Figure 1
Since our last report on 28 January 2004, a new moist phase of the Madden Julian Oscillation (MJO) has continued to intensify and move eastward from the Indian Ocean to the maritime continent and north Australian region. Section 1 discusses a circulation change related to the MJO that occurred during the period of 15-26 January 2004 and brought a cold and stormy regime to much of the USA. Section 2 gives an overview of the current climate situation, while Section 3 presents some predictive insights for weeks 1-3. We do not consider the possible effects of the MJO on forcing an oceanic Kelvin wave or on El Nino. Click here to see the latest ENSO diagnostic discussion from the Climate Prediction Center and for more information on the MJO and El Nino.
Figure 1 shows a time-longitude (Hovmoller) plot of outgoing longwave radiation anomalies (OLRA, a proxy for deep tropical convection) for two different latitude bands, one from 7.5S-7.5N (top) and the other from 2.5-15.0S (bottom). Superimposed are contours that illustrate time- space filtered convectively coupled modes of tropical convection, including the MJO, the Kelvin wave and the equatorial Rossby wave. The period of interest is from 15-26 January and is highlighted on the figures.
On 15 January the convectively active envelope of the December/January 2004 MJO #1 is over the western hemisphere and is concentrated in three regions: east of the dateline, near 40W over South America and near 20E over Africa. By the end of the period, active convection has consolidated near 80E and has weakened in the other three regions. The consolidation will be referred to as the "new" MJO #2. At the same time, the envelope of suppressed convection that followed MJO #1 moved from about 120 to 160E. Within the envelope of active equatorial convection, there are at least two convectively coupled Kelvin waves that propagate into the Indian Ocean from the west. South of the equator over the Indian Ocean, multiple tropical cyclones are present. There is a clear link between convection from MJO #1 and the start of MJO #2. For the circulation transition, the main point is that tropical forcing changes substantially during the 11-day period.
Figure 2 shows a Hovmoller of 250mb geopotential height anomalies for the latitude band of 30-60N. Some relevant features are highlighted, including the period 15-26 January 2004. On 15 January, a strong anomalous anticyclone was centered across the western part of the USA (110W) with +150 gpm anomalies. This feature was part of a L-H-L pattern across the PNA sector, which had been in place for about 8 days. Its development is discussed in our previous report. Fig. 1 showed that tropical convection was weakening rapidly across the central Pacific Ocean basin (~150W) as part of a major change in tropical forcing. In association with this change, two events are highlighted in Fig. 2: 1) two episodes of downstream wave energy dispersion that emanate from the region of the western Atlantic (dotted line segments labeled 1 and 5) and 2) the retrogression of the anticyclone toward the east Pacific (solid arrow). By 25 January, the anticyclone has shifted back to about 150W and a downstream trough has formed along 120W. At the same time, the second energy dispersion had worked its way around again to eastern Asia. This episode is best seen in synoptic maps since it occurs somewhat outside of the 30-60N latitude band. Note that the anomalous cyclone at around 60W continued to persist, moving only slightly eastward.
Figure 3 illustrates the pattern change in 250 mb wind and OLRA that occurred from 15 to 26 January 2004. The winds are a daily mean while the OLRA are a 7-day mean. On 15 January, there is enhanced convection around 10S/150W, tropical South America and equatorial Africa, and the extratropical flow shows an extended jet across the North Pacific and split flow over North America. By 26 January, the tropical forcing has consolidated over the Indian Ocean and the jet stream has contracted to west of the dateline. Split flow occurs across the eastern Pacific (mostly ridging) and a downstream trough is over North America. Despite the pattern change, some circulation features remain similar or persist. In particular anticyclones remain prominent poleward of 60N and the westerly flow remains strong at around 30N. More will be said later about contribution of other physical processes to these circulation anomalies.
First, however, we present some of the synoptic details of the 15-26 January 2004 transition. Figure 4 shows a sequence of daily mean 250mb vector wind anomalies spaced two days apart. Selected cyclones and anticyclones associated with five wavetrains are highlighted. The reader is referred to Fig. 2 for the "big picture", realizing of course that some of the wavetrains' centers are best developed outside the 30-60N latitude band. The sequence starts on 15 January with a pattern of L-H-L across the PNA region. Recall this wavetrain was initially forced by tropical convection around the dateline on ~3 January 2004. By 17 January, an apparent dispersion or breakdown of this anomaly pattern begins. A wavetrain (#1) emanates from an intensifying low over the western Atlantic and extends downstream over North Asia to a cyclonic anomaly near Japan by 19 January. During the same 15-19 Jan time span, the low over the eastern North Pacific moves inland and the high over North America retrogrades to the east Pacific. A cyclone and anomalous subtropical westerlies continue over the west Pacific as part of a continuous belt of anomalous westerlies around the hemisphere.
By 21 January energy propagates into the Indian Ocean from continuing developments over the North Atlantic (wavetrain #2) and contributes to a strong cyclonic gyre just north of the Arabian Sea. The cyclone helps intensify Indian Ocean convection and also interacts with the downstream low near Japan (previously the terminus of wavetrain #1). This low amplifies and shifts south. Energy from this development is then steered southeast toward the subtropical west Pacific (wavetrain #3) and a subtropical high-low couplet amplifies on either side of the dateline on 23 Jan 2004. This is followed by further amplification of the east Pacific anticyclone as energy propagates northeast across North America (wavetrain #4) on 25 January.
The transition in circulation anomalies is now complete. An anticyclone is in place over the Indian Ocean with strong cross-equatorial flow to its west from enhanced convection. A "downstream trough" is in place south of Hawaii with southward cross-equatorial flow toward strongly suppressed convection near the dateline. This circulation change placed most of the USA in an anomalously cold and stormy regime. However, note the persistence of the anomalous 250mb cyclones just to the east of North America, even after the transition (see also Fig. 2). In fact, energy coming out of the east Pacific on 25 January (wavetrain #4) contributes to another amplification of the west Atlantic low and another flow of energy across northern Asia (wavetrain #5). The cyclone within this wavetrain represents the beginnings of the extremely strong jet stream that is currently (9 February) over the Pacific Ocean. Its track of movement is highlighted on Fig. 2 (dashed arrow). More will be said about it in Section 2. First we discuss some other physical processes contributing to the observed circulation anomalies.
Fig. 5 shows a height-time plot of geopotential height anomalies over the polar cap (65-90N) and a time series of the Arctic Oscillation (AO). A sudden stratospheric warming (SSW) is evident in the top plot. It starts in the high stratosphere around mid-December 2003 and "propagates" downward reaching the surface in early January 2004. As a result of the high sea level pressures over the Arctic (recall the persistent anticyclones seen in Fig. 3), bitterly cold air has been free to move into mid-latitudes, including the lower 48 states and especially from the Plains to the Northeast. The SSW qualifies as an extreme event with positive height anomalies > 3 sigma in the stratosphere, and it shows little sign of weakening. Concurrently the AO has been in a negative phase. In terms of zonal momentum, this implies zonal mean easterlies at about 60N and zonal mean westerlies at about 30N.
Fig. 6 shows the anomalies in zonal mean and global atmospheric angular momentum for the last 90 days. In equatorial regions, a pattern of easterly anomalous flow in early December 2003 gives way to westerly flow in late December. These wind anomalies are forced by MJO #1. In early January 2004, as pressures rise over polar regions, strong westerly wind anomalies develop at ~15N and global AAM rises rapidly. These anomalies are south of the westerly anomalies associated with a negative AO, although easterly anomalies are present along 55-60N since about 6 January 2004.
Fig. 7 shows the zonal and global integral of the mountain torque, the primary forcing of global AAM. Several episodes of positive mountain torque are evident in the time series, including a strong and persistent event in early January 2004. The colored curves show that both North America and Asian mountains contributed to the global torque. Another positive mountain torque is evident from Asia around the time of our pattern transition, further contributing to strong westerlies over the subtropical Pacific. These mountain torques provide the momentum source for the anomalously strong subtropical westerly flow, which is currently moving poleward (see updated Fig. 6).
Of course, the mountain torque is not independent of high latitude sea level pressure anomalies that accompany a negative AO. A positive torque is produced by positive (negative) pressure anomalies on the eastern (western) slope of mountains, which can occur as cold air masses move southward from Arctic regions. Interactions among the AO, mountain torques, tropical convective forcing and Rossby wave/synoptic-scale energy dispersion are contributing to daily variations of the circulation.
Figure 8 is a phase space diagram of the MJO, which is currently in phase 4 near the maritime continent (see also Fig. 1). Figures 9 and 10 show convection relatively suppressed across the Indian Ocean, but quite active across the maritime continent. The mass of clouds northeast of New Guinea is a Kelvin wave and is associated with 20kt westerly surface winds. The MJO has been moving eastward rather rapidly, on the order of 4-7 m/s (~5 deg/day). As shown in Fig. 1, Kelvin wave activity has accompanied MJO #2 more so than MJO #1. This may account for the fast movement of MJO #2.
Figure 11 shows 7-day mean 150mb vector wind anomalies for the period of 30 Jan-05 Feb 2004. The pattern of zonal mean easterly flow anomalies near 60N (in the stratosphere) and westerly flow near 30N is clear. The circulation anomalies that developed during the transition discussed in Section 1 are also still in place. Broad twin subtropical anticyclones are present across the southern Asia and Indian Ocean sector and downstream twin cyclones are present over the subtropical Pacific. A split flow pattern over the east Pacific Ocean has an anticyclonic gyre near 40N/150W and an anomalous trough across the western USA. Finally, note the well-defined meridional structure of wind anomalies in place over Asia. The anticyclones linked to Indian Ocean convection are aligned with broad cyclones over Asia. A portion of this pattern has been moving slowly eastward as shown previously in Fig. 2 (dashed arrow at bottom). Anomalously cold air over Asia, which developed in association with wavetrain #5 in Fig. 4, combined with anomalous convection over the Indian Ocean to produce the very strong westerly wind anomalies seen along 35N over south Asia.
Figure 12 presents the daily mean of 150mb vector winds anomalies valid 5 Feb 2004. Consistent with eastward movement of the MJO, the daily mean circulation anomalies are displaced eastward relative to Fig. 11). Perhaps most relevant is the intense jet stream moving out of Asia, which we have linked with tropical-extratropical interactions there. The Pacific satellite picture (Fig. 10) suggests it may also be enhanced by the convective Kelvin wave moving eastward ahead of the MJO.
Given the current rate of movement of the MJO (assuming 6 m/s), it should reach the dateline during late week 2 (14-20 February). SSTA are still on the order of +1.0C just west of the dateline and maximum SSTs are 31C (~10S/160E). Convection anomalies are expected to intensify as the MJO moves over the warm water. However, there is still uncertainty about when the MJO will reach the region of positive SSTA. The most recent OLRA and satellite imagery indicate the MJO may be stalling near 140-150E (see also) . On the other hand a statistical prediction of the MJO supports the week 2 scenario.
Currently, a strong jet stream dominates the Pacific Ocean. This is likely a transient feature that will continue moving east. Such transient, mobile jet streams are not unusual in association with the MJO. They tend to "outrun" the MJO convection since they are embedded in strong westerly mid-latitude flow. The jet will help determine the weather conditions over the USA for at least the next 7-10 days. Strong difluent flow is already in place and its northern ridge and southern trough will have significant weather impacts on the USA. The northern ridge might be expected to retrograde after the jet moves into North America. Another more persistent jet extension is expected when the full convective anomaly of the MJO is established near the dateline.
Figure 13 shows the ensemble mean output from the NCEP operational and CDC experimental forecast systems. The maps are valid at 264 hours from 0000 UTC 5 Feb 2004 initial conditions. The CDC ensemble suggests a more amplified ridge over western Canada with stronger westerly flow approaching the west coast of the USA. Both show a trough over the northeast USA. Tropical precipitation forecasts may account for the difference in the circulation forecasts.
Fig. 14 shows the 200 mb velocity potential anomaly (i.e., large scale divergence anomaly) from the NCEP and CDC ensembles and from Wheeler's statistical scheme for week 2. The CDC and statistical model suggest southerly divergent outflow into the northern hemisphere west of the dateline, while the NCEP model's divergent flow tends to be northwesterly. Given the current situation with the MJO, the transient Pacific jet stream and the positive SSTA near the dateline, the CDC ensemble appears more reasonable and lends support to its ensemble mean forecast shown for week 2 in Fig. 13 (above).
Week 1 (11-17 February 2004) is expected to see a moderate surge of Arctic air initially into the northern Rockies and Plains, settling into the south central and southeast states late in the period. A more significant outbreak of bitterly cold Arctic air will likely affect the northeast states by the weekend, with that airmass persisting through at least the end of this time frame. Most of the precipitation should occur from the southern plains into the southeast USA before the weekend. Emphasis is then expected to shift back to the west coast by the weekend as the extended Pacific jet drives cyclonic systems inland, with a lower amplitude ridge and split flow moving into the Rockies. There should be some concern for light to moderate frozen precipitation across portions of the deep south early week 1, with at least seasonably strong storm systems for much of the west coast (perhaps mainly the northwest) by the weekend.
Week 2 (18-24 February 2004) should see the continuation of storms moving onto the west coast at least early in the period, with these cyclonic systems initially moving into the southern plains. There are important timing issues with the details of the evolving circulation in relation to the PNA sector. However, as the MJO moves into the tropical west Pacific, it appears plausible the ridge should retrograde into the east Pacific and amplify. Uncertainty exists if there will be a cold air source. Nevertheless, by late in this period much of the USA may again be anomalously cold and wet.
Week 3 (25 February - 2 March 2004) will depend on the development and persistence of convection near the dateline. A scenario by this time frame would have yet another extended anomalously strong Pacific jet impact the west coast, but farther south than week 1. That may drive cyclonic systems into the southwest states, leading to increased precipitation for an area that has experienced drought. These storms could then impact much of at least the southern half of the USA. There are serious timing issues to consider for forecasts at this lead time. For example, the speed of movement of the MJO is still uncertain. Continuous daily monitoring may add valuable lead time and help anticipate these developments.
