Real-time Weather-Climate Discussion and Predictive
Insights - 10 August 2005
Edward Berry, NWS and Klaus Weickmann, CDC
Since our last discussion (14 June 2005) tropical convective forcing associated
with a weak Madden-Julian Oscillation (MJO) became stationary across the
eastern hemisphere from roughly 40-120E, centered on the Bay of Bengal.
This behavior was part of a general westward shift of tropical forcing from
the western into the eastern hemispheres since the start of 2005, and has
coincided with a westward movement of positive sea surface temperatures anomalies
(~0.5-1.5C) from near the date line to 160E. Combined with the seasonal cycle
this has resulted in the center of warmest ocean temperature shifting further
west into the eastern hemisphere. Since about early July positive anomalies
of tropical convection have started moving slowly east, projecting onto
a weak MJO (see the latest Wheeler plot here
). While there has been some cooling of SSTs along the equator across
the central and eastern Pacific, they remain above normal across the western
Pacific and east Indian Oceans. Part I describes the behavior of the
tropical convective forcing and circulation anomalies since the start of
2005, followed by predictive insights in Part 2. For information on
the status of El Nino and the MJO (including a week 1-2 global hazards outlook)
see the following links:
Latest
CPC ENSO Advisory
Latest
CPC MJO Discussion and tools
Part 1. Weather-Climate Overview
Figure 1 (below) shows three panels consisting of two Hovmoller
plots of outgoing longwave radiation anomalies (OLRA), top, and a plot of
global relative atmospheric angular momentum (AAM) on the bottom. OLR anomalies
are used as a proxy for deep tropical convection anomalies. The top
left is for the last two years for just the MJO, while the plot to the right
focuses on the past six months not only for the MJO, but two other coherent
OLR modes. The contours isolate three coherent OLR modes determined
using a time-space filter, including the Madden Julian Oscillation (MJO,
blue contours), the Kelvin wave (green contours) and the equatorial Rossby
wave (black contours). The blue (red) shading denotes negative (positive)
OLR anomalies meaning enhanced (suppressed) convection. Annotated on
the top panels of Fig. 1 are: 1) "CP coupling" - the coupling of anomalous
convection to warm SSTs in the Nino 4 region (see 18 April discussion here),
2) "IO/WP" - the recent period of enhanced convection over Indonesia, 3)
"EH" - the period of enhanced convection over the eastern hemisphere, 4)
"WH" the period of enhanced convectionover western hemisphere and 5)
MJO#x" - the three MJOs. The lower panel of Fig. 1 is a plot depicting
the latitudinal distribution of zonal mean relative atmospheric angular momentum
(AAM) anomalies (top half) and a time series of global relative AAM (bottom
portion). Annotated are regions of westerly (W) and easterly (E) zonal
mean wind anomalies. See Discussion
number 3 (10 March 2004) for some additional information on AAM.
Figure 1 (recent OLRA plots: link for equator
; link
for north of eq. ; troposphere
AAM plot )
The colored horizontal lines on the top panels match the times of the vertical
lines on the AAM plot. These times are roughly a) mid February, black
lines, when CP was strong and westerly wind anomalies began propagating off
the equator, b) 21 March, brown lines, when easterly wind anomalies returned
to the equatorial regions associated with MJO#1, c) mid May, blue lines,
when westerly anomalies briefly appeared on the equator, and d) mid-June,
orange lines, when easterlies appeared throughout much of the tropics.
General observations from Fig. 1 include: 1) MJO activity was fairly regular
from late 2003 to at least mid 2004; 2) after a period of weak MJOs during
boreal fall 2004 strong coupling of anomalous convection to the relatively
warm central Pacific SSTs occurred, and there was also a general enhancement
of tropical convection across the WH; 3) three additional MJOs occurred after
CP, each one weaker and farther west; and 4) tropical convective forcing
became stationary across the EH centered at about 120E from about mid June-mid
July. Focusing on the top right Hovmoller, which is for the last six
months, and comparing it to the AAM plot, it is seen that as the enhanced
convective envelope shifted from around the date line to about 80E, and
convection decreased or shifted west over the WH. At the same time
AAM went from ~3-4 standard deviations above the 1968-1997 climatology to
2 standard deviations below climatology. In fact, just as westerly
anomalies ("W") moved off the equator during February, easterly anomalies
("E") come off the equator in July. The situations during February
and July involved stationary forcing, but their phases were opposite as the
result of which hemisphere had the enhanced convection (WH/anomalously high
AAM; EH/anomalously low AAM). As will be discussed in more detail in
Part 2, slow changes may again be in progress tied to what could be MJO#4
(For 2005). For the latest satellite imagery see: Latest Indian Ocean
Satellite Picture here
; Latest Western Pacific Satellite Picture here ; Latest GOES
West Satellite Picture here ; Latest GOES East
Satellite here .
While the intent of Fig. 1 was to provide a sense of the of the variability
for the tropical convective forcing, Fig.2 is meant to show more directly
the responses. The annotations are the same as Fig. 1. The top
left is a time-latitude section of 200mb zonal mean zonal wind anomalies,
while the top right is a Hovmoller plot of anomalies of 200mb velocity potential
(chi). For the latter, the blue (red) shading indicates negative (positive)
anomalies of chi suggesting upward (downward) vertical motion and a greater
likelihood of enhanced (suppressed) tropical convection. The bottom
plot is the same AAM graphic shown in Fig. 1.
As would be expected, the behavior of the 200mb zonal wind anomalies shown
on the time-latitude section is similar to the evolution of zonal mean relative
atmospheric angular momentum anomalies. As the westerly anomalies
moved off the equator from mid-February to mid-March, the anomalies of chi
were generally positive (negative) across the eastern (western) hemisphere
meaning suppressed (enhanced) convection. After about a 2-3 month
period of MJOs, the opposite pattern developed in early June. As mentioned
in the introduction and discussed for Fig. 1, this reversal of stationary
forcing was accompanied by the westward shift of anomalously warm SSTs from
the central Pacific into the Indian Ocean.
It is interesting to observe that from roughly mid-June to mid-July the
chi anomalies were generally less than zero around 90W while zonal mean easterly
wind anomalies propagated off the equator. Complex interactions of
the tropical convective forcing with midlatitude baroclinic wave energy dispersion
contributed to persistent twin subtropical anticyclones around the longitude
of 90W. These anticyclones were fundamental for providing the low
shear environment favorable for the tropical cyclogenesis of Major Hurricanes
Dennis and Emily.
Figure 2
Figure 3 presents composites of 150mb vector wind anomalies for the periods
when there was enhanced tropical convection: a) across the central Pacific
and western hemisphere (CP), top panel and b) over the Indian Ocean/western
Pacific and eastern hemisphere (IO/WP), bottom. The discussion issued
June 14 (link)
gives an overview of circulation events during the time interval between
the two panels.
During the period of CP, as would be expected, there were large westerly
anomalies across the tropics/subtropics. There was a signature
of twin cyclones around 120E with distorted twin anticyclones near 150W.
A coherent residual of a Rossby wave train linked to the subtropical anticyclones
hooks up with the large anomalous anticyclonic gyre just southeast
of Greenland, supporting a negative phase of the NAO. Other features
of note include the large cyclonic wind anomaly covering the North Pacific
allowing for a southward displaced polar jet, and the anticyclonic wind anomaly
present just northeast of Brasil (~5N, 50W). Anomalous southerly flow
can be seen entering the southwestern USA, contributing to above average
precipitation.
By the period of mid June through most of July, almost a complete circulation
reversal occurred across much of the globe. For instance, subtropical
westerlies were replaced by easterlies especially across Africa and the
Americas, central Pacific anticyclones were replaced by cyclones and a large
anticyclonic wind anomaly covered the North Pacific into at least the Great
Lakes. Other notable features include the anticyclonic wind anomaly
having shifted northwest to near Florida, the cyclone anomaly over the south
central states, and the large cyclonic gyre across northwest Canada.
For the USA, this distribution brought wet conditions to the southeast,
hot and dry weather to the Plains, a suppression of the southwest USA desert
monsoon and a favorable environment for tropical cyclogenesis across the
Carribean. Finally, the careful reader will see the symmetry of circulation
anomalies across the Americas linked to the tropical convective forcing over
the eastern (WESTERN?) hemisphere.
In some sense, these 2 panels represent opposite ends of the spectrum
of circulation states, with the period of January 25-March 5 similar to
an El Nino (warm event) composite with high AAM and tropical forcing centered
in the western hemisphere, and the reverse by the northern summer (La Nina
basic state). Figure 4 shows these same panels next to the subseasonal
synoptic-dynamic model (SDM) used to chacterize the global scale circulation
anomalies. The past late winter period corresponds with Stage 3 while
the recent June-July period with Stage 1. As already mentioned, this
change is linked to the cooling of anomalously warm SSTs in the central
Pacific. Combined with the normal seasonal cycle of SST, this has
the effect of shifting the center of warmest tropical SST west-northwest
into the eastern hemisphere.
Figure 3
Figure 4 (Most recent
150mb daily mean vector wind anomaly)
2. Predictive Insights
During the release of the last weather-climate discussion dated
June 14, the atmosphere had transitioned from Stage 4 to 1, and the MJO signal
was centered at around 10N/100E (MJO #3). It was understood the MJO
had been stationary for about a week, which added uncertainty to any prediction.
The outlooks reflected that. However, the feeling was to lean toward
the MJO progressing into the western Pacific by the end of week 2 or sometime
during week 3. Instead, the MJO remained stationary.
The forecast period was from June 15-July 5. The atmosphere was
expected to remain in Stage 1 through roughly June 28 (week 2), then transition
to Stage 2 by July 5. Stage 1 persisted this entire time; that is,
there was no Stage 1 to 2 transition. For weeks 1 and 2, the forecast
was for generally warmer than normal temperatures across the central and
eastern USA, and below normal temperature for the northwest states.
Mesoscale convective system (MCS) activity including severe local storms
was a concern for particularly the Central and Northern Plains and Upper
Mississippi Valley, with heavy rainfall being a hazard for the southeastern
states. For the period of June 29-July 5 the possibility of change
toward cooler temperatures for the Northern Plains and Great Lakes was offered,
along with the MCSs and severe thunderstorms spreading southward. The
east and southeast were expected to remain warmer than normal, with above
normal temperatures returning to the west coast. However, it was mentioned
that if the MJO did remain stationary, weather conditions for the USA would
be similar to week 1. Please see the June 14 discussion for details
of the forecast here.
Broadly speaking, the predictions for weeks 1 and 2 did quite well, and
those same weather conditions did persist into week 3. For the specifics
on the occurrences of severe storms, please see the SPC storm reports here. Details on temperatures
and precipitation can be found from the appropriate links on this site.
Referring back to Fig.1 (top right), notice that around July 20-22 there
is a consolidation of coherently propagating westward and eastward modes
of anomalous near equatorial convection. After this consolidation there
was a fairly abrupt eastward shift of the convection along the equator. North
of the equator the Hovmoller plots show (see link)
the anomalous convection remained stationary and was much more widespread,
as would be expected seasonally. During July 2005 consolidation, the
negative chi (see Fig. 2) did spread slowly east from ~80-140E, only to
become stationary again by early August.
The point is that because of behaviors such as consolidations, and interactions
with the extratropics, secondary anomalous convection can move east quickly
into the western hemisphere, while the primary area of forcing remains stationary
or shifts west very slowly (less than the MJO time scale) back in the eastern
hemisphere. This is believed to be true at this time, meaning a stationary
forcing is presently dominating the circulation. Because of the recent
eastward movement of equatorial convection the atmosphere is believed to
have transitioned from Stage 1 to Stage 2 during the period from early July
to the present (August 9). The recent increase of vertically averaged
westerly flow (and AAM link)
across the subtropics is one result of this change.
Currently satellite imagery shows the centroid of the primary tropical
convection at around 15N/135E, extending along 15N from eastern India to east
of the Phillipines. Three-day averaged OLR anomalies were ~ -50-70 W/M**2
(link).
Perhaps linked to SSTAs ~0.5-1.0 C (actual SSTs ~30-31C per TAO data link), a smaller area
of near equatorial thunderstorm activity was present around 160E, and appears
to be moving west perhaps as a Rossby mode. Yet another smaller area
of thunderstorms has developed across the relatively warm (SSTA ~0.5C) central
equatorial Indian Ocean apparently coupled to jet streak dynamics of the
southern extratropics.
The general stationarity of the tropical convective forcing, along with
secondary flare-ups across the Indian and west Pacific Oceans mean a great
deal of uncertainty to the future movement and evolution of this forcing for
at least weeks 1-3. Statistical and numerical models of the MJO are
inconclusive (see MJO forecasts,
Additional
MJO tools and forecasts). However, given the behaviors discussed
in Part 1 including the role of the SSTAs, the recent changes observed with
the equatorial SSTs (cooling across the central and eastern Pacific), and
seasonal cycle, it is believed the stationary component will dominate for
at least the following outlook period. In addition, while the atmosphere
has transitioned to Stage 2 (link
to latest 30-day animation of 150mb vector wind anomalies), there is
a concern there may be some slight shifts back toward Stage 1 (~Stage 1.5)
for at least the rest of August. For the latest on any Atlantic or
east Pacific tropical cyclone hazards (and other parts of the world) please
go to link. Daily monitoring
will be needed to improve predictions, especially of high impact weather.
Figure 5
Week 1 (10-16 August 2005): The atmosphere is expected to be in
a summer time version of SDM Stage 2. Interestingly, this circulation
state did occur about a year ago, and is documented in the August
16, 2004 weather climate report. After weeks of relatively tranquil
weather for much of the USA, a somewhat dramatic change is in the offing,
meaning an overall cool and stormy pattern. A cooler than normal continental
polar airmass should spread southward into much of the Northern and Central
Rockies, Central and Northern Plains and Great Lakes by the end of this
upcoming weekend. Some ensemble output suggests this airmass to be
~1-2 standard deviations below normal. The southeast states and locations
along the west coast should experience near to above normal temperatures.
The likelihood of severe local storms, including heavy rainfall, and MCS
activity are expected to be above climatology during this week, particularly
from the eastern slopes of the Northern and Central Rockies to much of Central
and Northern Plains to the Great Lakes and Ohio Valley. Indeed, many
of these areas have been experiencing some degree of dryness (see CPC
Drought Monitor), and the rainfall is needed. Please refer to the
latest official outlooks and statements from Storm Prediction Center not only for
severe storms, but also fire weather concerns across especially the western
states due to the expected trough. The west coast should remain dry
while the southeast states continues to have the usual diurnal thunderstorm
activity.
Week 2 (17-23 August 2005): The atmosphere is expected to remain
in Stage 2, meaning weather similar to week 1. However, enough westerly
flow may undercut the Alaska blocking to cause the trough ~ 105W to at least
deamplify. In fact, some of the numerical model output suggests the
trough may retrograde back to the west coast by the end of this period.
That scenario is possible since the circulation may shift back toward Stage
1. Should that occur, warmer (above normal) temperatures would return
to the central and eastern states (with possible excessive heat) while the
west cools down. The storm track would also shift back to the north
favoring MCS and overall thunderstorm activity from the Northern Rockies
to the upper Mississippi Valley. The Pacific Northwest may also see
some rainfall.
Week 3 (24-31 August 2005): Stage 1.5 to Stage 2 would be most probable.
By this time the effect of the seasonal cycle as the northern hemisphere
transitions into fall may become more important, and may also allow the MJO
to shift rapidly eastward. That would suggest the westerlies to come
south and increase per climatology. Should the circulation anomalies
occur as would be expected from SDM Stage 2, the 140W ridge-105W trough -
southeastern states ridge pattern relative to the lower 48 states may be
even more energetic and have greater amplitude/larger anomalies than week
1 given additional forcing from the seasonal cycle. That would suggest
an active late August regime for particularly the Rockies and Plains.
Latest CDC
Ensemble Forecast
Latest
NCEP Ensemble Forecast
Additional NCEP
Ensemble output
Latest Canadian
Ensemble Output