Figure 1
Since our last report on 10 February 2004, the convectively active phase of a Madden-Julian Oscillation (MJO) progressed eastward from the maritime continent/north Australian region to the west Pacific Ocean and then weakened considerably. As a result, tropical convective forcing of the circulation was less coherent during the last month. On the other hand, the circulation anomalies associated with the MJO strengthened during the month as strong westerly flow developed around 30N. The month has had a persistent negative phase of the North Atlantic and Arctic Oscillations, also consistent with strong westerly flow at 30N. In the last week, westerly anomalies have shifted north and a new round of easterly anomalies is developing in the subtropics. This would mean a northward-shifted storm track especially in the Northern Hemisphere.
The early February 2004 MJO did not produce a strong signal in the equatorial Pacific Ocean's thermocline or ocean currents. This time of year the maximum zonal surface wind anomalies tend to be too far south (10-15S) to give much of a signal on the equator. In fact, maximum positive SSTA in the west Pacific Ocean shifted west during the period. Click here to see the latest ENSO diagnostic discussion from the Climate Prediction Center.
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 depict space-time filtered convectively coupled modes of tropical convection, including the MJO, the Kelvin wave (Kw) and the equatorial Rossby wave (er1). For reference purposes, the dashed arrow shows the track of an intense jet streak that moved off of Asia and propagated coherently to North America.
The two recent MJOs (numbered 1 and 2) and the more complicated evolution of anomalies after approximately 15 February are evident in Fig. 1. At the time of our last discussion (10 February) the convectively active phase of MJO # 2 was located over the maritime continent and north Australian region. Based on the envelope of negative OLR anomalies ~<- 30 W/m2 near the equator, it was moving east at about 7 m/s, quite rapid for an MJO. By 15 February, the MJO had stalled, and weakened. Since then the variability has become more complicated but has involved eastward and westward propagating activity over regions of persistently positive SSTA at 80E and 160E. Toward the end of the period, a merger or consolidation of convection anomalies near ~120E can be seen especially in the upper panel. This projects weakly onto what could be a third MJO.
Figure 2 shows a Hovmoller of 250mb geopotential height anomalies averaged for the mid-latitude band of 30-60N. Again for reference purposes, the dashed arrow follows the leading edge of the jet streak that moved across the North Pacific Ocean in mid- February (same as on Fig. 1). We discussed some of the features seen here in our last two reports, including the retrogression of the ridge in the PNA sector in late January 2003. The longitudinal location of the ridge provides the best illustration of a mid-latitude response to the quasi-oscillatory forcing associated with the MJO activity. The heavy dots show a slow oscillation of the ridge around 120W in the last ~90 days, being farthest east when MJO convection is farthest east. The ridge pushed eastward again in February when the MJO convection moved into the Pacific Ocean region. Since then, ridging in the region has become more widespread as marked by the "Hs".
Another quantity to monitor is the strength of the westerly flow in the atmosphere, both for the atmosphere as a whole and for vertical and latitudinal averages. The latter provide information on the location and shape of the zonal mean mid-latitude jet stream. A time-latitude Hovmoller complements the information on transient and stationary waves (ridges and troughs) presented previously in Fig. 2. In Fig. 3, we first examine the zonal mean 150 mb zonal wind in latitude-time format. A prominent poleward propagation of westerly wind anomalies has been highlighted. We see that the anomalies are not small, exceeding 10 m/s in the northern hemisphere during January. Next we will examine a similar plot but for the atmospheric angular momentum (AAM, click here for more info ).
Figure 4 shows the time evolution of atmospheric angular momentum (AAM) or westerly flow, both for the atmosphere as a whole (bottom time series) and for an average around latitude circles (top plot). The most prominent feature in Fig. 3 (top panel) is again the poleward propagation of zonal mean zonal wind anomalies (actually AAM) from equatorial regions into mid-latitudes of both hemispheres; the anomalies are larger in the Northern Hemisphere. The "W's" follow the poleward movement of maximum westerly wind anomalies from the equator in early-January to 40N by the end of February. The westerlies then persist but they move back south during the second MJO before shifting north again most recently. The longitudes labeled along the equator show times when MJO convection anomalies were over Indonesia (120E). Clearly the zonal AAM anomalies that follow are very different. Strong subtropical westerlies develop near 15N with MJO#1 whereas westerly anomalies shift south to 30N with MJO#2.
Also highlighted n Fig. 4 is an arrow that extends from the time of a sudden stratospheric warming (SSW) in mid-Decmeber 2003 to the start of persistent easterly wind anomalies around 50-60N in early January 2004. The SSW had a strong influence on the atmospheric circulation anomalies in the Arctic for the last 60 days but now appears to be weakening.
The main issue for the realtime situation is that the enhanced westerly flow has shifted further poleward and strong easterlies are developing in the subtropics. The global AAM anomaly has decreased substantially in the last 1-2 weeks (Fig. 4). A large negative mountain torque over Asia during early March 2004 has been directly involved in spinning down the atmospheric westerly flow. The associated momentum and heat transports are contributing to the current zonal mean zonal wind anomalies of easterly at 20N and westerly at 40N. Their surface component is acting to counteract the AAM anomalies produced by the mountain torque and thereby bring the negative angular momentum anomaly back up toward zero. These zonal AAM or wind anomalies can be long-lived. For example, the strong subtropical westerly flow highlighted in Fig. 4 lasted for 20-25 days, shifting slightly poleward.
Figure 5 shows geopotential height anomalies over the polar cap (65-90N) and a time series of the AO. A sudden stratospheric warming (SSW) has been present since about 15 December 2003, and has been affecting surface pressures over the Arctic since about 1 January 2004. The most recent data suggests an easing of these anomalies as the circumpolar vortex contracts slightly.
Figure 6 is a phase space diagram for the MJO, which is currently in phase 4 near the maritime continent. As can be seen, there has been considerable variability in regard to the location of the MJO since 8 February. The current location is consistent with Fig. 1 and suggests there is a weak MJO that could start propagating eastward.
Figures 7 and 8 are current satellite images. Tropical convection is moderately active centered near 110E. Further west tropical cyclones a spinning up off of the equator in each hemisphere, a telltale residual of an MJO, and other equatorial convective flareups. A strong cyclone over the south Indian Ocean (see in Fig. 1 bottom) has moved toward south Africa where convection remains active. Over the Pacific, convection is well defined centered at 160E and may be enhanced by a Kelvin wave moving over the region (see Fig. 1, top).
Figure 9 shows the 7-day mean 150mb vector wind anomalies for the period of 03-09 March 2004. The longitudes contributing to the zonal mean westerly flow anomalies near 40N are: near 60W, near 20W, over the North Pacific and north of Saudi Arabia. Easterly anomalies dominate along 20N in the same regions. A pattern of twin anticyclones is embedded in the flow over the west Pacific with downstream cyclones over the east Pacific Ocean. These and the anticyclones along 30-60E over Africa are remnants of the circulation responding to and interacting with the tropical convection anomalies present since about 23 February (the last two weeks). Fig. 10 shows the daily mean 150 mb wind for 09 March 2004. Equatorial easterly winds over Africa now extend west to South America and westerly winds are still strong over the eastern equatorial Pacific. All indicators suggest another MJO is gathering but a life cycle of zonal wind anomalies linked to the Asian mountain torque in early March is also underway.
Figure 11 are the 500mb height ensemble mean output from NCEP and CDC, respectively, both valid at 0000 UTC Thursday 18 March, 2004. Note the NCEP ensemble mean is a 216-hour forecast while CDC is a 192-hour prediction. Both models indicate a trough moving into the western USA, with CDC a bit slower and with more of a suggestion of difluent flow (and hence possibly deeper). Both schemes also predict a split flow pattern across North America.
Figure 12 shows week 2 streamfunction (PSI) total anomaly forecasts for the period of approximately 17-23 March, 2004 (initial conditions and levels are given on the maps), for CDC, NCEP and Wheeler. The CDC and NCEP are numerical forecasts while the Wheeler prediction is statistical. All the prediction schemes show a L-H-L wave train signature approaching North America. In addition, for the Wheeler and CDC forecasts, there is a signal along and west of the dateline of subtropical anticyclone/cyclone couplets.
As discussed above, tropical convection has consolidated near ~120E during the past 7 to 10 days. The forecast for a convectively active phase of the MJO forms the basis of our outlook below. Assuming it propagates east at ~ 3 deg long/day (about 4 m/s), it will reach the dateline in 20 days, near the end of this outlook period. The precipitation forecast from the Wheeler has an MJO moving eastward (not shown) and the 200mb PSI forecast in Fig. 12 (third panel) has the expected pattern of twin anticyclones and cyclones straddling the west Pacific region.
The CDC ensemble has similar circulation anomalies over the west Pacific, apparently in response to tropical heating. By contrast the NCEP ensemble displays stronger zonally symmetric anomalies, more like those present at the initial time. It may be persisting the zonal mean zonal wind anomalies produced by the east Asian mountain torque on about 3 March 2004. The slower and slightly deeper solution of the CDC ensemble mean in Fig. 11 is preferred, partly because the model appears to be responding to the tropical forcing by week 2. On the other hand, the zonal mean wind anomalies will be important at least for week 1. Confidence is moderate for the following outlook.
Week 1 (11-17 March): As the ridge off the west coast of North America retrogrades to about 140W and amplifies, storm development downstream is expected; however, its magnitude is unclear. At this point the system appears progressive. The storm is expected to begin as an amplifying open wave across the Northern Rockies and Plains this weekend. The system should then move east and northeast toward the Great Lakes and eastern USA, and undergo additional baroclinic development. While heavy snowfall and perhaps blizzard conditions could impact much of the north central states, the southeast USA may experience severe local thunderstorms.
Week 2 (18-24 March): This is the period a western trough should affect much of the lower 48 states. The greatest impact from this development may be several rounds of severe local storms and heavy rainfall especially across the central and southern Plains. In addition, portions of the Rockies and Northern Plains may get late winter/early spring significant snowfall. While temperatures turn colder for much of the west (how cold depends on the cold air source), portions of the southeast states may have record high temperatures. Toward the end of this week 2 period the trough may start to move east, and a major outbreak of Arctic air could impact the central states, then spread east.
Week 3 (25 March - 1 April): This period may begin fairly cold for much of the central and eastern states, while the western USA becomes anomalously warm and initially dry. However, a fairly strong subtropical jet may bring more needed rainfall to the southwestern states. In addition, should the polar jet stream extend and come south, much of the west coast of the USA may be in for an active stormy period, especially California. Some of this precipitation may eventually move into the south central states, with perhaps an increasing chance of more severe thunderstorms (above what would be expected from climatology). The eastern USA would be expected to moderate from perhaps a period of below normal temperatures.