Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
1
Vertical Structure and Energetics of the Western Pacific Teleconnection Pattern 1
2
Sho Tanaka, Kazuaki Nishii, Hisashi Nakamura 3
4
Research Center for Advanced Science and Technology, the University of Tokyo, 5
Tokyo, Japan 6
7
Corresponding author: Kazuaki Nishii ([email protected]) 8
Research Center for Advanced Science and Technology, The University of Tokyo, 9
4-6-1 Komaba, Meguro-ku, Tokyo, 153-8904, Japan 10
11
Submitted to Journal of Climate in August 2015 12
13
14
15
16
17
18
19
2
Abstract 20
The Western Pacific (WP) pattern characterized by north-south dipolar anomalies in 21
pressure over the Far East and western North Pacific is known as one of the dominant 22
teleconnection patterns in the wintertime Northern Hemisphere. Composite analysis is 23
performed to investigate the three-dimensional structure of the WP pattern and assess 24
the energetics for the maintenance. The composited monthly height anomalies of the 25
WP pattern are found to exhibit baroclinic structure with their phase lines tilting 26
southwestward with height in the lower troposphere. Those anomalies can thus yield not 27
only a poleward heat flux across the climatological thermal gradient associated with the 28
strong Pacific jet but also a westward heat flux across the climatological thermal 29
gradient between the warmer North Pacific and the cooler Asian Continent. The 30
resultant baroclinic conversion of available potential energy from the 31
climatological-mean flow contributes most efficiently to the maintenance of the WP 32
pattern, acting against frictional and thermal damping, including the effects of 33
anomalous heat exchanges with the underlying ocean and anomalous precipitation in the 34
subtropics. The barotropic feedback forcing associated with modulated activity of 35
transient eddies also contributes positively to the maintenance of the WP pattern but it is 36
largely offset by the effect of anomalous transient heat flux and the net feedback forcing 37
is not particularly efficient. The present study suggests that the WP pattern has a 38
characteristic of a dynamical mode that can maintain itself by efficient energy 39
conversion from the climatological-mean fields even without external forcing. 40
41
3
1. Introduction 42
The Western Pacific (WP) pattern is a tropospheric teleconnection pattern characterized 43
by a north-south dipole of geopotential height anomalies over the Far East and western 44
North Pacific (Wallace and Gutzler 1981; hereafter WG81). The WP pattern is known 45
as one of the teleconnection patterns that modulate the East Asian winter monsoon 46
(Takaya and Nakamura 2005a, 2005b, 2013). Influence of the WP pattern is not limited 47
to on the winter monsoon. Stormtrack activity over the North Pacific is modulated in 48
association with the WP pattern (Nakamura et al. 1987; Lau 1988). Pavan et al. (2000) 49
and Rivière (2010) showed that frequency of blocking occurrence over the North 50
Pacific increases in months when the positive phase of the WP pattern is observed. 51
(Note that the positive phase of the WP pattern is defined when the northern height 52
anomaly is positive in this study.) Linkin and Nigam (2008) identified the WP pattern 53
with the North Pacific Oscillation (Walker and Bliss 1931; Rogers 1981; hereafter 54
NPO), calling it NPO/WP. They argued that its impact on surface air temperature 55
variability over the North America is comparable to that of Pacific North American 56
pattern (WG81; hereafter PNA) and El Niño/Southern Oscillation (ENSO; Walker and 57
Bliss 1931; Horel and Wallace 1981). They also mentioned that its influence on 58
variability of the Arctic Sea ice is stronger than that of other teleconnection patterns (In 59
section 2, we will briefly discuss the relationship between the WP pattern and NPO). 60
Meridional shift of the Aleutian Low in association with the WP pattern changes surface 61
wind stress curl over the North Pacific, which forces westward-propagating oceanic 62
baroclinic Rossby waves to move the northern boundary of the North Pacific 63
4
subtropical gyre and thereby alter sea-surface temperature (SST) in the 64
Kuroshio-Oyashio Extension (KOE; Ishii and Hanawa 2005; Sugimoto and Hanawa 65
2009). Moreover, Orsolini et al (2009) and Nishii et al. (2010) demonstrated that the 66
WP pattern acts to lower the polar stratospheric temperature over the Arctic and thereby 67
increase polar stratospheric clouds through suppression of upward planetary waves into 68
the stratosphere. 69
Linkage between the WP pattern and ENSO is known since Horel and Wallace 70
(1981). The linkage between the positive WP pattern and La Niña has been confirmed 71
as a third mode of singular value decomposition (SVD) between the global 72
mid-tropospheric geopotential height and SST in boreal winter (Koide and Kodera 73
1999). However, Kodera (1998) and Ose (2000) argued that the WP-ENSO linkage may 74
be affected by other processes, including snow cover over Siberia or precipitation over 75
the South China Sea. Influence of midlatitude SST variability on the WP pattern has 76
also been discussed. Hirose et al. (2009) showed that increasing transport in autumn of 77
the Tsushima warm current, which brings warm water from the Kuroshio to the Sea of 78
Japan through the East China Sea, tends to be linked with the positive phase of the WP 79
pattern in winter. By prescribing SST anomalies in the Sea of Japan as the boundary 80
condition for a regional atmospheric model, Yamamoto and Hirose (2011) obtained a 81
circulation response similar to the WP pattern. Frankignoul et al. (2011) found 82
circulation anomalies like the WP pattern in association with the decadal variability of 83
the KOE. Furthermore, a global circulation model (GCM) experiment with SST 84
anomalies over the North Pacific can simulate tropospheric height anomalies like the 85
5
WP pattern and a cooling in the polar stratosphere (Hurwitz et al. 2012). 86
Some previous studies attempted to understand the dynamics of the WP pattern from 87
a viewpoint of an internal dynamical mode of the midlatitude atmosphere that can be 88
excited even without external forcing like ENSO. Nakamura et al. (1987) argued that 89
kinetic energy (KE) conversion from the climatological-mean flow to the WP pattern 90
might be less efficient than that to the PNA pattern, in recognition of the fact that the 91
PNA pattern is located in the exit of the Pacific jet but the WP pattern is not. Lau (1988) 92
and Lau and Nath (1991) concluded that the WP pattern is close to the core of the 93
Pacific stormtrack and thus barotropic feedback forcing from synoptic-scale transient 94
eddies can be efficient for the maintenance of the WP pattern. 95
Most of the previous studies on low-frequency atmospheric variability, including 96
teleconnection patterns like the WP and PNA patterns, have focused on barotropic 97
processes. This is because circulation anomalies associated with low-frequency 98
variability appear to be equivalent barotropic, especially over the oceans (e.g., 99
Blackmon et al. 1979; Hsu and Wallace 1985). Black and Dole (1993) and Black (1997), 100
however, pointed out that height anomalies in association with the PNA pattern exhibit 101
a westward phase tilt with height and thus baroclinic processes may contribute to the 102
development of the PNA pattern. Linkin and Nigam (2008) found height anomalies 103
associated with the WP pattern also to be in baroclinic structure with a westward phase 104
tilt with height. In summer, the Pacific-Japan (PJ) pattern, a teleconnection pattern 105
characterized by meridionally aligned height anomalies over the Far East and western 106
North Pacific (Nitta 1987), exhibits baroclinic structure with westward and northward 107
6
phase tilts with height (Kosaka and Nakamura 2006, 2010). The PJ pattern thus 108
accompanies northward and eastward heat fluxes, acting to relax thermal contrasts 109
between the warmer Asian continent associated with the summer monsoon and the 110
climatologically cooler Siberia and North Pacific to the north and east, respectively. 111
These down-gradient heat fluxes act to reinforce the PJ pattern through conversion of 112
available potential energy (APE) from the climatological-mean flow. The baroclinic 113
structure of the PJ pattern in summer motivates us to investigate the vertical structure 114
and energetics of the wintertime WP pattern. As the zonal thermal contrast is reversed 115
from summer to winter, height anomalies associated with the WP pattern may exhibit a 116
southward phase tilt with height to yield a westward heat flux, in contrast to the PJ 117
pattern. We will show that is indeed the case for the WP pattern through composite 118
analysis for its high-amplitude events based on reanalysis data. 119
120
2. Data and analysis procedure 121
In the present study, monthly-mean data based on the European Centre for 122
Medium-Range Weather Forecasts (ECMWF) Reanalysis Interim (ERA-Interim; Dee et 123
al. 2011) are used. The horizontal resolution of the data is 0.75˚ x 0.75˚ in longitude and 124
latitude. We analyzed 32 winters (DJF) from 1979/80 through 2010/11. Diabatic 125
heating was estimated locally as the residual of the thermodynamic equation based on 126
6-hourly data. For the analysis for the earlier period, we use another reanalysis dataset 127
provided by the National Centers for Environmental Prediction (NCEP) and the 128
National Center for Atmospheric Research (NCAR) (NCEP–R1; Kalnay et al. 1996). 129
7
All the results shown in this paper are based on the ERA-Interim unless otherwise 130
stated. 131
Sub-weekly fluctuations associated with synoptic-scale transient eddies have been 132
extracted through digital high-pass filtering with a half-power cutoff period of 8 days, 133
following Kushnir and Wallace (1989) and Nakamura et al. (2002). In the following, the 134
fluctuations thus obtained are denoted by double primes. Momentum and heat fluxes 135
associated with the transient eddies (e.g., zonal momentum flux (u’’v’’) and meridional 136
heat flux (v’’T’’)) have been 8-day low-pass filtered and then averaged monthly at each 137
grid point before being used in the following composite analysis. We also used 138
precipitation data provided by Global Precipitation Climatology Project (GPCP; Adler 139
et al. 2003), whose resolution is 2.5˚ X 2.5˚. HadISST (Rayner et al. 2003) is used for 140
sea surface temperature (SST) and its resolution is 2.5˚x2.5˚. 141
The following WP pattern index (hereafter WP index) defined by WG81 is used in 142
this study: 143
𝑊𝑃 𝑖𝑛𝑑𝑒𝑥 = 12 𝑍!"" 60𝑁, 155.25𝐸 − 𝑍!"" 30𝑁, 155.25𝐸 (1)
In (1), Z500 represents a monthly-mean anomaly of 500-hPa geopotential height 144
normalized by its standard deviation at a given location for each calendar month. The 145
anomaly of a given variable for a given month is defined as a local deviation from its 146
32-year climatological mean for the particular calendar month. The index is suitable for 147
extracting a signature of dipolar height anomalies over the western North Pacific. In our 148
definition, the positive phase of the WP pattern corresponds to cyclonic and 149
8
anticyclonic anomalies at the southern and northern centers of action, respectively, and 150
vice versa for the negative phase. Composite maps of various anomalous fields are 151
constructed separately for the positive and negative phases of the WP pattern. For our 152
composites for strong positive events of the WP pattern, 18 months are selected for 153
which the WP index is positive and exceeds a unit standard deviation in strength (Table 154
1). Likewise, 14 months are selected for strong negative events of the pattern 155
Linkin and Nigam (2008) argued that the WP pattern is identical to the NPO pattern, 156
posing a question on the appropriateness of the usage of the particular WP index in our 157
analysis. For verification, we construct “teleconnectivity maps” for monthly anomalies 158
of 500-hPa height for boreal winter (DJF), following WG81 but separately for the 159
periods of 1962/63–1976/77 (Fig. 1a) and 1948/49–2010/11 (Fig. 1b) based on the 160
NCEP-R1 data. What is plotted at a given location in each of the maps is 161
“teleconnectivity”, defined as the absolute value of the strongest negative correlation 162
obtained in height anomalies between the particular location and any of the other 163
locations for the 45 months during a given 15-year period. Locations of strong 164
teleconnectivity thus correspond to centers of action of pressure seesaws. Constructed 165
for the same period as in WG81, the teleconnectivity map shown in Fig. 1a is very 166
similar to their Fig. 7b, in which the two centers of action of the WP pattern are 167
unambiguously depicted as local maxima of teleconnectivity (as indicated in Fig. 1a 168
with green dots). Two other local maxima in Fig. 1a correspond to centers of action of 169
the PNA pattern, one over the central North Pacific and the other over western North 170
America. Figure 1a suggests that in the particular period height variability associated 171
9
with the WP pattern was as prominent as that with the PNA pattern. In the 172
corresponding map for 1948/49–2010/11 (Fig. 1b), by contrast, the teleconnectivity 173
maxima that correspond to the WP pattern cannot be identified. We speculate that the 174
missing teleconnectivity maxima associated with the WP pattern in Fig. 1b may be due 175
to the dominant variability associated with the PNA pattern over that with the WP 176
pattern in the statistics for the 62-year period. Then, the centers of action of the WP 177
pattern would emerge in the teleconnectivity map if based on partial correlation from 178
which the variability of the PNA pattern is statistically removed. The variability is 179
defined on the basis of the following PNA index: 180
𝑃𝑁𝐴 𝑖𝑛𝑑𝑒𝑥 = 12 𝑍!"" 20𝑁, 180𝐸 − 𝑍!"" 50𝑁, 190𝐸 (2)
Note that the two reference grid points in (2) are shifted slightly from their counterpart 181
in WG81, so as to correspond to the centers of action of the PNA pattern in the period 182
1948/49–2010/11 recognized as the teleconnectivity maxima in Fig. 1b. The 183
teleconnectivity map based on the partial correlation is shown in Fig. 1c, where the 184
signature of the PNA pattern over the North Pacific is no longer recognized. Instead, 185
four teleconnectivity maxima are identified within the North Pacific, and two of them 186
are located in the immediate vicinities of the centers of action of the WP pattern (red 187
crosses). We confirmed that strong negative correlation is observed between the height 188
anomalies at these locations. The two other teleconnectivity maxima around 160°W 189
(blue crosses in Fig. 1c) may correspond to centers of action of another teleconnection 190
pattern. This pattern, if exists, is located so close to the WP pattern that the two patterns 191
10
cannot be spatially orthogonal. It is therefore unlikely that they are extracted in separate 192
modes of variability through an empirical orthogonal function (EOF) analysis, as used 193
by Linkin and Nigam (2008) to identify the NPO/WP pattern. Recognizing the clear 194
signature of the WP pattern in Fig. 1c, we use the WP index based on its definition as in 195
(1) throughout this study. 196
The other teleconnection pattern suggested in Fig. 1c may be similar to the NPO 197
pattern, but detailed comparison between the NPO and WP pattern is beyond the scope 198
of this study. The comparable dominance of the WP pattern to the PNA pattern for the 199
period of 1962/63–1976/77 (Fig. 1a) may suggest possible long-term modulations of the 200
WP pattern, which is also beyond the scope of this study. 201
202
3. Three dimensional structure of the WP pattern 203
As shown in Fig. 2a, 500-hPa height anomalies composited for the 18 strongest monthly 204
events of the positive WP pattern are characterized by a north-south dipole, in good 205
agreement with WG81. The northern anticyclonic anomaly is over eastern Siberia, the 206
Kamchatka Peninsula and Sea of Okhotsk, while the southern cyclonic anomaly is 207
elongated zonally from the Far East to as far as Hawaii Islands. The center of the latter 208
anomaly does not coincide with the southern reference grid point for the WP index 209
(southern green point in Fig. 2a), but it shifts slightly northwestward. This may be 210
because the intra-seasonal and inter-annual variance of 500-hPa height in winter is 211
larger over the eastern North Pacific than over the central and western North Pacific 212
(Blackmon et al. 1984). 213
11
A zonal-height section along 30˚N (Fig. 2c), the latitude very close to that of the 214
center of the southern anomaly of the WP pattern, shows that the phase of the 215
composited height anomaly tilts westward with height. Furthermore, a meridional 216
cross-section along 155˚E (Fig. 2e), the longitude of the two reference grid points for 217
the WP index, shows a southward tilt of the height anomalies with height in the lower 218
and mid-troposphere. This baroclinic structure of the WP pattern seems inconsistent 219
with the previous studies that show equivalent barotropic structure of low-frequency 220
variability over the oceans (e.g., Blackmon et al. 1979; Hsu and Wallace 1985). The 221
apparent barotropic structure of low-frequency anomalies may be due to coarse 222
horizontal and vertical resolutions of observational data at that time. In fact, recent 223
studies pointed out baroclinic structure of the WP and PNA patterns exhibit a westward 224
phase tilt with height (e.g., Black and Dole 1993; Black 1997; Linkin and Nigam 2008), 225
although the southward phase tilt has not been pointed out. This baroclinic structure 226
implies baroclinic energy conversion from the climatological-mean flow to the WP 227
pattern, which will be quantified in section 4. 228
As the composite anomalies of the negative WP pattern (Figs. 2b, d, f) look almost 229
like a mirror image of their positive counterpart (Figs. 2a, c, e), we focus only on its 230
positive phase in the following. In association with the positive WP pattern, a cold 231
anomaly with the enhanced northwesterlies is observed near the surface over the East 232
China Sea and south of Japan (Fig. 3a). This is consistent with Takaya and Nakamura 233
(2005a, b), who pointed out that a blocking high whose height anomaly has a strong 234
projection onto the WP pattern tends to amplify the surface Siberian High and thus 235
12
gives rise to a cold surge over East Asia. Cold anomalies with the anomalous westerlies 236
are also observed in the lower and middle troposphere over the East China Sea and 237
south of Japan (Figs. 3b-c). Around the Sea of Okhotsk, by contrast, a warm anomaly is 238
observed with anomalous easterlies in the lower and middle troposphere (Fig. 3b-c). 239
These wind and temperature anomalies are consistent with the baroclinic structure of 240
the WP pattern with a southward phase tilt (Fig. 2e) and equivalent to a westward heat 241
flux (Fig. 3c). Acting on the climatological zonal thermal gradient between the warmer 242
Pacific and cooler Asian continent, this westward heat flux is down gradient and thus 243
implies baroclinic energy conversion from the climatological flow to the WP pattern. 244
In the upper troposphere (Fig. 3d), the positive WP pattern accompanies anomalous 245
easterlies and westerlies to the north and south, respectively, which suggests southward 246
shift of the subtropical jet over the western North Pacific. The easterly anomaly is 247
consistent with the suppression of storm-track activity, as indicated by anomalous 248
lower-tropospheric poleward heat flux associated with sub-weekly disturbances (Fig. 249
3e) and by the upper-tropospheric variance of sub-weekly fluctuations in meridional 250
wind in the (Fig. 3f). The weakening of the storm-track activity is consistent with the 251
relaxed tropospheric meridional temperature gradient in association with the positive 252
WP pattern (not shown). 253
Diabatic heating anomalies associated with the WP pattern are observed mainly in the 254
lower and middle troposphere. In the lower troposphere, a positive anomaly of diabatic 255
heating is observed over the midlatitude/subtropical North Pacific, while a negative 256
anomaly is over the Sea of Okhotsk and Bering Sea (Fig. 4a). The former and latter 257
13
anomalies seem consistent with the local enhancement and reduction, respectively, of 258
sensible heat flux (SHF) from the ocean to the atmosphere (Fig. 4c). By analyzing two 259
other reanalysis datasets of NCEP-R1 and JRA-25 (Onogi et al. 2007) that provide 260
diabatic heating data, we confirm that those diabatic heating anomalies are mainly due 261
to anomalous vertical diffusion (not shown), which is consistent with our speculation. 262
Induced by strengthening and weakening of cold air outflow from the continent (Figs. 263
3a-b), the enhancement and reduction of upward SHF associated with the WP pattern 264
act to cool and warm the ocean surface over the respective domains and thereby induce 265
the SST anomalies as observed (Fig. 4b). In fact, those SST anomalies are hardly 266
observed in the preceding months of the positive monthly events selected for the 267
compositing (not shown). A recent numerical study suggested that SST anomalies 268
similar to Fig. 4b can generate anomalous atmospheric circulation like the positive WP 269
pattern (Hurwitz et al. 2012). Thus, there may be positive feedback between the WP 270
pattern and associated SST anomalies. In the mid-troposphere, a positive diabatic 271
heating anomaly observed over the eastern North Pacific (Fig. 4d) corresponds to 272
enhanced precipitation (Fig. 4e), which is mainly due to enhanced convective heating, 273
as represented in the NCEP-R1 and JRA-25 data. The enhanced convective 274
precipitation around 30°N is consistent with reduced static stability associated with the 275
cyclonic anomaly (Figs. 3b-c) . 276
277
4. Energetics of the WP pattern 278
To examine mechanisms for the maintenance of the monthly anomalies associated with 279
14
the WP pattern, we quantify each of the energy conversion/generation terms relevant to 280
the WP pattern based on its composited anomalies. In the composite maps, anomalies 281
that were observed in the individual events but unrelated to the WP pattern are 282
suppressed, and this procedure is necessary for our assessment of the energetics of the 283
WP pattern. 284
We estimated barotropic energy conversion (or KE conversion; CK) from the 285
climatological-mean state into the anomaly field associated with the WP pattern, 286
following Hoskins et al. (1983), Simmons et al. (1983), and Kosaka and Nakamura 287
(2006, 2010): 288
𝐶𝐾 = !!!!!!!
!!!!"− !!
!"− 𝑢!𝑣! !!
!"+ !!
!". (3) 289
Here, overbars and primes denote climatological mean and monthly anomalies, 290
respectively, and positive CK means that KE (defined as (u’2+v’2)/2 for the anomalies) 291
is converted from the climatological flow into anomalies. Figure 5 shows a map of CK 292
at the 250-hPa level, where horizontal wind shear is particularly strong climatologically 293
along the Pacific jet. As shown in Fig. 5a, positive CK maxima are observed to the 294
northern and southern flanks of the Pacific jet downstream of its core region. The first 295
term of CK (namely CKx in (3)) contributes positively to CK on the northern (southern) 296
portion of the jet exit region (Fig. 5b), where anomalous easterlies (westerlies) induce 297
anomalous advection of the westerly momentum that acts to reinforce themselves. The 298
second term of CK (CKy) also contributes positively in the southern portion of the jet 299
exit (Fig. 5c), where the anomalous winds have a slight northerly component and thus 300
15
yield anomalous westerly advection across the lateral shear of the jet that acts to 301
reinforce the anomalous winds. Positive CKy is also observed on the northern flank of 302
the jet core, across which the anomalous northeasterlies yield anomalous easterly 303
advection to reinforce themselves. 304
Baroclinic energy conversion (CP) through which available potential energy (APE) is 305
converted from the climatological-mean state to the monthly anomalies is estimated, as 306
in Kosaka and Nakamura (2006, 2010): 307
𝐶𝑃 = − !!!!
𝑢!𝑇! !!!"−𝑣!𝑇! !!
!". (4) 308
Here, Sp = (R/p) [(RT/pCp) – (dT/dp)] evaluated for the climatological-mean state, where 309
R and Cp denote the gas constant and the specific heat at constant pressure (p), 310
respectively. Figures 6a and 6b show maps of CP at the 500 and 850-hPa levels, 311
respectively, associated with the positive WP pattern. In Figs. 6a and 6b, positive CP 312
maxima are found to the south of the northern reference point and to the west of the 313
southern reference grid point of the WP pattern. The northern and southern CP maxima 314
are mainly contributed to by the first (CPx) and second (CPy) terms, respectively, of (4). 315
Around the northern CP maximum around the Sea of Okhotsk and Bering Sea, zonal 316
temperature gradient is climatologically strong associated with planetary waves (Fig. 317
6c) and/or thermal contrasts between the warmer Pacific Ocean and colder Eurasian 318
continent (Fig. 6d). Acting on this temperature gradient, the westward heat flux, which 319
reflects the southward-tilting height anomalies as discussed before (Figs. 2e and 3c), 320
yields large positive CP. Around the southern CP maximum over the 321
16
midlatitude/subtropical western Pacific, in contrast, APE is converted through poleward 322
heat flux acting on the climatologically strong temperature gradient (Fig. 6c) 323
accompanied by the Pacific jet (Fig. 3c). The poleward heat flux is consistent with 324
baroclinic structure of the WP pattern with the westward-tilting height anomalies (Fig. 325
2e). In the lower troposphere (Fig. 6b), those positive CP maxima are contributed to 326
also by relatively low static stability (i.e., small Sp), which arises climatologically from 327
the monsoonal outflow of cold continental air onto the warmer ocean. 328
Anomalous APE generation by anomalous diabatic heating (CQ) is defined as, 329
𝐶𝑄 = !!!!!!
𝑇!𝑄!, (5) 330
where Q on the RHS denotes diabatic heating. In the mid-troposphere (Fig. 7a), 331
negative CQ over the central/eastern North Pacific south of 40°N results from 332
anomalous diabatic warming due to increased precipitation (Figs. 4d-e) coinciding with 333
the cold anomaly in association with the WP pattern (Fig. 3c). In the lower troposphere 334
(Fig. 7b), warm and cold anomalies tend to be damped by negative and positive SHF 335
anomalies, respectively, from the ocean (Figs. 3b and 4c), yielding negative CQ (Fig. 336
7b). Those anomalies in air temperature and SHF arise from modulated cold surges in 337
association with the positive WP pattern, as mentioned in the preceding section. 338
Previous studies have pointed out the importance of feedback forcing from transient 339
eddies in the maintenance of the WP pattern (Nakamura et al. 1987; Lau 1988). The KE 340
and APE gains of the WP pattern due to modulated activity of high-frequency transient 341
eddies can be evaluated as follows; 342
17
𝐶𝐾!" = −𝑢! ! !!!!!!!
!"+ ! !!!!!! !
!"− 𝑣! ! !!!!!! !
!"+ ! !!!!!! !
!", (6) 343
𝐶𝑃!" = − !!!
!!!
! !!!!!!!
!"+ ! !!!!!!
!
!", (7) 344
In (6) and (7), double primes denote sub-weekly fluctuations associated with the 345
transient eddies, and their products implicitly indicate monthly statistics. Hereafter the 346
KE and APE gains defined in (6) and (7) are referred to as barotropic feedback (CKHF) 347
and baroclinic feedback (CPHF), respectively. Figure 8a shows CKHF at the tropopause 348
level, where not only the monthly wind anomalies but also sub-weekly wind 349
fluctuations tend to be larger than at any other levels. In agreement with previous 350
studies (Lau 1988; Lau and Nath 1991), upper-tropospheric CKHF contributes positively 351
to the maintenance of the monthly WP pattern by acting to increase KE over the central 352
and eastern North Pacific (Fig. 8a), most prominently in the downstream portion of the 353
storm-track core. Since synoptic-scale eddies migrating along the stormtrack act to 354
accelerate the westerlies through converging eddy momentum fluxes, the weakening of 355
the stormtrack activity associated with the positive WP pattern induces anomalous 356
easterly acceleration with anomalous divergence of eddy momentum flux where the 357
monthly anomalies are easterly (Fig. 8c). Likewise, CKHF is also contributed to by 358
anomalous convergence of a southerly momentum flux associated with transient eddies 359
that acts to reinforce the anomalous southerlies over the eastern North Pacific (Fig. 8e). 360
In contrast to CKHF, CPHF generally contributes negatively to the maintenance of the 361
WP pattern (Fig. 8b), as anomalous heat flux associated with transient eddies is 362
generally down-gradient and thus acting to relax anomalous temperature gradient 363
18
associated with the monthly WP pattern, as shown in the previous studies (Lau and 364
Nath 1991). The weakening of the stormtrack activity associated with the positive WP 365
pattern accompanies the reduced poleward eddy heat flux associated with transient 366
eddies (Fig. 3c), thus yielding anomalous divergence and convergence of the flux where 367
cold and warm anomalies are located to the south and north of the stormtrack axis, 368
respectively (Fig. 8d). In this manner, transient eddies act to damp the monthly-mean 369
thermal anomalies and thereby reduce APE associated with the WP pattern in acting to 370
render the monthly anomalies less baroclinic. 371
For more quantitative assessment of the energetics for the WP pattern, we integrated 372
each of the energy conversion/generation terms horizontally over the entire extratropical 373
Northern Hemisphere (20˚ ~ 90˚N, 0˚~360˚E) and vertically from the surface to the 374
100-hPa level. Dividing this value by the sum of KE and APE (i.e., total energy) of the 375
WP pattern that has been integrated over the same three-dimensional domain yields the 376
efficiency of each of the energy conversion/generation terms, which is shown in Table 2. 377
In the table, a negative value signifies that the particular term contributes negatively to 378
the maintenance of the WP pattern. Among the several conversion/generation terms 379
discussed above, the process that can contribute to the maintenance of the WP pattern 380
with the highest efficiency is baroclinic energy conversion (CP), through which the total 381
energy of the WP pattern could be replenished within five days. Barotropic energy 382
conversion (CK) can also contribute positively to the maintenance of the WP pattern, 383
but its efficiency is less than one third of that of CP. Through their evaluation of CK 384
and KE only at the 500-hPa level, Nakamura et al. (1987) argued that barotropic energy 385
19
conversion may be important for the maintenance of the WP pattern, but the present 386
study reveals the greater importance of the baroclinic processes. In agreement with Lau 387
(1988), barotropic feedback from transient eddies (CKHF) is also important for the 388
maintenance of the WP pattern, but its efficiency is only 40% of that of CP. 389
Furthermore, the net contribution from transient eddies is even less important; its 390
efficiency is only 10% of that of CP, because of the large offset by the negative 391
contribution through CPHF (Lau and Nath 1991). When combined, all the energy 392
conversion terms with the climatological-mean flow and transient eddies are highly 393
effective for the maintenance of the monthly anomalies of the WP pattern, as its total 394
energy could be replenished only within three or four days. This efficiency appears to 395
be sufficient for maintaining the WP pattern against thermal and frictional damping. As 396
indicated in Table 2, anomalous diabatic heating as a net acts as thermal damping, 397
through which the total energy can be consumed within 10 days. The efficiency of the 398
frictional damping cannot be evaluated from the dataset. The same evaluation of the 399
energetics has been repeated for the composited anomalies for the negative phase of the 400
WP pattern, and qualitatively the same result is obtained as shown in Table 2. 401
402
5. Summary and discussion 403
Through composite analysis, we have revealed a three-dimensional structure of the WP 404
pattern, which exerts significant influence on winter climate over the Far East and 405
western North Pacific. We have also elucidated the energetics of the composited 406
monthly anomalies associated with the WP pattern to clarify the mechanisms for its 407
20
maintenance. For the compositing, we selected large amplitude events of the WP pattern 408
based on the index defined by WG81. At first glance, the anomalies appear to be in 409
equivalent barotropic structure, but a close inspection reveals that it is in baroclinic 410
structure with a southwestward phase tilt with height from the surface to the 411
mid-troposphere. Although previous studies have suggested the baroclinic nature of the 412
PNA and WP patterns with a westward phase tilt with height (Black and Dole 1993; 413
Black 1997; Linkin and Nigam, 2008), the southward-tilting height anomalies of the 414
WP pattern have not been pointed out. 415
Energetics of the WP pattern has revealed that baroclinic APE conversion from the 416
climatological-mean flow to the monthly-mean anomalies through a horizontal heat flux 417
contributes the most to the maintenance of the WP pattern owing to its baroclinic 418
structure. Of particular importance is the westward component of the heat flux 419
associated with the southwestward-tilting height anomalies of the WP pattern. The 420
particular flux component, especially around the northern center of action, yields APE 421
conversion in acting on the eastward gradient of climatological-mean tropospheric 422
temperature over the Far East. The eastward temperature gradient reflects thermal 423
contrasts between the warmer Pacific Ocean and the colder Eurasian continent and the 424
planetary waves. The southwestward-tilting height anomalies of the WP pattern also 425
yield a poleward heat flux, which also contributes to the APE conversion, especially 426
around the southern center of action, in crossing the southward temperature gradient 427
associated with the Pacific jet. As its southern center of action is close to the jet core 428
region, the WP pattern can induce barotropic KE conversion from the climatological 429
21
Pacific jet. Its efficiency is, however, found much lower than that of the baroclinic 430
conversion. In addition, feedback forcing due to modulated activity of the Pacific 431
stormtrack also contributes positively to the maintenance of the WP pattern. This 432
process is, however, of secondary importance because of a large offset between the 433
barotropic feedback through eddy momentum fluxes that acts to increase KE and the 434
baroclinic feedback through eddy heat fluxes that acts to reduce APE. These processes, 435
most importantly the baroclinic APE conversion from the climatological-mean flow, are 436
so efficient for the maintenance of the monthly anomalies of the WP pattern against 437
frictional and thermal damping that the total energy of the pattern could be replenished 438
within four days. In fact, anomalous diabatic heating acts as an efficient damping 439
process of the WP pattern, through enhanced precipitation within the cold cyclonic 440
anomaly in the subtropical Pacific and through anomalous SHF from the ocean due to 441
modulated monsoonal outflow from the Eurasian continent. 442
Our finding of the efficient APE conversion from the wintertime climatological-mean 443
flow suggests that the WP pattern has a characteristic of a dynamical mode. The WP 444
pattern thus owes its existence to the particular wintertime climatological-mean 445
conditions over the Far East and the western North Pacific characterized by the strong 446
Pacific jet and zonal thermal contrasts between the warmer Pacific and the cooler Asian 447
continent. Recently, Kosaka and Nakamura (2006, 2010) have found that the 448
summertime PJ pattern also has a characteristic of a dynamical mode. Both the WP and 449
PJ patterns are characterized by meridional dipoles of zonally elongated height 450
anomalies over the western North Pacific and, these anomalies are tilting westward with 451
22
height to allow efficient APE conversion from the westerly Pacific jet. Interestingly, the 452
height anomalies associated with the PJ pattern are also tilting northward with height, 453
which yields eastward heat flux from the warmer Asian Continent into the cooler North 454
Pacific in summer. The opposing meridional tilting of height anomalies between the 455
WP and PJ patterns is consistent with the seasonal reversal of the thermal gradient 456
between the Asian Continent and North Pacific, so as to yield down-gradient heat fluxes 457
for APE conversion for the maintenance of those patterns. An obvious distinction 458
between the two patterns can be found in the role of anomalous diabatic heating due to 459
anomalous precipitation, which acts as thermal damping for the WP pattern but not for 460
the PJ pattern. Rather, anomalous convective heating acts to generate APE, and 461
anomalous surface winds act to enhance anomalous evaporation and anomalous 462
moisture transport for sustaining anomalous convection (Kosaka and Nakamura 2006, 463
2010). We therefore argue that the wintertime WP pattern is a dry dynamical mode 464
whereas the summertime PJ is a moist dynamical mode. 465
Previous studies have pointed out a linkage between the WP pattern and ENSO (e.g., 466
Horel and Wallace 1981; Kodera 1998). As a brief confirmation, we classified 186 467
winter months (DJF) for the period of 1949/50 – 2010/11 according to the phases of the 468
WP pattern (i.e., positive, negative, and neutral) and ENSO (El Niño, La Niña, and 469
neutral). In any of the months of the positive (negative) phase of the WP pattern chosen 470
for analysis, WP index exceeds a unit standard deviation positively (negatively). The 471
other classification of the winter months is based on two definitions of ENSO events 472
used operationally by the Japan Meteorology Agency (JMA) and National Oceanic and 473
23
Atmospheric Administration / Climate Prediction Center (NOAA/CPC) (Table 3).474
According to the JMA definition, the occurrence of an El Niño (La Niña) event is 475
officially identified when 5-month moving averaged Niño 3 index is 0.5 ˚C (-0.5 ˚C) or 476
higher (lower) for at least 6 consecutive months. Likewise, the definition by 477
NOAA/CPC of an El Niño (La Niña) event is such that 3-month moving averaged Niño 478
3.4 index exceeds 0.5 (is below -0.5). As shown in Table 3, those two definitions yield 479
only slight differences, and therefore the discussion below is mainly on the basis of the 480
JMA definition. 481
Table 3 indicates that only a single monthly event of the positive WP pattern was 482
observed during El Niño, which is much fewer than during La Niña (13 months). The 483
negative WP events were more frequent during El Niño (10 months) than during La 484
Niña (6 months). In other words, El Niño tends to set conditions highly unfavorable for 485
the occurrence of the positive phase of the WP pattern. Likewise, La Niña tends to set 486
conditions unfavorable for the occurrence of the negative WP pattern. These results 487
seem consistent with previous studies (Horel and Wallace 1981; Trenberth et al. 1998), 488
but Table 3 indicates that the positive WP pattern was observed most frequently when 489
the Tropical Pacific SST is close to normal (i.e., neutral). In other words, the linkage 490
between the WP pattern and ENSO is not particularly strong, and the WP pattern can 491
develop through internal atmospheric dynamics as a dynamical mode even without 492
remote influence of ENSO. Table 3 nevertheless shows that ENSO is still influential in 493
which phase of the WP pattern is likely to be triggered. In addition, recent studies have 494
suggested that SST variability in the western North Pacific (Hurwitz eta al. 2012; 495
24
Frankignoul et al. 2011) and/or in the Sea of Japan (Hirose et al. 2009) may trigger the 496
WP pattern. Further study is necessary on the mechanism behind the linkage between 497
the WP pattern and SST variability both in Tropics and extratropics. 498
In this study, the processes for the maintenance of the WP pattern are discussed 499
through composite analysis of monthly anomalies. For our understanding of the 500
formation mechanisms and predictability of the WP pattern, we need to examine time 501
evolution of the WP pattern in detail based on daily data. We also need to clarify the 502
impacts of global warming on the WP pattern in recognition of its influence on winter 503
climate over the Far East and ozone variability in the polar stratosphere. The findings of 504
the present study may lead us to the speculation that the WP pattern may undergo some 505
modulations by the projected weakening of thermal contrasts between the Asian 506
Continent and the Pacific. Detailed analysis on any possible projected changes in the 507
structure and/or dynamics of the WP pattern will be presented in another paper. 508
509
Acknowledgments 510
This study is supported in part by the Japanese Ministry of Environment through the 511
Environment Research and Technology Development Funds A-1201 and 2-1503, by 512
Japanese Ministry of Education, Culture, Sports, Science and Technology (MEXT) 513
through a Grant-in-Aid for Scientific Research in Innovative Areas 2205, and by the 514
Japan Society for the Promotion of Science (JSPS) through a Grant-in-Aid for Scientific 515
Research (B) 25287120. K.N. is also supported by JSPS through a Grant-in-Aid for 516
Young Scientists (B) 25800261. The JRA-25 dataset is provided by the Japan 517
25
Meteorological Agency and the Central Research Institute of Electric Power Industry. 518
The Grid Analysis and Display System (GrADS) was used for drawing figures. NCEP 519
Reanalysis data is provided by the NOAA/OAR/ESRL PSD, Boulder, Colorado, USA, 520
from their Web site at http://www.esrl.noaa.gov/psd/ 521
522
References 523
Adler, R. F., and Coauthors, 2003: The version-2 Global Precipitation Climatology 524
Project (GPCP) monthly precipitation analysis (1979–present). J. Hydrometeor, 4, 525
1147–1167. 526
Black, R. X., and R. M. Dole, 1993: The dynamics of large-scale cyclogenesis over the 527
North Pacific Ocean. J. Atmos. Sci., 50, 421-442. 528
Black, R. X., 1997: Deducing anomalous wave source regions during the life cycles of 529
persistent flow anomalies. J. Atmos. Sci. 54. 895-907. 530
Blackmon, M. L., R. A. Madden, J. M. Wallace, and D. S. Gutzler, 1979: Geographical 531
variations in the vertical structure of geopotential height fluctuations. J. Atmos. Sci., 532
36, 2450–2466. 533
Blackmon, M. L., Y. -H. Lee, and J. M. Wallace, 1984: Horizontal structure of 500 mb 534
height fluctuations with long, intermediate and short time scales. J. Atmos. Sci. 41. 535
961-979. 536
Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: configuration and 537
performance of the data assimilation system. Q. J. R. Meteorol. Soc., 137, 553-597. 538
Frankignoul, C., and N. Sennéchael, Y.-O. Kwon, M. A. Alexander, 2011: Influence of 539
26
the meridional shifts of the Kuroshio and the Oyashio Extensions on the atmospheric 540
circulation. J. Climate.. 24, 762-777. 541
Hirose, N., K. Nishimura, and M. Yamamoto, 2009: Observational evidence of a warm 542
ocean current preceding a winter teleconnection pattern in the northwestern Pacific. 543
Geophys. Res. Lett., 36, L09705, doi:10.1029/2009GL037448. 544
Horel, J. D., and J. M. Wallace, 1981: Planetary-scale atmospheric phenomena 545
associated with the Southern Oscillation. Mon. Wea. Rev., 109, 813–829. 546
Hoskins, B. J., I. James, and G. H. White, 1983: The shape, propagation and mean-flow 547
interaction of large-scale weather systems. J. Atmos. Sci., 40, 1595-1612. 548
Hurwitz, M. M., P. A. Newman, and C. I. Garfinkel, 2012: On the influence of North 549
Pacific sea surface temperature on the Arctic winter climate. J. Geophys. Res., 117, 550
D19110, doi:10.1029/2012JD017819. 551
Hsu, H. –H., and J. M. Wallace, 1985: Vertical structure of wintertime teleconnection 552
patterns in Northern Hemisphere. J. Atmos. Sci., 42, 1693-1710. 553
Ishii, Y., and K. Hanawa, 2005: Large-scale variabilities of wintertime wind stress curl 554
field in the North Pacific and their relation to atmospheric teleconnection patterns. 555
Geophys. Res. Lett., 32, L10607, doi:10.1029/2004GL022330. 556
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-year reanalysis project. Bull. 557
Amer. Meteor. Soc., 77, 437-471. 558
Kodera, K. Consideration of the origin of the different midlatitude atmospheric 559
responses among El Nino events. J. Meteor. Soc. Japan, 76, 347-361. 560
Koide, H., and K. Kodera, 1999: A SVD analysis between the winter NH 500-hPa 561
27
height and surface temperature fields. J. Meteor. Soc. Japan, 77, 47-61. 562
Kosaka, Y., and H. Nakamura, 2006: Structure and dynamics of the summertime 563
Pacific-Japan (PJ) teleconnection pattern. Quart. J. Roy. Meteor. Soc., 132, 564
2009-2030. 565
Kosaka, Y., and H. Nakamura, 2010: Mechanisms of meridional teleconnection 566
observed between a summer monsoon system and a subtropical anticyclone. Part I: 567
The Pacific-Japan pattern, J. Climate, 23, 5085-5108. 568
Kushnir, K., and J. M. Wallace, 1989: Low-frequency variability in the Northern 569
Hemisphere winter: geographical distribution, structure and time-scale dependence. J. 570
Atmos. Sci., 46, 3122–3143. 571
Lau, N.-C., 1988: Variability of the observed midlatitude storm tracks in relation to 572
low-frequency changes in the circulation pattern. J. Atmos. Sci., 45, 2718-2743. 573
Lau, N-. C., and M. J. Nath, 1991: Variability of the baroclinic and barotropic transient 574
eddy forcing associated with monthly changes in the midlatitude storm tracks. J. 575
Atmos. Sci., 48, 2589-1613. 576
Linkin, M. E., and S. Nigam, 2008: The North Pacific Oscillation-West Pacific 577
teleconnection pattern: Mature-phase structure and winter impacts. J. Climate, 21, 578
1979–1997. 579
Nakamura, H., M. Tanaka, and J. M. Wallace 1987: Horizontal structure and energetics 580
of Northern Hemisphere wintertime teleconnection patterns. J. Atmos. Sci., 44, 581
3377-3391. 582
Nakamura, H., T. Izumi, and T. Sampe, 2002: Interannual and decadal modulations 583
28
recently observed in the Pacific storm track activity and East Asian winter monsoon. 584
J. Climate, 15, 1855–1874. 585
Nishii, K., H. Nakamura, and Y. J. Orsolini, 2010: Cooling of the wintertime Arctic 586
stratosphere induced by the Western Pacific teleconnection pattern. Geophys. Res. 587
Lett., 37, L13805, doi:10.1029/2010GL043551. 588
Nitta, T., 1987: Convective activities in the tropical western Pacific and their impact on 589
the Northern Hemisphere summer circulation. J. Meteor. Soc. Japan, 65, 373-390. 590
Onogi, K., and Coauthors, 2007: The JRA-25 reanalysis. J. Meteor. Soc. Japan, 85, 591
369-432. 592
Orsolini, Y. J., A. Y. Karpechko, and G. Nikulin, 2009: Variability of the Northern 593
Hemisphere polar stratospheric cloud potential: The role of North Pacific disturb- 594
ances. Quart. J. Roy. Meteor. Soc., 135, 1020-1029. 595
Ose, T., 2000: A biennially oscillating sea surface temperature and the Western Pacific 596
Pattern. J. Meteor. Soc. Japan, 78, 98-99. 597
Pavan, V., Tibaldi, S., and Branković, Č., 2000: Seasonal prediction of blocking 598
frequency: Results from winter ensemble experiments. Q.J.R. Meteorol. Soc., 126: 599
2125–2142. 600
Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, 601
E. C. Kent, and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, 602
and night marine air temperature since the late nineteenth century. J. Geophys. Res., 603
108, 4407, doi:10.1029/2002JD002670. 604
Rivière, G., 2010: Role of Rossby wave breaking in the west Pacific teleconnection. 605
29
Geophys. Res. Lett., 37, L11802, doi:10.1029/2010GL043309. 606
Rogers, J. C., 1981: The North Pacific Oscillation. J. Climatol., 1, 39–57. 607
Sugimoto, H., and K. Hanawa, 2009: Decadal and interdecadal variations of the 608
Aleutian low activity and their relation to upper oceanic variations over the North 609
Pacific. J. Meteor. Soc. Japan, 87, 601-614. 610
Takaya, K., and H. Nakamura, 2005a: Mechanisms of intraseasonal amplification of the 611
cold Siberian High. J. Atmos. Sci., 62, 4423-4440. 612
Takaya, K., and H. Nakamura, 2005b: Geographical dependence of upper-level 613
blocking formation associated with intraseasonal amplification of the Siberian High. 614
J. Atmos. Sci., 62, 4441-4449. 615
Takaya, K., and H. Nakamura, 2013: Interannual variability of the East Asian winter 616
monsoon and associated modulations of the planetary waves. J. Climate, 26, 617
9445-9461. 618
Trenberth, K. E., G. W. Branstator, D. Karoly, A. Kumar, N.-C. Lau, and C. 619
Ropelewski, 1998: Progress during TOGA in understanding and modeling global 620
teleconnections associated with tropical sea surface temperatures. J. Geophys. Res., 621
103, 14,291–14,324. 622
Walker, G. T., and E. W. Bliss, 1932: World weather V. Mem. Roy. Meteor. Soc., 4, 53–623
84. 624
Wallace, J. M., and D. S. Gutzler, 1981: Teleconnections in the geopotential height field 625
during the Northern Hemisphere winter. Mon. Wea. Rev. 109, 784-812. 626
Yamamoto, M., and Hirose, 2009: Possible modification of atmospheric circulation 627
30
over the northwestern Pacific induced by a small semi-enclosed ocean. Geophys. Res. 628
Lett., 38, L03804, doi:10.1029/2010GL046214. 629
630
31
Table 1. The 18 and 14 months selected for composites of the positive and negative 631
phases, respectively, of the WP pattern. 632
Positive Negative Jan. 1980 Dec. 1982 Dec. 1980 Jan. 1987 Feb. 1982 Dec. 1987 Jan. 1984 Jan. 1988 Feb. 1984 Jan. 1989 Feb. 1985 Feb. 1989 Feb. 1986 Jan. 1990 Dec. 1989 Feb. 1990 Jan. 1991 Jan. 1992 Feb. 1991 Feb. 1999 Feb. 1994 Feb. 2001 Feb. 1995 Feb. 2006 Dec. 1995 Dec. 2006 Jan. 1997 Jan. 2007 Feb. 2003 Jan. 2004 Dec. 2005 Dec. 2010
633
634
635
32
Table 2. Efficiency (day-1) of each of the conversion/generation terms (1) through (7) in 636
the text. Efficiency represents how fast a particular term alone could replenish the total 637
energy (KE + APE) based on composited monthly-mean anomalies associated with the 638
WP pattern. The total energy and respective conversion/generation terms are integrated 639
horizontally over the entire extratropical North Hemisphere (20˚ ~ 90˚N, 0˚ ~ 360˚E) 640
and vertically from the surface to the 100-hPa level. 641
Efficiency in POS. Efficiency in NEG. CK 0.062 0.048 CP 0.220 0.243 CQ -0.106 -0.113 CKHF 0.090 0.093 CPHF -0.065 -0.065
642
643
33
Table 3, Classification of 186 winter months (DJF) in the period 1949/50 - 2010/11 644
according to the phases of the WP pattern and ENSO events. Calculation of the WP 645
pattern index is based on the NCEP-R1 data. The numbers in parentheses are statistics 646
based on the ENSO definition by NOAA/CPC, while others are based on the definition 647
by JMA. 648
El Niño Neutral La Niña WP positive 1(2) 18(19) 13(11) WP neutral 31(44) 67(42) 32(44) WP negative 10(14) 8(6) 6(5)
649
650
651
34
Figure captions 652
Figure 1. Teleconnectivity (contoured for every 0.05) maps of monthly geopotential 653
height at the 500-hPa level based on the NCEP-R1 data for winters (DJF) of (a) 1962/63 654
– 1976/77, the same period as in WG81, and (b) 1948/49 – 2010/11. (c) Same as in (b), 655
but the teleconnectivity is based on partial correlation from which the variability 656
associated with the PNA pattern is excluded. Confidence levels of the teleconnectivity 657
(i.e., negative correlation) at the significance levels of 10–6, 10–8, 10–10 and 10–12, based 658
on the Student’s t-static are indicated by different colors. The centers of action of the 659
WP pattern are indicated by a pair of red crosses at (30˚N, 155˚E) and (60˚N, 155˚E). 660
The two blue crosses in (c) denote another pair of teleconnectivity maxima that are 661
mutually correlated. 662
663
Figure 2. Monthly height anomalies (contoured for every 20 m; dashed for negative) 664
composited for the 18 and 14 strongest months of (a, c, e) positive and (b, d, f) negative 665
phases, respectively, of the WP pattern. Yellow (blue) shading represents the anomalies 666
that are positively (negatively) significant at the 95% confidence level based on the 667
t-statistic. Green dots represent the reference grid points at (30˚N, 155˚E) and (60˚N, 668
155˚E) used for the definition of the WP index. (a, b) 500-hPa height anomalies over 669
the Asian-Pacific region (20˚N~90˚N, 65˚E~115˚W). (c, d) Zonal-height sections for 670
30˚N (the latitude close to the southern reference grid point for the WP index (WG81)). 671
(e, f) Meridional sections for 155˚E (the longitude for the two reference grid points for 672
the WP index). Zero lines are omitted. 673
35
674
Figure 3. As in Fig. 1a, but for (a) temperature anomalies at the lowest model level (σ = 675
0.995) (countered for every 0.5 K), (b c) temperature anomalies (countered for every 0.5 676
K) at the 850-hPa (b) and (c) 500-hPa levels, (d) 250-hPa zonal wind anomalies (2 m 677
s-1), (e) 850-hPa anomalous heat flux due to transient eddies (2 K m s-1), and (f) 678
variance of 250-hPa meridional wind fluctuations associated with transient eddies (10 679
m2 s-2). Zero lines are omitted. In (a), vectors represent the total (anomaly + 680
climatology) wind associated with the WP pattern at the lowest model level. In (b, c), 681
vectors represent the wind anomalies associated with the monthly WP pattern at the 682
corresponding levels. Brown line in (d-f) indicates (d) climatological jet axis, (e) 12 m 683
s-1, and (f) 150 m2 s-2 of corresponding climatological-mean fields. 684
685
Figure 4. As in Fig. 1a, but for (a) anomalous diabatic heating at the 850-hPa level 686
(contoured for every 0.5 K day-1), (b) SST anomalies (0.2 K), (c) anomalous upward 687
surface sensible heat flux (5 W m-2), (d) anomalous diabatic heating at the 500-hPa level 688
(0.5 K day-1), and (e) precipitation anomalies of GPCP (0.3 mm day-1). Zero lines are 689
omitted. 690
691
Figure 5. (a) Local barotropic KE conversion (CK) at the 250-hPa level (shading; 10-4 692
m2 s-3) associated with the positive WP pattern. Brown contours represent 693
climatological-mean zonal wind (contoured for 30, 40, 50 m s-1, …) in winter (DJF), 694
and arrows 250-hPa wind anomalies (m s-1) associated with the WP pattern. (b) As in 695
36
(a), but for the KE conversion related only to the diffluence/confluence of the 696
climatological-mean jet (CKx). (c) As in (a), but for the KE conversion related mainly 697
to the lateral shear of the climatological-mean jet (CKy). 698
699
Figure 6, (a, b) Local baroclinic APE conversion (CP) at the (a) 500-hPa and (b) 700
850-hPa levels (shading; 10-4 m2 s-3) associated with the positive WP pattern. Contours 701
represent the climatological-mean temperature (every 10 K) at the given level. (c, d) As 702
in (a, b), respectively, but for APE conversion related only to the zonal gradient of the 703
climatological-mean temperature (CPx). (e, f) As in (a, b), respectively, but for APE 704
conversion related only to the meridional gradient of the climatological-mean 705
temperature (CPy). 706
707
Figure 7. APE generation (shading; 10-4 m2 s-3) by anomalous diabatic heating (CQ) at 708
the (a) 500-hPa and (b) 850-hPa levels, associated with the positive WP pattern. 709
710
Figure 8. (a) Local barotropic KE generation as feedback forcing by anomalous activity 711
of transient eddies (CKHF) at the 250-hPa level (shading; 10-4 m2 s-3) associated with the 712
positive WP pattern. (b) As in (a), but for baroclinic APE generation as feedback 713
forcing by anomalous activity of transient eddies (CPHF) at the 850-hPa level. (c) 714
Anomalous flux of westerly momentum at the 250-hPa level associated with transient 715
eddies (arrows) and its convergence (shading; m s-2). Contours represent monthly-mean 716
250-hPa zonal wind anomalies (every 4 m s-1; dashed for anomalous easterlies; zero 717
37
lines are thickened) associated with the positive WP pattern. (d) Anomalous 718
temperature flux at the 850-hPa level associated with transient eddies (arrows) and its 719
convergence (shading; K s-1 Contours represent monthly-mean 850-hPa temperature 720
anomalies (every 1 K; dashed for negative; zero lines are thickened). (e) As in (c), but 721
for anomalous flux of southerly momentum. Contours are for monthly-mean 250-hPa 722
meridional wind anomalies (every 2 m s-1; dashed for anomalous northerlies; zero lines 723
are thickened). 724
725
726
727
728
729
38
730 Figure 1. Teleconnectivity (contoured for every 0.05) maps of monthly geopotential 731
height at the 500-hPa level based on the NCEP-R1 data for winters (DJF) of (a) 1962/63 732
– 1976/77, the same period as in WG81, and (b) 1948/49 – 2010/11. (c) Same as in (b), 733
but the teleconnectivity is based on partial correlation from which the variability 734
associated with the PNA pattern is excluded. Confidence levels of the teleconnectivity 735
(i.e., negative correlation) at the significance levels of 10–6, 10–8, 10–10 and 10–12, based 736
on the Student’s t-static are indicated by different colors. The centers of action of the 737
WP pattern are indicated by a pair of red crosses at (30˚N, 155˚E) and (60˚N, 155˚E). 738
The two blue crosses in (c) denote another pair of teleconnectivity maxima that are 739
mutually correlated. 740
741
39
742 Figure 2. Monthly height anomalies (contoured for every 20 m; dashed for negative) 743
composited for the 18 and 14 strongest months of (a, c, e) positive and (b, d, f) negative 744
phases, respectively, of the WP pattern. Yellow (blue) shading represents the anomalies 745
that are positively (negatively) significant at the 95% confidence level based on the 746
t-statistic. Green dots represent the reference grid points at (30˚N, 155˚E) and (60˚N, 747
155˚E) used for the definition of the WP index. (a, b) 500-hPa height anomalies over 748
the Asian-Pacific region (20˚N~90˚N, 65˚E~115˚W). (c, d) Zonal-height sections for 749
30˚N (the latitude close to the southern reference grid point for the WP index (WG81)). 750
(e, f) Meridional sections for 155˚E (the longitude for the two reference grid points for 751
the WP index). Zero lines are omitted. 752
753
40
754 Figure 3. As in Fig. 1a, but for (a) temperature anomalies at the lowest model level (σ = 755
0.995) (countered for every 0.5 K), (b c) temperature anomalies (countered for every 0.5 756
K) at the 850-hPa (b) and (c) 500-hPa levels, (d) 250-hPa zonal wind anomalies (2 m 757
s-1), (e) 850-hPa anomalous heat flux due to transient eddies (2 K m s-1), and (f) 758
variance of 250-hPa meridional wind fluctuations associated with transient eddies (10 759
m2 s-2). Zero lines are omitted. In (a), vectors represent the total (anomaly + 760
climatology) wind associated with the WP pattern at the lowest model level. In (b, c), 761
vectors represent the wind anomalies associated with the monthly WP pattern at the 762
corresponding levels. Brown line in (d-f) indicates (d) climatological jet axis, (e) 12 m 763
s-1, and (f) 150 m2 s-2 of corresponding climatological-mean fields. 764
41
765
Figure 4. As in Fig. 1a, but for (a) anomalous diabatic heating at the 850-hPa level 766
(contoured for every 0.5 K day-1), (b) SST anomalies (0.2 K), (c) anomalous upward 767
surface sensible heat flux (5 W m-2), (d) anomalous diabatic heating at the 500-hPa level 768
(0.5 K day-1), and (e) precipitation anomalies of GPCP (0.3 mm day-1). Zero lines are 769
omitted. 770
771
42
772
Figure 5. (a) Local barotropic KE conversion (CK) at the 250-hPa level (shading; 10-4 773
m2 s-3) associated with the positive WP pattern. Brown contours represent 774
climatological-mean zonal wind (contoured for 30, 40, 50 m s-1, …) in winter (DJF), 775
and arrows 250-hPa wind anomalies (m s-1) associated with the WP pattern. (b) As in 776
(a), but for the KE conversion related only to the diffluence/confluence of the 777
climatological-mean jet (CKx). (c) As in (a), but for the KE conversion related mainly 778
to the lateral shear of the climatological-mean jet (CKy). 779
780
43
781 Figure 6, (a, b) Local baroclinic APE conversion (CP) at the (a) 500-hPa and (b) 782
850-hPa levels (shading; 10-4 m2 s-3) associated with the positive WP pattern. Contours 783
represent the climatological-mean temperature (every 10 K) at the given level. (c, d) As 784
in (a, b), respectively, but for APE conversion related only to the zonal gradient of the 785
climatological-mean temperature (CPx). (e, f) As in (a, b), respectively, but for APE 786
conversion related only to the meridional gradient of the climatological-mean 787
temperature (CPy). 788
789
44
790 Figure 7. APE generation (shading; 10-4 m2 s-3) by anomalous diabatic heating (CQ) at 791 the (a) 500-hPa and (b) 850-hPa levels, associated with the positive WP pattern. 792
793
794
45
Figure 8. (a) Local barotropic KE generation as feedback forcing by anomalous activity 795 of transient eddies (CKHF) at the 250-hPa level (shading; 10-4 m2 s-3) associated with the 796 positive WP pattern. (b) As in (a), but for baroclinic APE generation as feedback 797 forcing by anomalous activity of transient eddies (CPHF) at the 850-hPa level. (c) 798 Anomalous flux of westerly momentum at the 250-hPa level associated with transient 799 eddies (arrows) and its convergence (shading; m s-2). Contours represent monthly-mean 800 250-hPa zonal wind anomalies (every 4 m s-1; dashed for anomalous easterlies; zero 801 lines are thickened) associated with the positive WP pattern. (d) Anomalous 802 temperature flux at the 850-hPa level associated with transient eddies (arrows) and its 803 convergence (shading; K s-1 Contours represent monthly-mean 850-hPa temperature 804 anomalies (every 1 K; dashed for negative; zero lines are thickened). (e) As in (c), but 805 for anomalous flux of southerly momentum. Contours are for monthly-mean 250-hPa 806 meridional wind anomalies (every 2 m s-1; dashed for anomalous northerlies; zero lines 807 are thickened). 808