Edward Berry, NWS and Klaus Weickmann, CDC

Since our last discussion (dated 12 January 2005) there has been a general eastward shift of tropical convective forcing across the eastern hemisphere, while MJO activity has been very weak (see activity estimate for MJO).  Since  ~1 September 2004 two areas of tropical forcing have been dominant.  One area has been centered across the Indian Ocean with the other over the western Pacific near 150E.  Much of Indonesia has experienced suppressed conditions.  By late January 2005 a change occurred.  Tropical convective forcing became dominant in the equatorial date line/South Pacific region, the area where sea-surface temperature (SST) anomalies have been at least 1 deg C above normal since about July 2004.  The above normal SSTs are the result of an on-going central tropical Pacific warming that projects on the Nino 4 region of El-Nino.  The following describes both the change in the tropical convective forcing, and the subsequent response of the atmospheric circulation.  Uncertainties about the weather-climate anomalies during the next 2-3 months involve whether this represents an important ocean-atmosphere coupling or whether weak MJO activity will shift convection back to the west for a period of time.  Predictive insights follow in Part 2.

Latest CPC ENSO Advisory

Latest CPC MJO Discussion
 

Part 1.   Weather-Climate Overview

Figure 1 (below) shows Hovmoller plots of outgoing longwave radiation anomalies (OLRA) for the equatorial and southern tropics.  OLR anomalies are used as a proxy for deep tropical convection anomalies.  The contours isolate three coherent OLR modes, including the Madden Julian Oscillation (MJO), the Kelvin wave and the equatorial Rossby wave.  The red (blue) shading denotes (positive) negative OLR anomalies meaning enhanced (suppressed) convection. The sloping dotted lines enclose the region of most persistent convective forcing and the - +  - sequences follow the "energy" of an MJO packet consisting of two MJOs. The straight black dotted line shows the equatorial 27C SST contour while the red dotted line shows the axis of 30C SST.  The SST contours remind the reader that the equatorial Pacific is currently in the cooling phase of the seasonal cycle of SST and that the cycle has been unusually well-defined since the 1997-98 El Nino.  

Until about mid December 2004, the dominant area of tropical convective forcing was located generally across the Indian Ocean, while flareups with periods of about 20-30 days occurred over the west Pacific Ocean. The flareups coincide with the axis of warmest water in equatorial "warm pool" regions as seen in the top panel of Fig. 1.  Starting around 22 December a weak MJO shifted east from the Indian Ocean and on 13 January 2005 interacted with an equatorial Rossby mode near 160E.  The resulting consolidation of tropical convective forcing was part of the eastward shift of the convective envelope from the Indian Ocean into the South Pacific Ocean.  Only intermittent episodes of anomalous convection have occurred across the remainder of the tropics since the eastward shift.  Coherent mode atmospheric Kelvin waves (green contours on Fig. 1) have become somewhat more prominent as the total convection becomes more expanded over the global tropics.  The persistent tropical convection across the South Pacific likely represents a coupled ocean-atmosphere response to the moderately large Nino 4  SST anomalies.  
 
The coherent OLR modes suggest another weak MJO has developed farther east near the date line since about 5 February.  Recent satellite images continue to show strong tropical thunderstorm activity across the South Pacific and the South Pacific Convergence Zone (SPCZ). Several tropical cyclones have been spawned across this region during the last few weeks, including category 5 Hurricane Olafl.  A strong subtropical jet (STJ) can be seen exiting this activity and, through interaction with yet another closed low moving onshore into southern California, heading across Mexico and into southern USA.  This is an example of the linkage between this forcing and the recent weather events across that country.  For the latest satellite imagery see: Latest Indian Ocean Satellite Picture;  Latest Western Pacific Satellite Picture and
Latest GOES West Satellite Picture

  Figure 1  latest images

Figure 2 (below) is a time-latitude section of 200mb zonal mean zonal wind anomalies.  It depicts a zonal averaged view of recent weather-climate anomalies. The discussion will focus on the transition of the circulation during late January 2005 and the wave-mean flow interaction that accompanied it. First, however, a difference in time scales along 30N and 60N is worth noting. At 30N, 20-30 day quasi-oscillations are prominent.  The mean separation between the positive zonal wind anomalies is ~25 days, similar to the 20-30 day convective foring over the ocean "warm pool".  At 60N ~10 day variations dominate (see dotted line along 60N in this figure). These represent the zonal mean response to baroclinic life cycles that can include several synoptic events or a family of cyclones.  At times there is a meridional coupling or propagation that links these time scales across latitude bands.  This figure shows a blowup of the first 75 days of Fig. 2 and illustrates two of these events (labelled "LC1 wave break").  

The most recent transition is associated with a time-space quadrapole of strong zonal wind tendencies that is highlighted on Fig. 2.  The quadrapole develops as the tropical convective envelope moves east and represents a transient tropical extratropical interaction involving Rossby wavetrains, energy propagation and wave breraking.  The "+  -" on ~24 January 2005 represents a momentum transport into the subtropics followed by a transport back out of the subtropics 10 days later.  By around 5 February, tropical convection became persistent across the South Pacific and a zonal mean split flow type pattern returned - similar to what was observed several times during September-December 2004.  There were, however, some differences.  As shown in Figure 2, subtropical westerly wind anomalies are now present in both hemispheres having shifted from the equator only recently. This is the atmosphere's response to the Nino 4 central Pacific SST anomalies.  It has similarities to an El Nino composite response and to Stage 3 of the Synoptic Dynamic Model (for updated version click here).

Figure 3 (below) shows the flux convergence of zonal momentum for the last 90 days.  The two black arrows give the sense of the momentum transports. For timing purposes 24 January 2005 has been highlighted on Fig. 3 and also on both Figs. 1 and 2.  The wavetrain evident at this time (panel 2 of Fig. 3) is linked to tropical forcing from the eastward shift of convection. It is tied to the development of a blocking anticyclone (and negative phase of the North Atlantic Oscillation (NAO link )) across the North Atlantic.  In panel 3 of Fig. 3, the wavetrain is seen to bifurcate over Europe with a portion of the energy moving east and a portion southeast.  An anticyclonic wave break is also evident downstream of the Atlantic anticyclone. 

Figure 3.

  2.   Predictive Insights

In the 15 January 2005 discussion, we suggested that after relatively strong subtropical westerly flow typical of Stage 3 of the subseasonal Synoptic Dynamic Model (SDM) ( link to discussion with SDM ) during week 1, tropical convection would redevelop across the Indian Ocean and Indonesia. This assumed Stage 3 would be followed by 4, 1, 2 and then Stage 3 again. The tropical convection did not follow this trajectory, instead persistenting in the central equatorialPacific and keeping the circulation anomalies in Stage 3. 

For the USA we believed, for week 1, the weather would be relatively tranquil across much of the west and central, with colder and somewhat wetter than normal conditions across the Northeast (typical for a positive PNA).  It was mentioned that cyclogenesis off the New England coast would have to be monitored.  Confidence was good.  The prediction for week 1 did reasonably well.  Although, and very importantly, we did not explicitly predict the record setting snowstorm that affected much of the Northeast states.

For weeks 2 and 3, with very low confidence, a negative PNA pattern was predicted, meaning cold/wet (warm/relatively dry) for much of the west (east).  As discussed in Section 1, an apparent coupling to the weak central Pacific El Nino occurred, meaning a general persistence of a circulation state depicted by Stage 3 of the SDM.  We did not forecast this transition although a "climate watch" for such behavior was appropriate given seasonal cycle considerations and the Nino 4 SSTs. 

For the following (through 16 March 2005) the overall coupled response to the weak central Pacific El Nino is expected to continue (SDM Stage 3).  However, there will be variations in its amplitude and the synoptic details are unclear after week 1.  Adding further to the uncertainty is the split flow pattern affecting the USA, especially with an enhanced STJ/southern storm track.  Even though there is good confidence about the flow pattern, confidence in the predictions are low, especially for weeks 2 and 3.  For example, any coupling of the northern and southern branches of the westerlies could easily lead to a high impact baroclinic synoptic weather event (with all relevant hazards including winter storm and severe thunderstorms), especially given the seasonal cycle.

Week 1 (24 Feb -2 March 2005):  A ridge (mid/upper troposphere) is expected to persist along or just off the coast of western Canada, with a trough across the southwest states (with perhaps closed lows).  This will mean a strong southern storm track and a source of cold air for the north central and northeast states.  Much of the area from California through the Central and Southern Plains to a good portion of the east coast may have above normal precipitation.  The northwest states will likely stay dry (where drought continues).  Locations from the Front Range of the Rockies to southern New England may experience significant winter precipitation.  Severe local storms may be a concern for parts of the southern and southeast states.  Much of the country may experience below normal temperatures.  The exceptions may be the Pacific Northwest states and along the Gulf coast. 

Latest CDC Ensemble Forecasts

Latest NCEP Ensemble Forecasts

Additional NCEP Ensemble Output

Latest Canadian Ensemble Output