Calhoun: The NPS Institutional Archive
DSpace Repository
Faculty and Researchers Faculty and Researchers' Publications
2015-04
A Composite Perspective of the Extratropical
Flow Response to Recurving Western North
Pacific Tropical Cyclones
Archambault, Heather M.; Keyser, Daniel; Bosart, Lance
F.; Davis, Christopher A.; Cordeira, Jason M.
Montly Weather Review, Volume 143, pp. 1122-1141, April 2015
http://hdl.handle.net/10945/45742
This publication is a work of the U.S. Government as defined in Title 17, United
States Code, Section 101. Copyright protection is not available for this work in the
United States.
Downloaded from NPS Archive: Calhoun
A Composite Perspective of the Extratropical Flow Response to Recurving
Western North Pacific Tropical Cyclones
HEATHER M. ARCHAMBAULT*
Department of Meteorology, Naval Postgraduate School, Monterey, California
DANIEL KEYSER AND LANCE F. BOSART
Department of Atmospheric and Environmental Sciences, University at Albany, State
University of New York, Albany, New York
CHRISTOPHER A. DAVIS
National Center for Atmospheric Research, Boulder, Colorado
JASON M. CORDEIRA
Department of Atmospheric Science and Chemistry, Plymouth State University, Plymouth, New Hampshire
(Manuscript received 22 August 2014, in final form 18 December 2014)
ABSTRACT
This study investigates the composite extratropical flow response to recurving western North Pacific
tropical cyclones (WNP TCs), and the dependence of this response on the strength of the TC–extratropical
flow interaction as defined by the negative potential vorticity advection (PV) by the irrotational wind asso-
ciated with the TC. The 2.58 NCEP–NCAR reanalysis is used to construct composite analyses of all 1979–2009
recurving WNP TCs and of subsets that undergo strong and weak TC–extratropical flow interactions.
Findings indicate that recurving WNP TCs are associated with the amplification of a preexisting Rossby wave
train (RWT) that disperses downstream and modifies the large-scale flow pattern over North America. This RWT
affects approximately 2408 of longitude and persists for approximately 10 days. Recurving TCs associated with
strong TC–extratropical flow interactions are associated with a stronger extratropical flow response than those
associated with weak TC–extratropical flow interactions. Compared with weak interactions, strong interactions
feature a more distinct upstream trough, stronger and broader divergent outflow associated with stronger midlevel
frontogenesis and forcing for ascent over and northeast of the TC, and stronger upper-level PV frontogenesis that
promotes more pronounced jet streak intensification. During strong interactions, divergent outflow helps anchor
and amplify a downstream ridge, thereby amplifying a preexisting RWT from Asia that disperses downstream to
North America. In contrast, during weak interactions, divergent outflow weakly amplifies a downstream ridge,
such that a RWT briefly amplifies in situ before dissipating over the western-central North Pacific.
1. Introduction
Tropical cyclones (TCs) undergoing extratropical tran-
sition (ET), a change from a warm-core, axisymmetric
system to a cold-core, asymmet ric system (e.g ., Klein et al.
2000; Jones et al. 2003; Harr and Archambault 2015), may
significantly perturb their environment as they recurve into
the midlatitudes (e.g., McTaggart-Cowan et al. 2007; Harr
and Dea 2009; Cordeira and Bosart 2010). The poleward
and upward transport of low potential vorticity (PV) air
associated with TC recurvature tends to amplify a ridge and
intensify a jet streak along the meridional PV gradient
marking the jet stream (e.g., Agustí-Panareda et al. 2004,
Publisher’s Note: This article was revised on 10 June 2015 in order
to include an updated acknowledgements section.
* Current affiliation: NOAA/OAR Climate Program Office,
Modeling, Analysis, Predictions and Projections Program, Silver
Spring, Maryland.
Corresponding author address: Heather M. Archambault, 1315
East–West Hwy., Silver Spring, MD 20910.
E-mail: heather.archambault@noaa.gov
1122 MONTHLY WEATHER REVIEW VOLUME 143
DOI: 10.1175/MWR-D-14-00270.1
Ó 2015 American Meteorological Society
2005; Riemer et al. 2008; Riemer and Jones 2010; Cordeira
2011; Grams et al. 2013a; Archambault et al. 2013).
Ridge amplification and jet streak intensification are
enhanced by diabatic heating associated with the recurv-
ing TC (e.g., Atallah and Bosart 2003) and attendant
frontogenesis in its northeast quadrant (e.g., Torn 2010).
Above the level of maximum diabatic heating associated
with clouds and precipitation, PV is reduced in conjunc-
tion with diabatic vertical redistribution. In addition, the
divergent outflow generated by diabatic heating advects
low PV toward the PV gradient/jet stream (Archambault
et al. 2013, their Fig. 4).
Given a strong, continuous meridional PV gradient
(i.e., waveguide; Martius et al. 2010), ridge amplification
and jet streak intensification will excite or amplify a baro-
clinic Rossby wave train (RWT) that disperses eddy
kinetic energy downstream while initia ting surfa ce cy-
clogenesis downstream (e.g., Riemer et al. 2 008; Harr
and Dea 2009; Keller et al. 2014). Thus, a recurving TC
may indirectly reconfigure the extratropical flow pattern
and influence the sensible weather thousands of kilome-
ters downstream within a few days (e.g., Archambault
et al. 2013, their Fig. 1).
On average, the North Pacific flow pattern becomes
significantly amplified for approximately four days fol-
lowing western North Pacific (WNP) TC recurvature
(Archambault et al. 2013). However, the extratropical flow
response to TC recurvature may range from a marked flow
amplification [e.g., with TC Oscar (1995); Archambault
et al. (2013),theirFig.1;Harr and Archambault (2014)], to
a strengthened jet stream without substantial flow amplifi-
cation [e.g., with TC Jangmi (2008); Grams et al. (2013a,b)],
to little change in the flow pattern [e.g., with TC Opal
(1997); Harr and Elsberry (2000)]. Anticipating the
extratropical flow response to TC recurvature is crucial
given that flow amplification downstream of a recurving
TC may induce extreme weather (e.g., Cordeira and
Bosart 2010; Grams et al. 2011; Chaboureau et al. 2012;
Pantillon et al. 2014) and contribute to reduced mid-
latitude predictability (e.g., Harr et al. 2008; Anwender
et al. 2008; Reynolds et al. 2009; Keller et al. 2011;
Pantillon et al. 2013, 2014; Harr and Archambault 2015).
The primary factors influencing the downstream flow
response to TC recurvature are (i) the large-scale flow
pattern into which the TC is moving (e.g., Harr and Dea
2009; Riemer et al. 2008; Riemer and Jones 2010,
2014),
and (ii) the interaction between the TC and an extra-
tropical disturbance such as a trough or jet streak (e.g.,
Klein et al. 2002; Ritchie and Elsberry 2007; Riemer and
Jones 2010, 2014; Grams et al. 2013b; Keller et al. 2014).
Characteristics of the recurving TC, such as size and
intensity, are thought to be secondary factors (e.g., Harr
and Dea 2009; Archambault et al. 2013).
The large-scale extratropical flow pattern into which
a TC recurves varies by time of year and by ocean basin.
Compared with August–November, June and July are
relatively unfavorable for the North Pacific flow pattern to
become significantly amplified following WNP TC re-
curvature (Archambault et al. 2013,theirFig.15),con-
sistent with the tendency for the North Pacific waveguide/
jet stream to be relatively weak in June and July com-
pared with August–November. A compositing study by
Quinting and Jones (2014) indicates that, whereas RWT
activity is significantly above average following WNP TC
recurvature, it is not significantly above average following
North Atlantic TC recurvature. Based on case studies
(e.g., Grams et al. 2011; Pantillon et al. 2013, 2014), anti-
cyclonic Rossby wave breaking instead of Rossby wave
dispersionmaybemoretypicalfollowingNorthAtlantic
TC recurvature, perhaps owing to the climatologically
short and weak waveguide/jet stream over the North At-
lantic compared with over the North Pacific.
The large-scale extratropical flow response to TC
recurvature also depends on the interaction of the TC
with disturbances in the extratropical flow. The TC–
extratropical flow interaction, or phasing, can be con-
sidered favorable or unfavorable. An example of an
unfavorable phasing between the TC and extratropical
flow is the recurvature of WNP TC Jangmi (2008). As
the TC recurved into the base of a WNP trough, the jet
stream was enhanced downstream but Rossby wave
amplification and dispersion did not occur (Grams et al.
2013a,b). Through numerical modeling experiments,
Grams et al. (2013a) demonstrate that had TC Jangmi
recurved ahead of the trough rather than into the base of
the trough, a high-amplitude RWT likely would have
been induced. They find that the shift in TC position
required to discriminate between RWT and no-RWT
scenarios is only 130 km, which they note is the average
48-h TC position error for ensemble mean ECMWF
forecasts (Lang et al. 2012).
In a recent climatological study (Archambault et al.
2013), the top quintile of TC–extratropical interactions,
defined by the magnitude of negative PV advection by
the irrotational wind associated with the recurving TC, is
found to be associated with a sustained, highly statisti-
cally significant amplification of the North Pacific flow.
In contrast, the bottom quintile of TC–extratropical flow
interactions is found to be associated with a shorter-
lived, less statistically significant amplification of the
North Pacific flow. In the present study, the climato-
logical study of Archambault et al. (2013) is used as the
basis for a comprehensive composite analysis of the
extratropical flow response to WNP TC recurvature. In
particular, key synoptic–dynamic differences are illus-
trated between strong and weak TC–extratropical flow
APRIL 2015 A R C H A M B A U L T E T A L . 1123
interactions, which are important to understanding dif-
ferences in the downstream extratropical flow response
to TC recurvature.
The remainder of this paper is organized as follows.
Section 2 describes the methodology. Section 3 presents
composite analyses of the midlatitude flow response to
recurving WNP TCs and compares composite analyses
of strong and weak TC–extratropical flow interactions
during WNP TC recurvature. Section 4 provides a dis-
cussion, and section 5 contains a summary and an
overview of future work.
2. Methodology
a. Recurvature-relative compositing methodology
To examine the extratropical flow response to recurv-
ing WNP TCs, all 292 recurving WNP TCs identified
during 1979–2009 from the Japan Meteorological Agency
best track dataset by Archambault et al. (2013) are com-
posited. Tropical cyclone recurvature is defined following
Archambault et al. (2013) as a change in TC heading from
westward to eastward as a TC moves poleward. The re-
curvature point corresponds to the most westward posi-
tion of the recurving TC. All recurving TCs are required
to be at tropical storm intensity or greater at the time of
recurvature (i.e., T 1 0 h), and to eventually become ex-
tratropical (i.e., complete ET). The recurving TCs mainly
occur during May–December, with a peak in activity
during August–October. September has the highest in-
cidence of TC recurvature of any month, with 82 cases
identified in 31 years, a frequency of 2.6 yr
21
.
Composite analyses of the recurving WNP TCs are
constructed in a recurvature-relative framework. That is,
fields for each case are shifted such that the recurvature
point is collocated with the mean recurvature point at the
time of recurvature (T 1 0 h). As illustrated by a compar-
ison of conventional geography-relative and recurvature-
relative TC tracks (Figs. 1a,b), the recurvature-relative
framework reduces composite smearing that would arise
from the spatial variability of the recurving TC tracks
(Fig. 1a). The track variability partially reflects the in-
fluence of the time of year: Between May and August, TC
recurvature shifts poleward as the North Pacific jet stream
weakens and shifts poleward, and thereafter shifts equa-
torward through December as the jet stream strengthens
and shifts equatorward (e.g., Archambault et al. 2013,
section 3e).
The 6-hourly 2.58 NCEP–NCAR reanalysis fields
(Kalnay et al. 1996; Kistler et al. 2001)areusedtocon-
struct the composite analyses. Although higher-resolution
reanalyses such as the 0.58 Climate Forecast System Re-
analysis (Saha et al. 2010) are available, the NCEP–NCAR
reanalysis is used for consistency with the Archambault
et al. (2013) climatology. As discussed in Archambault
et al. (2013, their section 2d), the 2.58 NCEP–NCAR re-
analysis is able to capture synoptic signatures of TC–
ext ratropical flow interactions.
The statistical significance of the composite fields with
respect to climatology is assessed using a two-sided
Student’s t test (e.g., Wilks 2006, see section 5.2.1).
The climatology is produced from 1979–2009 21-day
running means of recurvature-relative 2.58 NCEP–
NCAR reanalysis fields.
b. Interaction-relative compositing methodology
Composite analysis is used to compare the extratropical
flow response to strong and weak TC–extratropical
FIG. 1. Recurving WNP TC tracks (2000–09 only; N 5 101) for T 2
48 to T 1 72 h plotted in (a) geography-relative and (b) recurvature-
relative frameworks. The thick black curve denotes the composite
track of all recurving TCs (N 5 292) for T 2 48 to T 1 72 h. The TC
symbol denotes the composite recurvature point (24.98N, 134.08E) for
all recurving TCs.
1124 MONTHLY WEATHER REVIEW VOLUME 143
flow interactions during WNP TC recurvature. For each
recurving TC, the strength of the interaction (i.e., the
interaction metric) is defined as the magnitude of the
spatially and temporally averaged 250–150-hPa layer-
averaged PV advection by the irrotational wind associ-
ated with the TC computed from 6-hourly 2.58 NCEP–
NCAR reanalysis (Archambault et al. 2013, their section
2d). The spatial (1583158) and temporal (48 h) averaging
is centered on the point and time (i.e., T 1 0 h), re-
spectively, of maximum interaction, which is defined as
the highest instantaneous magnitude of negative PV ad-
vection by the irrotational wind associated with the TC.
As noted by Archambault et al. (2013, their Fig. 5a) for
the case of WNP TC Oscar (1995) and by Cordeira (2011,
p. 111) for the case of WNP TC Dale (1996), the region
of strongest negative PV advection by the irrotational
wind generally tends to be closely aligned with the up-
stream edge of the TC cirrus shield. A point of maximum
interaction is not identified for 20 of the 292 recurving TC
cases because the negative PV advection by the irrotational
wind associated with the TC never exceeds an arbitrary
threshold of 1 PVU day
21
(1 PVU 5 10
6
Kkg
21
m
2
s
21
).
These cases are considered no-interaction cases. For the
remaining 272 recurving TC cases, those associated with an
interaction metric in the top quintile are categorized as
strong interactions (N 5 54), whereas those associated with
an interaction in the bottom quintile are categorized as
weak interactions (N 5 54).
As discussed in Archambault et al. (2013, their section
2d), the negative PV advection by the irrotational wind is
useful for diagnosing the strength of the TC–extratropical
flow interaction because divergent outflow impinging
upon the PV waveguide/jet stream is a key signature of
Rossby wave amplification induced by a TC (e.g., Harr
et al. 2008; Riemer et al. 2008; Hodyss and Hendricks
2010; Pantillon et al. 2013, 2014; Grams et al. 2013a,b).
Ridge amplification and jet streak intens ification
occur in conjunction with Rossby wave amplification
as the divergent outflow of the TC deforms PV con-
tours poleward and strengthens the PV gradient
(Archambault et al. 2013, their Fig. 4). T he effect s of
diabatic heating are included in this framework, as
upper-level low-PV air that arises from diabatic
heating is advected by diabatically driven divergent
outflow toward the waveguide/jet stream.
To examine strong and weak TC–extratropical flow
interactions, composites are constructed such that they
are centered on the point of maximum TC–extratropical
flow interaction. This ‘‘interaction-r elative’’ framework
is used to maximize the sharpness of the composite
synoptic features associated with the TC–extratropical
flow interactions. To construct these composites, the 2.58
NCEP–NCAR fields for a given case are shifted such
that the point of maxi mum TC–e xtratro pical flow in-
teraction is collocated with the mean point of maxi-
mum interaction (Fig. 2). The statistical significance of
the composite fields with respect to climatology is
evaluated using a two-sided Student’s t test. The cli-
matology is produced from 1979–2009 21-day running
means of interacti ve-r elative 2.58 NCEP–NCAR re-
analysis fields.
For all WNP TCs undergoing recurvature, the point of
maximum interaction tends to be located approximately
68–148N of the TC center (Fig. 2). However, for strong
interaction cases the location of the mean point of
maximum interaction is significantly farther northward
(11.68 vs 8.18 latitude) and westward (21.18 vs 1.58 lon-
gitude) relative to the TC compared with weak in-
teraction cases (Fig. 2; Table 1).
The synoptic signatures and dynamics of strong and
weak TC–extratropical flow interactions are compared
using a variety of plan-view composite analyses pro-
duced at the time of maximum interaction (i.e., T 1 0 h).
The irrotational and nondivergent wind fields are com-
puted for each individual case prior to compositing. All
derived variables, including PV advection, Q vectors,
and Q-vector divergence (described subsequently in
section 2d), and PV frontogenesis
1
[i.e., the rate of
change of the magnitude of the horizontal PV gradi-
ent; e. g., Davies and Rossa (1998), Cordeira (2011,
p. 105)], are computed on isobaric l evels from the
composite fields.
c. Assessment of characteristic magnitudes of various
quantities for composite cases
Although useful for identifying common synoptic fea-
tures, composite analyses have limited utility in de-
termining the characteristic magnitudes of these features
(e.g., Archambault et al. 2008). To compensate for this
limitation, the maximum values of various quantities for
the individual cases that constitute the composites are
analyzed for strong and weak TC–extratropical flow in-
teractions. For each case, the maximum magnitude of
a given quantity in the NCEP–NCAR reanalysis is ob-
tained by searching within a 208 latitude 3 208 longitude
box centered on the point of maximum interaction.
Whether the mean of the maximum magnitude of a given
quantity is significantly different for strong versus weak
interactions is tested using a nonparametric two-sided
Wilcoxon–Mann–Whitney rank-sum test (e.g., Wilks
2006, section 5.3.1).
1
The PV frontogenesis is calculated by replacing potential
temperature with PV in the simplified 2D form of the Miller (1948)
frontogenesis equation [e.g., Novak et al. 2004, their Eq. (1)].
A
PRIL 2015 A R C H A M B A U L T E T A L . 1125
