Wednesday, July 12, 2006

Above Average Global Westerly Flow is about to Collapse

The following is a link to our recently accepted paper by MWR which discusses the GSDM (Weickmann and Berry 2006).

From taking into consideration the interactions of 4 different subseasonal time scales, a sequence of maps depicting a coherent set of repeatable events has been derived for the Northern Hemisphere cold season from November-March. This set is broken up into 4 stages, referred to as GSDM (for Global Synoptic-Dynamic Model) Stages 1-4 in the text of my Blog. Figure 13 in our paper presents a schematic of the GSDM. Ideally it would be advantageous to post our weather-climate discussions (link at the bottom) with greater frequency to provide additional detail while having a more complete weather-climate record of attribution and prediction. In these discussions I adapt the GSDM for the warm season. Our list of work includes a seasonally adjusted rendition of the GSDM.

Global tropical SSTs continued a general cooling trend during the past week, especially across portions of the Indian Ocean, extreme west Pacific and Atlantic Oceans. SSTs in these regions lowered by as much as 1C, and were largely a feedback from increased clouds and rainfall. Still, near to above average SSTs remained from the South Pacific into the Tropical Northwest Pacific, as well as along the East Pacific equatorial cold tongue. In fact, anomalies ~ plus 1-2C extended to roughly 200m deep just to the east of the equatorial date line, and that may be the start of another oceanic Kelvin wave per latest TAO data. Actual SSTs in excess of 29C still dominate from the South Pacific into the Indian Ocean, and nearly all of the Tropical Northwest Pacific. Additional global SST information can be obtained from latest TAO data here, ESRL/PSD data here, CPC data,

and BMRC at .

The following are links to ENSO discussions (recently updated).

Tropical convective forcing has strongly re-emerged (and flared-up) into the central and eastern equatorial Indian Ocean toward the Bay of Bengal and Southeast Asia during the past week. Anomalies of outgoing longwave radiation (OLRA) have been ~minus 30-50 W/m**2. The time filtered coherent modes Hovmollers present very weak evidence of convectively coupled Kelvin waves coming back around from the Western Hemisphere initiating this tropical rainfall. However, daily monitoring also suggests jet streak dynamics linked to austral winter Southern Hemisphere baroclinic wave activity may have also contributed, particularly since the convection did start south of the equator. Whatever the case may be, we continue to observe the ~30 day variability of tropical convective forcing across the Indian Ocean. Additionally, time-latitude sections of OLR/A show at least 2 episodes of northward propagation from the equatorial Indian Ocean into southern Asia since early May, reflecting the seasonal cycle. I think we are seeing another such episode, working along with the currently well established and very intense active Indian and Southeast Asia Monsoon systems.

Farther to the east tropical convection across the west central and northwest Pacific remains strong while shifting northwest. At least 3 tropical cyclones have developed in this region (see for details) during the past week. As discussed last week, this region of enhanced tropical thunderstorm activity has been part of a ~4-6 week persistent area of SST boundary forcing extending from the South Pacific then toward Southeast Asia. Latest Hovmoller and other plots suggest suppression is returning to the equatorial west Pacific (as enhancement shifts northwest), while ~minus 5 W/m**2 OLRA persists south of the equator. Latest full disk satellite imagery shows a weaker region of thunderstorm activity from ~140E-180 within 10 deg. of the equator. Hence there are currently two areas of tropical forcing (once again) across the eastern hemisphere, with the western area easily the most intense.

A weak eastward propagating convectively coupled dynamical signal is also present across the Western Hemisphere tropics, part of the fast eastward propagating modes observed since April. Per Hovmoller plots of velocity potential anomalies and additional animations of combined fields, this signal was centered ~90-120W as of July 10th, moving east at about 12 m/s (there has been a loose wave 1 signature of the global velocity potential field at times). The coherent modes do not project this signal onto a convectively coupled Kelvin wave. At least 2 tropical cyclones have recently developed across the East Pacific perhaps as a result of strengthened cross equatorial low level flow (please see TPC link below for details).

TAO buoy data through July 11th show that surface wind anomalies have become very weak within 5 deg. of the equator east of 140E to near the South American coast. The equatorial westerly wind anomalies observed near the date line a week ago have propagated into both hemispheres linking up with the northwest Pacific tropical cyclones and Southern Hemisphere frontal activity. However, Hovmoller plots of daily mean (through July 11th) near equatorial surface wind show that while weak easterly anomalies have returned from just west of the date line to Indonesia, westerly anomalies ~5-7 m/s were present from ~60-90E (consistent with the tropical forcing). Monitoring will be needed to see if this next round of surface westerlies comes out to the date line along the equator.

As stated last week, it is somewhat unusual to see surface westerly flow on the equator this time of year (usually north). However, we observed this kind of behavior during summer 2004 (see weather-climate discussions), prior to the 2004-05 warm event. Per above, the TAO buoy data already suggest weak evidence that a downwelling oceanic Kelvin wave was initiated by the westerly wind event (WWE) last week. Another WWE may intensify that process, possibly leading to a warming of the equatorial eastern Pacific.

Empirical, statistical and numerical prediction tools continue to be inconclusive for useful information about the future evolution of the tropical convection. Please see ESRL/PSD MJO tools , BMRC MJO tools, CPC MJO tools, and for the details (and draw your own conclusions). These tools generally rely on a moderate to strong MJO signal, which is nearly non-existent at this time. West Pacific SST boundary forced, fast dynamical signals and intense eastern hemisphere monsoon systems dominate the tropical convective forcing.

Zonally averaged anomalies of upper tropospheric zonal wind continue to propagate poleward and downward (see July 5th posting). Time-latitude sections of 200mb zonal mean zonal wind anomalies still indicate most of the tropical and subtropical atmospheres to be dominated by westerlies. Largest magnitudes were ~ 10S and 30N, with values of nearly 10 and 5 m/s, respectively. However, especially from 5-10N, values were returning to near normal.

A much more complete understanding of the poleward propagation of circulation anomalies and forcing-response-feedback relationships requires a vertical average of zonal mean anomalies and a global summation (regional considerations and other dynamical processes understood). This is one reason to make use of the atmospheric angular momentum (AAM) plots from ESRL/PSD (links below). While tropospheric global relative AAM is still ~1.5 standard deviations (sigma) above the 1968-1997 reanalysis climatology (1 sigma above the operational 1979-1998 operational data climatology), the tendency is becoming negative (as of July 9th for the reanalysis data). Much of the tendency is coming from the mountain and frictional torques, globally ~ minus 10 and 15 Hadleys, respectively. Of course, the Southern Hemisphere is contributing “a lot” to these signals this time of year.

Weather-climate linkage signals have been incredibly mixed and confusing for the past several weeks. However, the above discussed AAM behaviors are consistent with tropical convective forcing re-emerging into the Indian Ocean along with other subseasonal modes of variability making up the GSDM as discussed in our paper. One of these may have included a “mountain-frictional torque index cycle”. Anomalous westerly flow is reaching the Earth’s surface of the subtropical (trades are weakening across the Tropical North Atlantic, for example) and extratropical atmospheres while negative mean sea level pressure anomalies (MSLPA) are starting to be observed along the east slopes of major north-south mountain barriers globally. Negative MSLP anomalies are also appearing across the Arctic (links to various animations are given below). The point is that the Earth’s surface is removing angular momentum (westerly flow) from the atmosphere, allowing the Earth to gain it (meaning a faster rotation rate and a shortening of the length of day by a few milliseconds!). Hence a collapse of the anomalous atmospheric westerly flow (including the anomalously strong North Pacific jet for this time of year) is expected during the next few weeks.

Animations of daily mean anomalies of 150mb and 250mb vector winds show distorted twin subtropical anticylones ~140-150E (the northwest Pacific tropical cyclones are contributing), with a signal of redevelopment across the Indian Ocean. The latter is tied to the tropical convective flare-up discussed above (easterly anomalies starting to appear at 150mb). There is a fast Rossby wave energy dispersion interacting with the twin subtropical anticyclones leading to yet another anomalous anticyclonic gyre across the north central Pacific. These result is a strong North Pacific Jet with 250mb wind anomalies ~30-40m/s around 50N, and a subtropical jet (STJ) with 150mb anomalies ~plus 15-20 m/s extending into the Desert Southwest (suppressing the monsoon system there). As most global models show, this strong North Pacific Jet will move across the northern part of the USA over the next several days.

Within the GSDM framework, still largely from the zonal mean wind signal from the subtropical atmosphere and the AAM budget (including a possible mountain-frictional torque index cycle), I think there is now a reasonable argument for a boreal summertime rendition of a Stage 3 circulation, about to transition to Stage 4. The strong westerly flow across the North Pacific is supportive of this notion, where strong interactions with the extratropics have contributed.

Uncertainty remains very high for the outlooks that follow, but maybe a bit less than recently given the above. Reasonably good model agreement also minimizes some of that concern for week 1. However, only broad and hopefully statistically useful statements can be offered.

My own thoughts would be to see the renewed Indian Ocean convection move out into the west central and northwest Pacific by late week 2 or 3, behaving similar to the previous 2 events discussed above. This would enhance the SST boundary forced tropical convection in that region, and may contribute to a retrogression of extratropical circulation anomalies across the Pacific-North American (PNA) sector. The role the fast dynamical signal contributes is unclear. There has been a slight tendency for that signal to become less coherent while the convection across the Eastern Hemisphere intensifies. A change to GSDM Stage 4 may occur during week 1, followed possibly by Stage 1 sometime weeks 2-3. One big concern I have is Stage 1 getting bypassed and going to GSDM Stage 2 “quickly”. Interestingly this possibility was mentioned in my last posting dated July 5th. Let’s just see what happens.

It is possible that a transition from the previous cold event (La-Nina) to perhaps an oncoming warm event (El-Nino) may be occurring. Stay tuned for additional information on the latter per links given, especially from CPC.

Week 1 (13-19 July 2006): Strong westerly flow across the northern USA with a broad central states ridge looks like a good bet. Whether or not there are multiple centers to this ridge are unclear synoptic details. I am also unclear just how long the STJ across the Desert Southwest will persist (suppressing their local monsoon system). GSDM Stage 4 would suggest twin subtropical anticyclones across the Western Hemisphere. Perhaps these may set up around 60-90W given the dynamical signal per above, maintaining the STJ across the southwest USA.

The well advertised heat wave (at least loosely) should be on track for most of the country, with the largest positive temperature anomalies centered from the Northern Rockies into the Upper Mississippi Valley. Some locations in these areas will exceed 100F, with high points well into the 70s for locations such as Iowa. There may be some pockets of intense/severe thunderstorms across the northern part of the country linked to jet streaks and frontal activity. Most/all drought stricken parts of the country will not get much relief. With the trades relaxing across the Tropical North Atlantic and the possibility of reduced shear due to twin subtropical highs, there may be a concern for tropical cyclones particularly across the Caribbean later this period going into week 2. Please see for the latest tropical cyclone information. Other types of summertime weather/hazards around the country in this situation should be apparent.

Week 2 (20–26 July 2006): An option that must be monitored is retrogression with amplification across the PNA sector. This may result in a ridge from western North America into Alaska, a central USA/Great Lakes states trough (~90W?) and a ridge across the southeastern part of the country/western Atlantic Ocean (one summertime version of GSDM Stage 2). Much cooler weather for the North Central States with above average rainfall for roughly the eastern half of the country would be expected if this occurred (similar to late June).

Week 3 (27 July – 2 August 2006): Same as week 2, otherwise unclear.

Hot daytime temperatures with maximums from the mid 90s to around 105F and little organized rainfall are likely for Southwest Kansas through much of week 1. I cannot rule out a heavy rainfall surprise somewhere in this part of the world. If there is any truth to the possibility of GSDM Stage 2, much cooler temperatures and an improved opportunity for rainfall would exist sometime during weeks 2-3.

The time -filtered coherent modes Hovmoller plots of OLR and OLRA are at, velocity potential Hovmollers at , and an animation of velocity potential overlayed on OLRAs are at

Satellite imagery and other information can be found from the following links: eastern hemisphere, full-disk west Pacific, mtsat, IO, Africa, ; other imagery here. Latest tropical cyclone statements can be found from, while the latest 3-day averages of OLR totals and anomalies and other data can be found here (animations of various fields from the operational data) (Global Tropical

Hazards Assessment available from this site, along with other useful information) (reanalysis AAM plots) (operational AAM plots)

Latest CDC Ensemble Forecast

Latest NCEP Ensemble Forecast

Additional NCEP Ensemble Output

Latest Canadian Ensemble Output

Latest Deterministic ECMWF Forecasts (link to our Weather-
Climate discussions) (model performance; please navigate to others)

Please see the CPC Drought Monitor for areas of dryness and the latest official outlooks and statements from the Storm Prediction Center not only for severe storms, but also fire weather concerns. Finally, the CPC USA Hazards Assessment for offers additional insights not only for possible week 1 high impact weather, but week 2 as well.

I will try to do another posting this upcoming weekend.
Ed Berry

No comments: