the equatorial Pacific Ocean cold tongue, especially east of 140W. Based on the SSTs, the 2005-06 event peaked around February 2006, while this may be in the process of maturing as I type. However, there are complex coupled subseasonal dynamical forcing-response-feedbacks, etc., working with the seasonal cycle. Hence it is still unclear to me when this cold event will mature.
From this point I am not going to repeat what I wrote nearly a week ago since it is still valid, particularly in the interest of striving for some brevity (which failed!). Subseasonal variation #4 of the tropical convective forcing has progressed east-northeast into the west central and TNWP. Full disk satellite imagery and other monitoring tools suggest this forcing is centered ~10-15N/140E, with negative OLR extending from this region into portions of the very warm South Pacific Ocean. An almost separate region of enhanced rainfall is ~20N/60-90E. Finally, linked with a Southern Hemisphere Rossby wave energy dispersion (RWD), a separate intense tropical convective flare-up is developing ~60-70E just south of the equator. The latter may be part of a dynamical process to initiate what could become subseasonal event #5.
However, the most intense tropical forcing currently is across the west central Pacific Ocean, and is impacting the global circulation such as adding weak zonal mean westerly wind flow anomalies to the northern subtropical atmosphere (more said below). Surface westerly wind anomalies ~5m/s extend from the equator into the northwest Pacific Ocean, and it is probable 1 or more tropical cyclones will get spun off from the intense tropical thunderstorm clusters. At this point I do not think an oceanic Kelvin wave will get generated by the weak equatorial surface westerly anomalies.
Satellite animations and OLR/A Hovmoller plots do suggest the west Pacific convection to be drifting generally west, and may consolidate with the forcing loosely centered on the Bay of Bengal. Adding a bit more detail, during ~weeks 1-2 I would expect a portion of the west central Pacific convection to shift back to the northwest possibly in the form of tropical cyclones while the South Pacific Convergence Zone (SPCZ) becomes enhanced. Tropical convection may once again become quite intense from 60-90E while consolidating with the forcing farther east and north. I would speculate that during the next few weeks, tropical convective forcing should be nicely consolidated centered in the region of 100-120E along and just north of the equator. The latter is a consistent tropical convective signal with both La-Nina and the seasonal cycle. Again, make no mistake about the valid concerns I have of our nemesis. Whatever the case may be, the problem of 2 regions of Eastern Hemisphere tropical convective forcing will be a global weather-climate issue through at least DJF 2007-08.
Global relative AAM remains well below the R1 data climatology (updated through 24 September) by roughly 2 atmospheric momentum units (1 AMU = 1*10**24 Kg m/s**2) which is also approximately 2 standard deviations. A week ago I thought given the fairly strong poleward transport AAM signal across the Northern Hemisphere midlatitudes we would see a global response from the surface torques. However, that has not occurred. Updated through 21 September the Global Wind Oscillation (GWO) was still stuck at GSDM Stage 1 at roughly 2 standard deviations, and I do not think there has been much change since. From a zonal mean perspective there has been a recent weakly positive East Asian mountain torque, with a positive frictional torque of roughly 10 Hadleys (1 Hadley = 1*10**18 Kg m**2/s**2 = 1*10**18 J) at 30N compensated by dissipation at 50N. The latter are linked to an increase of storm track activity across the Northern Hemisphere midlatitudes.
The most significant point in regard to global relative AAM and zonal mean variations is that zonal mean anomalous easterly wind flow throughout the troposphere has been propagating off the equator into the subtropical atmospheres of both hemispheres since around 10 August (~10m/s at 200mb). This propagation has been to about 35N and 20S, and was tied to subseasonal tropical events #3 and # 4. In fact, the tropical convective flare-up across the TWNP discussed above has actually added weak zonal westerly wind flow anomalies (coming from the Western Hemisphere Pacific Ocean) from the equator to ~20N. As a La-Nina matures, it is typical to observe this kind of atmospheric behavior.
On the poleward flanks of these anomalous zonal mean easterly winds, westerly wind flow anomalies have also been increasing, especially across the higher latitudes of the Northern Hemisphere. The latter is not only due to seasonal transition into boreal autumn, but also a direct response to poleward shifted anomalous ridges dynamically forced by the meridional propapagtion of the zonal mean easterly wind flow anomalies discussed above. In fact, the zonal mean earth AAM is ~plus .2-.4 AMUs around 40-50N meaning an increase of atmospheric mass suggestive of anomalously high surface pressures (see the references in WB2007). I believe this is why extratropical storm track activity across the Northern Hemisphere midlatitudes has been increasing. This whole diagnosis is yet another example how our evolving La-Nina is already impacting the global synoptic variability (and has been since boreal winter). I disagree with any statements to the contrary from anywhere, particularly from the viewpoint of subseasonal atmospheric variability.
The global circulation is coupled to the developing cold event. RWDs from the west central and TNWP tropical convective forcing have also been strengthening westerly flow across the North Pacific Ocean. In fact, a fairly well defined residual of RWD activity linking both the Indian Ocean and west central Pacific forcing can be observed from weekly mean anomalies of 150mb vector winds. Wind speed anomalies greater than 15m/s have been common with these RWDs. Animations of upper tropospheric daily mean vector wind anomalies show twin tropical/subtropical anticyclones across both the Indian and west central Pacific Ocean basins, with the usual downstream cyclones generally around the date line, typical of GSDM Stage 1. So far the equatorial westerly wind anomalies tied to these twin lows have been mostly propagating into the Southern Hemisphere via South America (through RWDs), increasing their extratropical westerly flow.
Remembering the AAM considerations discussed above, these twin anticyclones are anchoring RWDs which are in the process of generating an extremely strong poleward shifted (~50-60N) north Pacific Ocean jet. This past JJA tropical forcing remained generally in the region of India, and in the presence of the boreal summer base state, troughs stayed offshore of North America favoring “Devil Ridges”. Now, anomalously intense troughs are probable to pound the western half of the USA for at least the next couple of weeks. This situation is stronger than I would have expected a week ago. The first of these will be an anticyclonically wave breaking event typical of GSDM Stage 1 this weekend. However, that character may locally change across North America given the west central Pacific Ocean signal. I do think there will be a mountain-frictional torque index cycle like variation during the next few weeks meaning a brief GWO orbit toward GSDM Stage 2.
The global numerical ensemble prediction systems (as part of a complete forecast process including the GSDM) continue to struggle with the synoptic details. However, there is good agreement for a western USA trough and eastern states ridge. There are some ensemble members suggesting a Stage 2 like pattern for North America starting late week 2. That response suggests a ridge west of North America well into Alaska and a deep central USA trough.
The bottom line is much of the USA is in for an unusually active early autumn regime for at least the next couple of weeks. Most of Alaska should have colder than normal temperatures with periods of rain and snow on the north side of the strong Pacific jet. In addition to significant mountain snowfall particularly west of the Continental Divide (for the CONUS), early season severe downslope wind storms are probable at times across the central and Northern Rockies. Periods of heavy rainfall along with episodes of severe local storms including tornadoes are probable for much of the Plains, starting this weekend. Much of the east will continue with summertime warmth. During week-2 colder polar air (still need the Arctic Ocean to “freeze” to have a good cold air source) should dump into much of the Rockies and even portions of the Northern Plains. It would not surprise me to see an early season significant snowstorm for locations such as eastern Wyoming-Montana into the western Dakotas.
I would expect a respite from this active pattern during the latter part of October into November. However, I will restate that as boreal winter approaches, western USA troughs should become stronger and more persistent, leading to an active southwest flow storm track across the Plains. This type of pattern can be favorable for Arctic outbreaks combined with intense precipitation. Should our nemesis become the rule, perhaps GSDM Stage 4-1 may start out the boreal cold season leading to GSDM Stage 3-4 if an El-Nino starts to develop during winter-spring 2008.
Careful daily monitoring of tropical forcing across the very warm west central Pacific Ocean is a must starting now. For instance, my assessment for the Eastern Hemisphere tropical forcing to again consolidate may get "blown out of the water (pun?)" if the current west central Pacific Ocean convection takes over now given the very warm SSTs. Ramifications could then be completely different much sooner, which is simply a testament to growing uncertainty.
Please see the latest statements from the NOAA/NWS/Tropical Prediction Center for tropical cyclone statements. Tropical cyclone activity did become more robust that I thought a week ago, with much of that coming from “Gabrielles”. Right now Tropical Storm Karen is fighting wind shear and stability per concerns discussed in past postings. Nevertheless, attention still needs to be paid to the tropical cyclone hazard not only from the Atlantic Ocean but also the Caribbean. The USA is by no means “out of the woods”.
The west central into the northwest Pacific Ocean is probable to continue with a tropical cyclone hazard through week 2. Impact areas include the Philippines possibly north into Japan. Another area of concern is the Bay of Bengal including coastal sections. In general, locations from India into Southeast Asia are still probable to continue with rounds of intense/severe thunderstorms through at least week 2. Much of the North Atlantic into Europe is probable to continue with active weather also for at least the next couple of weeks. On the latter, there are some ensemble members suggesting a period for troughs to remain west Europe especially week-2, meaning warmer and dryer weather.
An experimental quasi-phase space plot of the GSDM utilizing a time series of normalized relative AAM tendency anomaly (Y-axis) and normalized relative AAM anomaly time series (X-axis) can be found at
We call the behavior of this plot the Global Wind Oscillation (GWO). While the intent of the GSDM is to extend current thinking beyond the MJO, the purpose of the GWO is to illustrate the non-oscillatory stochastically forced component of the GSDM.
These are probabilistic statements, and work is ongoing to quantify in future posts (for example, risk assessment maps, signal to noise ratio plots and shifts of probability). We hope that an opportunity will arise for us to have a dedicated web page effort to expedite more objectively, with rigor, thoroughness and verification. The WB (2007) paper on the GSDM has been published in the February issue of MWR. I will be on travel and my next discussion will not be until ~ 4-5 October at the earliest. Difficulty will increase keeping these discussions current.