Real-time
Weather-Climate
Discussion and Predictive Insights –
Edward Berry, NWS and Klaus Weickmann, CDC
Since our last discussion (7 October 2005), equatorial sea surface temperatures (SSTs) across the central and eastern Pacific have cooled considerably, leading to widespread negative anomalies (SSTAs) of ~ 1-2C. SSTAs across much of the western and south Pacific have remained positive 0.5-1.5C with actual SSTs in excess of 29C. The SST pattern resembles a basin wide La Nina cold event. The signal from the Madden-Julian Oscillation (MJO) has continued to be weak. The Wheeler plot shown here has been depicting a very strong MJO while the coherent modes have a weak to moderate MJO. This disparity and the role of La Nina is discussed later. Tropical convective forcing became strong across the region of Indonesia by early December, and then late in the period extended east and southeast along the South Pacific Convergence Zone (SPCZ) in the area of very warm SSTs mentioned above.
Part I presents an overview of the SST and tropical convective evolutions for the past year, followed by more recent behaviors of the tropical forcing and circulation anomalies. This particular report gives the opportunity to document a transition from El-Nino to La-Nina during the last year. A synopsis of the current weather-climate situation with predictive insights follows in Part 2. For information on the status of El Nino and the MJO (including a week 1-2 global hazards outlook) please see the following links:
Latest CPC MJO Discussion and tools
Part 1.
Weather-Climate
Overview
Figure 1 (Hovmoller
plots of SSTs and SSTAs; latest images of various SST monitoring
tools are here
)
Figure 2 (Hovmoller plot of seasonally
varying total OLR; latest image here)
Figure 3 are time-longitude plots of three-day averages of outgoing longwave radiation
anomalies (OLRA) for the period mid-October 2005 to early
February 2006. Time and
space filtering are
used to isolate coherent
convectively coupled modes such as the MJO, equatorial Rossby
waves and Kelvin waves (see 10
August discussion and previous issuances for details).
Contours depict these modes and go from upper left to lower right
(upper right to lower left) to depict eastward (westward)
propagation. Here
we will focus on the more recent evolution of the tropical convection.
The top (bottom) panel is for the near-equatorial (southern tropics).
The ENSO cycle consolidation is evident in the anomalies by early
December
along the equator and especially across the southern tropics at the
same time (indicated by the
dashed vertical line). During December positive anomalies of
tropical convection remained stationary across much of
Indonesia.
Beginning late December 2005, strong tropical convection developed
across South America, apparently linked to interactions with
the northern extratropics. A coherent atmospheric Kelvin
wave (KW) was initiated by this activity (see the green
square in
the Fig. 3 top panel just after Christmas) and propagates to 0E by 9
January 2006. The KW moved at 15-20 m/s
and interacted with an equatorial Rossby
wave to give an
intense flareup of tropical thunderstorm activity on ~18 January.
Thus Event
#1
on Fig. 3 could be characterized as a subseasonal consolidation of
convection. The KW
remained coherent as it propagated well into the western hemisphere
(WH). An important ramification
of the flareup was the development of a slow eastward moving
component of the tropical convection, which projected onto a
weak-moderate MJO. The MJO is best seen on the southern tropics
OLRA plot (Fig. 3 bottom panel; blue contours, 10 Wm-2
interval).
The slow OLR component, identified as a weak MJO, propagated at 4-5 m/s
into the WH along the SPCZ. There were two additional
flare-ups of convection (numbers 2 and 3) within the envelope, with
intense
thunderstorm clusters located over the South Pacific subtropical SST
gradients. Recently, the convection has stalled and even started
drifting slowly northwest. A weakened
signal of the MJO has propagated into the WH.
Figure 3 (recent Hovmoller OLRA plots: equator
; northern
tropics. ; southern
tropics
)
Figure 4 (below) is the first in a series of charts that will focus on the circulation variability in the northern extratropics and interactions with the tropics. It shows the 250mb zonal wind anomaly averaged between 25-40N; Ws and Es indicate westerly or easterly flow anomaly. The longitude band from 100E to 120W is the region that determines the strength and position of the East Asian Jet (EAJ) or North Pacific Jet Stream. The EAJ strengthened around 10 November 2005 but extended barely past the date line until 10 December (with the exception of a spectacular jet streak in late November). Thereafter it became more intense and expanded to 120W during 15 December-4 January 2006, with anomalies in excess of 40 m/s at times. Anomalous storm track activity with the jet impacted the USA west coast from Northern California to particularly the Pacific Northwest. Strongest tropical convective forcing during this time remained across the EH ~120-140E, a location statistically related to a retracted jet, not the observed expanded jet. More discussion of this jet extension will be presented below. After 4 January, the EAJ collapsed and easterly zonal wind anomalies appeared for the first time since early December 2005 (Fig. 4).
Figure 4 (Hovmoller plot of 250mb zonal wind anomalies from 25-40N. Click here to do additional time-longitude plots)
Figure 5 (below) reviews the behavior of atmospheric relative angular momentum (AAM) since about 1 November. The top panel is a time-latitude plot of zonal mean anomalies of tropospheric AAM (warm colors for westerly/cold colors for easterly flow), along with the total (not anomalous) zonal and vertical mean AAM (contours). The bottom panel is the time series plot of global AAM anomaly for the troposphere (to 100 mb). A stratosphere plot seen here shows the weakening negative phase of the equatorial QBO, and the major sudden stratospheric warming (SSW) that developed recently over the Arctic.
A set of vertical dashed lines depict the period when the EAJ was extended across the Pacific. Zonal mean westerly anomalies appeared along 30N in early November (shown by vertical red line) and they became stronger after mid December 2005. Furthermore, there is a hemispheric symmetry of the zonal wind anomalies, with westerlies in mid-latitudes and easterlies poleward and equatorward. The strong mid-latitude westerlies were supported by strong flux convergence of AAM out of both the subtropical and polar regions toward ~35-40N (see plot here if interested). We speculate wave energy sources linked to east Asian topography and cold air outbreaks contributed to the maintenance of the extended EAJ. Some of these waves and other disturbances are highlighted later.
After early January, the mid-latitude westerlies collapsed, particularly across the northern midlatitudes. Additionally, negative (easterly) AAM anomalies propagated into the subtropics between 25-30N and 25-35S. The recent appearance of equatorial westerly flow is attributable to the MJO signal discussed above, which is moving into the WH. Good symmetry of zonal mean AAM anomalies continues and the midlatitude westerlies are displaced poleward from their locations during December.
Figure 5 (most recent plot of tropospheric AAM here; other plots here )
The Fig. 6 sequence (below) is a selection of daily mean maps of 250mb vector wind anomaly and are used to summarize the observed circulation anomalies, including the ridges and troughs that influence local weather. In all of the maps that follow, the red H's (L's) are the anticyclonic (cyclonic) circulation anomalies (see scale at the bottom), the dashed orange ovals indicate the twin anticyclones forced by the persistent Indonesian convection, and the black arrows give a sense for the wind directions.
We start the sequence on 9 November (Fig. 6) when Rossby wave dispersion (RWD) occurred across the Pacific Rim (white dashed curved line), linked to a convective flareup over the west Pacific. The orange oval highlights the location of twin anticyclones linked to tropical convection anomalies. The flareup is also outlined by a purple square on the top panel of Fig. 3. As part of the RWD, a blocking anticyclone developed over northeast Asia, the first in a series of such events. At least 3 more RWDs linked to the Indonesian tropical convection occurred afterwards (15 Nov, 28 Nov and 4 Dec) as is shown in Fig. 6. During this period (9 November-4 December), not only did the northeast Asian blocking anticyclone persist, but it expanded and contributed to large anticyclonic anomalies across most of the Arctic. This produced negative projections onto both the AO and NAO. The growing easterly anomalies in high latitudes (55-90N), suggest a southward momentum flux into mid-latitudes. The negative tilt of disturbances in the 40-80N band on 4 and 21 December 2005 would give a southward flux. Additionally, positive mountain torques contributed to a momentum increase along 30N (not shown). This was part of a persistent dipole pattern of positive torque near 30N and negative torque near 50N (not shown). The pattern reversed itself ~27 December just before the breakdown of the EAJ.
Figure 6 (sequence
of 250mb daily mean vector wind anomalies; latest animations of various
fields are here
)
The jet was most expanded across the Pacific from 10 December until
early January 2006. An example of this expansion is
shown on 21 December. The meridional tilt of the perturbations on
either side of
~35N suggests a flux convergence of momentum into the
midlatitudes from both the subtropics and the Arctic. Such postulated
"eddy" feedbacks can help mantain the EAJ against tropical convective
forcing that may seek to weaken it. In terms of our synoptic
dynmaic
model (SDM), Stage 3 has a strong extended jet over the
Pacific but the tropical convection and subtropical circulation are
quite different from what has been observed. This period has
proven difficult to predict in, both for models and human forecasters.
Recall from Fig. 3 that after 21 December tropical convection became intense across northern South America and near Africa. A long-lived KW developed out of this activity. Examination of daily animations from 21 December-4 January revealed subtropical wavetrains interacting with and amplifying the circulation features over South America and Africa, possibly helping to excite the convection there. Also, several anticyclonic wavebreaking episodes across the eastern Pacific led to equatorial westerly flow anomalies over the east Pacific by 4 January (see the yellow "W"). These westerlies shift west and then expand eastward during mid-late January. By 4 January the jet was starting to collapse across the Pacific as convection began increasing across Africa and the Indian Ocean. Easterly anomalies (with globally averaged AAM as low as 2 standard deviations below the 1968-1997 climatology) then dominated much of the tropics and extratropics as convection was active across Indonesia and northern Australia, shifting south with the seasonal cycle. From 4-18 January, the circulation would have been represented by SDM Stage 1. However, troughs and ridges remained progressive across the Pacific and North America, and that allowed more severe cyclonic storms to impact the Pacific Northwest.
On 18 January, enhanced divergence was observed
across the western Pacific due to the tropical flareup from the KW and
Rossby wave interaction mentioned earlier. One response
was the twin
subtropical cyclones east of the date line by 24 January and the
start of strong westerly wind anomalies in the upper equatorial
regions.. As
the MJO developed and moved east (recall above), convection became very
intense along
the SPCZ.
The globally averaged AAM increased rapidly to about 1
standard deviation above climatology by early February 2006. By 1
February poleward displaced twin anticyclones had moved to about the
date line. The poleward displacement is consistent with the
anomalous
easterly flow throughout the subtropics linked to La-Nina. The 5
February map is a good example of not only the
twin anticyclones moving into the WH, but also linkages with fast
moving midlatitude wavetrains. Notice that the cyclonic
circulation anomaly around southwest Alaska on 1 February retrogrades
to
near Asia by 5 February. This is part of a wave 0-2 retrogression
that has been in progress across the north polar latitudes for roughly
a week. SDM Stage 3 would best describe the global circulation
after about 1 February.
Brief mention will be made of the recent major
SSW since it appears to be influencing the troposphere. The top
panel of Fig. 7 is a time series of 30mb temperature
centered on the North Pole, from the Tokyo Climate Center.
A warming of ~50C is evident during January 2006 (smooth curve
is climatology). The bottom panel depicts a time-height section
of geopotential anomalies averaged over 65-90N and an AO index time
series from CPC.
The appearance of mid/upper tropospheric positive height
anomalies after about 8 November can be seen, only to intensify to at
least 2 standard deviations above climatology during December (green
rectangle). Evidence from other diagnostics (e.g., Eliassen-Palm
fluxes) show an increased upward propagation of wave
eneregy into the stratosphere near 1
January, 2006; around the time the warming commenced at this level (at
10 mb it looks earlier, at least in the winds). The arrow
pointing
up on the time-height section suggests when upward propagation into the
stratosphere occurred. During the latter part of
January, ~4 sigma height anomalies are shown in the
stratosphere. There have also been at least 2 episodes of
downward
extension of positive height anomalies from the stratosphere to the
earth's
surface. These mainly where linked with the anticyclonic wind
anomalies and high heights seen over the eastern hemisphere Arctic on
18 January (at 80N, 40E) and 24 January (at 70-80N,
20-90E), as shown at 250 mb in Fig. 6.
Figure 7 (top:
time series of 30mb temperatures from 65-90N (?) lastest
image; bottom height-time section of geopotential height anomalies
with AO time series, latitude band 65-90N latest
image).
Figure 8 (Description of the
SDM is presented here)
2. Predictive Insights
La-Nina continues with negative SSTAs of 1-2C covering nearly
all of the central and eastern equatorial Pacific Ocean basin, and
extending down to depths of at least 200m. However, cool
anomalies have weakened slightly during the past few days. Actual
SSTs along the
equatorial cold tongue are as low as 25C, while values from 29-31C
cover much of the western and south Pacific. The warmest SSTs are
at ~150E with the coolest at 140W, allowing for an enhanced SST
gradient
which is once again supporting an intensification of the trade winds
throughout much of the central and western Pacific. Finally,
anomalous easterly flow has become quite pronounced across the
subtropical atmosphere. Latest prediction from CPC expects these
cold event conditions to continue for at least the next 3-6 months (see
latest TAO data here).
At this time the signal from the MJO is weak with
the upper tropospheric divergent response moving into the EH.
Satellite
imagery (eastern
hemisphere,
full-disk east Pacific,
full-disk west Pacific,
full-disk Indian Ocean,
full-disk Africa;
other
images available here)
has enhanced convection across tropical South America and especially
South Africa, with slow intensification occurring across the South
Indian Ocean and Indonesia. Strong interactions with the
extratropics can be seen from much of this activity through well
defined subtropical jets (STJs). While convection has weakened
considerably across the SPCZ during the past week, enhancement has been
drifting slowly west and north, now located ~ 5N/140E (latest 3-day
averages of OLR
total and anomalies here).
Additionally, very strong suppression of convective activity has
been present from the South Indian Ocean across Australia into portions
of the South Pacific, with enhancement to the north (latest 7-day
average here).
Statistical and
numerical
models of the MJO (see MJO forecasts, Additional
MJO tools and forecasts)
generally support the notion of enhanced convection returning to the
Indian Ocean/Indonesia regions by the end of week 2. SSTs have
also warmed to slighly above average across portions of the central and
southern Indian Ocean (~+ 5.-1.5C), which can be a precursor to a
re-emerging MJO into the EH. Finally, SSTs also remain ~.5-1.5C
above normal across the tropical southwest Pacific, as well as the
subtropical North and South Pacific and Atlantic Oceans. These
basins will also contribute tropical convective forcing at times (plot of
weekly
global SSTAs here).
Monitoring experience suggests a
consolidation of the tropical forcing will occur near 120E and extend
east and southeast along the SPCZ to perhaps 160E, by the end of week
2. This pattern would also be consistent with La-Nina.
The global AAM
anomalies are still
about 1 standard deviation above climatology, due to enhanced zonal
mean westerly flow across particularly the southern subtropics
With the
current pattern of tropical forcing and the appearance of strong
WH STJs, the atmosphere is believed to be transitioning to SDM
Stage 4. The strong tropical forcing from south of the equator in
the Pacific in the last two weeks, interacting with the retrogressive
wave 0-2 high
latitude transient discussed in Section 1, has produced a robust
positive PNA pattern
across the Asia-North American sector during the past weekend.
During this upcoming week, as the convective
signal re-emerges over the EH,
the most significant lobe of the polar voretx should lock in across
central and east Asia, maintaining a contracted EAJ. That would
argue for a transition to SDM Stage 1, meaning a negative PNA pattern.
Additionally, given the north polar latitude easterly wind
(positive geopotential height) anomalies, it is plausible that any
ridging in the central Pacific may link up with possible blocking at
the higher latitudes. Thus in addition to a negative PNA, there
may also be a negative AO and NAO. We would expect the
circulation response as shown by SDM Stage 1 to mature during week
2. Variations in amplitude of the
ridge-trough-ridge negative PNA pattern across the USA would be
expected. The most
meaningful insight offered for week 3 is peristence of the situation
from
week 2, given the La-Nina. The CDC week 2 ensemble mean 500mb
height
anomaly forecast, along with the calibrated temperature and
precipitation tercile pobabilities, is presented below to assist with
the low confidence forecast that follows.
Figure 9. (Week 2 ensemble mean 500mb height anomaly from the CDC ensemble; latest images here)
Figure 10. (Week 2
calibtrated tercile probabilities for temperature from the CDC
ensemble; latest
forecast)
Figure 11. (Same as 10 but for precipitation; latest forecast)
Week 1 (15-21 February 2006): As stated above, a transition to SDM Stage 1 is probable by the end of this period, particularly if tropical convection becomes intense ~120E. That would lock in the developing western USA trough with well below normal temperatures and lead to the opposite for the southeast USA. True Arctic air, perhaps from northern Asia and Siberia, is likely to penetrate at least into the northern Rockies and Plains, and even spill over into the Pacific Northwest. The coldest temperatures of this winter are likely for many parts of the north and west CONUS (at least 2 standard deviations below climatology -- see CDC and NCEP ensemble links below), including possible records for places such as California. Some of the bitterly cold air will spread into the northeast states by the end of this period.
As shown by the models, an initial event of
baroclinic cyclogenesis is expected across the Plains during the next
couple of days. Even though moisture transport from the Gulf of
Mexico and other lower latitude sources will initially be limited due
to the recent surges of cold air, heavy snow and blizzard conditions
for much of the northern Plains into the Great Lakes seems probable.
Intense thunderstorm activity and heavy rainfall (including over the
snow cover across the Northeast raising flooding concerns) may occur in
the warm sector as moisture becomes more plentiful.
Another concern is to monitor how far off the west coast troughs may dig in this pattern. A cold and wet scenario is not out of the question for locations such as California, with possibly even snow for locations such as Seattle. In fact, the first of these events may occur this upcoming weekend particularly if there is interaction with the current east Pacific STJ.
Most models suggest the above mentioned west coast system will get kicked out into the Plains by early next week, which is reasonable. By this time moisture transport should not be an issue. Thus a second storm system on the Plains may occur, and then move into the Great Lakes/Ohio Valley and eastern states. The track for this second storm may be shifted slightly to the south from its predecessor. Heavy snowfall for much of the northern and central Rockies, including perhaps the Front Range, speading across the central and northern Plains/Upper Mississippi Valley would be a concern. Strong winds combining with cold temperatures could again lead to blizzard conditions. Meanwhile, heavy rainfall and even severe thunderstorms may occur anywhere from the Ohio Valley into the southeastern states. Finally, icing is a possibility in between the winter and spring-like weather, which may include parts of the Northeast.
Please see the CPC Drought Monitor for areas of dryness and the latest official outlooks and statements from 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.
Week 2 (22-28 February 2006): Please see Figs. 9-11. We think these are reasonable forecasts and other models are similar. SDM Stage 1 would be most probable suggestive of a storm track from the Pacific Northwest to the central Plains. Perhaps a modification we would make would be to increase the probabilities for wetness along the Pacific Northwest, as well as the northern and central Rockies into the Plains.
Week 3 (1- 7 March 2006): We would expect the week 2 situation to persist during much of this period although an eastward shift of the ridge-trough pattern may occur.
Additional NCEP Ensemble output
Latest Canadian Ensemble Output
Latest CDC Week 2 Forecasts from the CDC EnsembleLatest Deterministic ECMWF Forecasts