Generated using version 3.1.2 of the official AMS LATEX template

Mechanisms of variability within the Pacific storm track:1

upstream seeding and jet core strength2

Sandra M. Penny ∗ David S. Battisti

University of Wasington, Department of Atmospheric Sciences

3

Gerard H. Roe

University of Washington, Department of Earth and Space Sciences

4

∗Corresponding author address: Sandra Penny, University of Washington, 408 ATG Building, Box 351640,

Seattle, Washington 98195-164.

e-mail: latex@ametsoc.org

1

ABSTRACT5

This paper examines how variations in two mechanisms, upstream seeding and jet core6

strength, relate to storminess within the cold season (Oct - Apr) Pacific storm track. It is7

found that about 17% of observed storminess co-varies with the strength of the upstream8

source, and the relationship is robust throughout the cold season (Oct - April) and for9

both the Pacific and Atlantic basins. This analysis also supports the conclusion that the10

midwinter suppression in storminess over the North Pacific Ocean is due to a notable lack11

of storminess upstream of the Pacific storm track in the heart of winter, consistent with12

previous work. Further analyses of the intraseasonal variability in the strength and structure13

of the wintertime Pacific jet stream draw upon both Eulerian-variance and feature-tracking14

statistics. We diagnose why winter months with a strong-core jet stream have a weaker15

Pacific storm track than those with a weak-core jet stream. Most notably, these results show16

that disturbances are significantly weaker and shorter-lived when the wintertime jet core is17

strongest. Connections with the lifecycles of individual storms and the role that transient18

eddies play in maintaining the jet stream are also discussed.19

1

1. Introduction20

Linear theories of baroclinic instability were first proposed shortly after the formulation21

of the quasigeostrophic equations (e.g., Charney 1947; Eady 1949), and they have provided22

remarkable insight into the controls on midlatitude synoptic storms. These linear scaling23

relationships have been invoked to make predictions about the large-scale controls on clima-24

tological storminess as a function of mean climate state, both future and past, with a great25

deal of success (e.g., Nakamura and Shimpo 2004; Li and Battisti 2008).26

However, no linear theory can be a complete description of storm-track dynamics, and27

wealth of studies have investigated other aspects of mid-latitude eddies (e.g., Thorncroft et al.28

1993). Observations of the Pacific storm track have also called into question the applicability29

of linear theory. In particular there are two aspects of its behavior, the climatological seasonal30

cycle of storminess and the intraseasonal variability of storminess in the winter season, that31

apparently contravene linear predictions. We describe each below.32

A midwinter minimum (MWM) is observed in the strength of the climatological Pacific33

storm track (Nakamura 1992). The minimum occurs despite the fact that the regional34

temperature gradients and jet stream winds are strongest in winter, which linear theory35

would predict leads to the largest growth rates. Many possible dynamical reasons for this36

behavior have been proposed and investigated (e.g., Christoph et al. 1997; Zhang and Held37

1999; Chang 2001; Nakamura et al. 2002; Nakamura and Sampe 2002; Yin 2002; Harnik38

and Chang 2004). In a recent study, Penny et al. (2010, hereafter PRB10), used a feature-39

tracking technique (Hodges 1994) to show that the MWM occurs primarily because there is a40

midwinter reduction in storminess over Asia, upstream of the Pacific storm track. As a result,41

fewer and smaller seed disturbances propagate into the Pacific storm track. PRB10 showed42

that this is, in fact, broadly consistent with linear theory: seed disturbances from Asia are43

reduced in midwinter consistent with an increase in low-level static stability associated with44

the Siberian High. However PRB10 also showed that while growth rates of storms do not45

have a minimum in midwinter, they do not have a maximum either, as linear theory would46

2

predict. Instead the growth rates stay relatively constant throughout the winter months,47

suggesting that there is indeed behavior that is not accounted for in a wholly linear picture.48

The idea that upstream effects may be important is not new, in fact it was proposed49

by Nakamura (1992) as a possible explanation for the MWM. The results of PRB10 were50

surprising for two reasons. First, nonlinearities of winter Pacific storm track have been well51

documented by others (e.g., Chang 2001; Nakamura et al. 2002; Nakamura and Sampe 2002)52

and were previously advocated as the cause of the MWM. The results of PBR10 show that it53

is not necessary to invoke such nonlinearities to understand why there is a MWM, and that54

they are not the primary contributor to the phenomenon. Second, others who have looked55

did not find a clear relationship between upstream seeding and storminess in the midwinter56

suppression season (Zhang 1997; Chang and Guo 2011).57

Explorations of the relationship between upstream seeding and storm-track variability58

are sparse, however a few studies do exist. Zurita-Gotor and Chang (2005) employed both59

theory and simple modeling experiments and found that Earth’s atmosphere exists in a60

regime where both local effects (e.g., baroclinicity) and non-local effects (e.g., seeding, trop-61

ical teleconnections) contribute to the overall strength of a storm track. Orlanski (2005)62

performed 50-day integrations of a high-resolution numerical model for the Pacific storm63

track. He nudged the upstream boundary by random upper-level waves and showed a clear64

relationship between the storm track and upstream seeding. Donohoe and Battisti (2009)65

found evidence that a reduction in seeding contributes to North Atlantic storminess in GCM66

simulations of the Last Glacial Maximum. PRB10 was the first study that we are aware of to67

document the importance of seeding in observations on climatological timescales. One aim68

of this study is to further dissect the relationship between upstream seeding and downstream69

storminess in observations of Pacific and Atlantic storm tracks throughout the cold season.70

The second aspect of the winter Pacific storm track that has been argued to contravene71

linear behavior is its intraseasonal variability in the winter (DJF) season. Previous work72

(Chang 2001; Nakamura et al. 2002; Nakamura and Sampe 2002) has shown that there73

3

is an inverse relationship between winter storm track intensity and jet core strength in74

this region. That is, faster-than-normal Pacific jet streams are accompanied by weaker-75

than-normal storm track intensity. Numerous publications have evaluated how dynamical76

mechanisms that operate locally (i.e., within the storm track itself) may give rise to the77

observed intraseasonal and interannual variability. For example, studies have evaluated the78

extent to which variability in diabatic heating (e.g., Nakamura et al. 2002; Chang 2001) or a79

modification to the linear models that accounts for a narrow jet stream (Harnik and Chang80

2004) can account for the observations. Others studies have evaluated the importance of81

non-local forcings, such as winter monsoonal flow (Nakamura et al. 2002) and wave trapping82

by a strong subtropical jet stream (Nakamura and Sampe 2002). Taken together, these83

studies make a significant contribution to our current understanding of midlatitude storm84

track dynamics, but still much remains that is poorly understood.85

The inverse relationship between winter storm track intensity and jet core strength on86

intraseasonal timescales is surprising for the same reason that the MWM was surprising:87

linear theory predicts that midlatitude storminess should maximize when the jet stream,88

and its associated vertical shear, is strongest. Due to this similarity, it is often assumed that89

the physics of the climatological MWM are the same as those associated with intraseasonal90

variations in the wintertime Pacific storm track. However, as we will show, this assumption91

is not valid. While the climatological patterns of monthly average storminess and jet stream92

strength are similarly anticorrelated for the climatological MWM (fall/spring compared to93

winter) as well as on intraseasonal timescales (months with weak compared to strong jet94

streams in winter), the mechanisms giving rise to the two are different (see also Chang 2001;95

Nakamura et al. 2002, PRB10).96

This paper is divided into two parts. We first investigate the relationship between up-97

stream seeding and downstream storminess in the Pacific and Atlantic storm tracks on both98

climatological (the climatological seasonal cycle, from October through April) and intrasea-99

sonal (wintertime month-to-month variability, including December, January, and February)100

4

timescales. We will show that there is a general and robust correlation with seeding for the101

two Northern Hemisphere storm tracks that persists despite dramatic changes in baroclin-102

icity and bottom boundary condition. These results also expand upon the conclusions of103

PRB10 that the MWM occurs primarily because the upstream wave source is reduced in104

winter compared to fall and spring. Second, we use feature-tracking statistics to investigate105

why there is an inverse relationship between jet core strength and Pacific storminess on the106

month to month timescale in winter. We will show that during winter when the jet core is107

strongest, nonlinearities that are local to the Pacific storm track result in weak-amplitude,108

short-lived disturbances. For these intraseasonal variations in the wintertime Pacific storm109

track there is no statistically significant signal due to seeding.110

2. Methods111

In the climate literature, storm tracks are generally viewed as a geographic region of112

enhanced synoptic wave activity in the climatological sense. In this paper we adopt that113

definition and use two complementary perspectives - one based on eddy variance one and114

the other based on feature tracking - to analyze storminess within the Pacific storm track.115

We focus our attention primarily on upper-level (300hPa) geopotential height, but do also116

discuss results from other fields and levels where appropriate.117

For eddy variance statistics, we represent storminess as the variance of geopotential height118

that has been 2-10 day bandpass filtered to isolate wave activity on the time scale of synoptic119

storms, as is common in the climate literature (e.g., Blackmon 1976; Blackmon et al. 1977).120

All of the calculations we will show have also been performed for the meridional wind field121

and the results are essentially unchanged.122

For feature-tracking statistics, we use the algorithm documented by Hodges (1994, 1995,123

1999) to compile an inventory of all Northern Hemisphere disturbances in the 6-hourly124

European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40)125

5

dataset (Uppala et al. 2005) from 1958 to 2001. As in PRB10 and Hoskins and Hodges126

(2002), disturbances must travel at least 1000 km and last 48 hours to be included in our127

analysis, and the minimum threshold value (relative to the background field) for the exis-128

tence of a disturbance is 3 dm. The filtering scheme we implement prior to feature tracking is129

deliberately quite weak. While we obviously wish to remove any time-average features (e.g.,130

the Aleutian and Icelandic Lows) we also want to retain as much synoptic-scale wave activ-131

ity as possible. As in PRB10, we apply a 90-day high-pass Butterworth filter to remove the132

seasonal cycle, and apply a spatial filter that admits only planetary wavenumbers between 5133

and 42 to isolate synoptic-scale disturbances. Geopotential height is ideal for our purposes134

because: features have an easily identified center; central magnitude is a meaningful mea-135

sure of overall intensity (unlike relative vorticity); and distinguishing between cyclonic and136

anticyclonic disturbances is straightforward (unlike meridional wind). See PRB10 for more137

discussion of our filtering and tracking choices.138

The feature-tracking algorithm can be used to determine the density of storm tracks,139

which we call “number density”. This is calculated from the number of storms to pass140

within 100 km of a latitude and longitude gridpoint, and not the number of storms to pass141

within a gridbox. We make this choice, similar to what others have done (e.g., Hoskins and142

Hodges 2002), because a simple count by gridbox tends to overemphasize the importance of143

slow-moving disturbances and it is also subject to noisy results when few storms are present.144

Although we mostly present results for geopotential height at 300hPa, we have also145

performed similar calculations using relative vorticity and meridional wind, and at levels146

between 250 and 500 hPa. Results from these fields are very similar. In addition, results147

found using the whole record (1958 - 2001) also hold for the satellite era (1979 - 2001).148

Unless otherwise stated, all confidence intervals are determined from a student’s t-test, and149

95% confidence is required for statistical significance. We reserve the word “significant” to150

mean that a test for statistical significance has been performed.151

6

3. Upstream seeding and the Northern Hemisphere storm152

tracks153

a. Upstream seeding and the Pacific storm track154

We first explore the relationship between Pacific storm-track intensity and upstream155

seeding by comparing geopotential height variance at 300hPa over Asia (35◦-65◦N, 70◦-156

120◦E, hereafter called the “upstream domain”) to the same variance within the Pacific157

storm track (20◦-65◦N, 160◦E-160◦W, hereafter called the “downstream domain”) for all158

cold-season (defined in this paper as October through April) months. The upstream and159

downstream domains, together with a random sample of the disturbances that propagate160

through the upstream domain, are shown in Figure 1. The upstream domain’s bounds are161

chosen because previous work (e.g., Wallace et al. 1988; Hakim 2003; Chang 2005) has shown162

that mid-latitude Asia is the dominant source of seed disturbances for the Pacific storm track.163

The downstream domain’s bounds are chosen to cover a large portion of the Pacific storm164

track, and also to be several Rossby radii downstream of the upstream domain. Results are165

insensitive to modest changes (i.e. ∼ 10◦− 20◦ in any direction) to the location and sizes of166

these two domains.167

A scatter plot comparing 300hPa storminess in the upstream and downstream domains is168

shown in Figure 2a, where storminess is defined as the band-passed variance in geopotential169

height (see Section 2). Every symbol represents the monthly average storminess for each of170

the seven cold season months and for 43 years between October 1958 and April 2001. Very171

similar results are obtained for meridional wind variance and for feature tracking statistics172

at the same level (not shown).173

The first impression from Figure 2a is that there is a great deal of scatter. This em-174

phasizes that there is a large amount of seasonal and interannual variability in storminess,175

and that mean climate phenomena emerge from this variability. However it is also visually176

obvious that there is a positive correlation between the upstream and downstream domains.177

7

For all months together, the correlation coefficient of the data shown in Figure 2a is 0.41,178

which is highly statistically significant (> 99%, see Table 1), and implies that about 17%179

of the variability in Pacific storm track intensity is accompanied by variations in upstream180

storminess. While 17% may seem like a small fraction of the variance, it should be noted181

that this is comparable to other notable and meaningful fluctuations in storm track intensity.182

For example, the MWM is manifest as a 8% reduction in the intensity of the Pacific storm183

track in winter (DJF) relative to the transition (ON, MA) months (Table 2, also Nakamura184

(1992)) and comparing extreme strong and weak jet months in winter (DJF), we find a 25%185

reduction in storminess in the downstream domain (Table 3, column 1, also Chang 2001).186

Figure 2a also demonstrates the role of seeding and is consistent with the results of187

PRB10: midwinter months (filled black symbols) are clustered in the lower left of Figure188

2a (i.e., weak upstream seeding corresponds to weak storm track intensity), whereas the189

transition seasons (open gray symbols) are not. On average, in winter upstream seeding190

is 14% weaker, and the Pacific storm track is 8% weaker, than in the transition months191

(see also Table 2). The difference between both of these means is statistically significant192

with greater than 99% confidence. This degree of covariance is sufficient by itself to account193

for the observed seasonal cycle in Pacific storminess, and is consistent with the result from194

PRB10 that the MWM occurs primarily because there is a wintertime reduction in seeding.195

Like many climatic indices, however, this mean signal emerges after averaging over a large196

amount of scatter.197

The relationship between upstream and downstream storminess persists throughout the198

cold season. For all months the correlation is 0.41; for only the transition-season months199

(October, November, March, and April) it is 0.38; and for winter months (December, Jan-200

uary, and February), it is 0.39 (Table 1). These correlations are statistically indistinguishable201

from each other. In this paper we focus on seasonal data because we want to maximize the202

amount of available data, but we have also considered year-to-year changes in a set calendar203

month. With 95% confidence, six of the seven individual cold season months show correla-204

8

tions between upstream and downstream storminess that are statistically indistinguishible205

from the seven-month average correlation (0.41, see Table 1). The odd month out is April206

(correlation coefficient is 0.07), which is indistinguishable from 0.41 if we require 96% con-207

fidence. The weak relationship in April is consistent with results from PRB10, where we208

found a notable spike in growth rates over the Pacific storm track in late spring.209

It is important to note that the relationship in Figure 2 cannot simply be due to the direct210

effect of disturbances that propagate from the upstream domain to the downstream domain.211

Indirect processes, such as downstream development, must also be involved. This can be212

seen most easily from feature-tracking statistics, such as the sample of 100 disturbances that213

propagate through the Pacific upstream domain, shown in Figure 1. Overall, about 9.5% of214

the cold season cyclonic disturbances that cross through the center of the downstream domain215

(180◦E) also cross the center of the upstream domain (95◦E, this fraction is 7.3% for Dec-Jan,216

and 11% for Oct, Nov, Mar, and Apr, consistent with there being fewer seeds in winter.).217

This places an estimate on the importance of the upstream domain due to direct seeding of218

the Pacific storm track. For example, there is a 14% reduction in upstream variance during219

winter (DJF) relative to the transition months (ON, MA, see Table 2). Therefore if all of the220

disturbances in the upstream region disappeared (something that is physically unrealistic),221

then there would still be a 2.5% (i.e., 12%-9.5%) reduction in downstream variance in winter222

relative to the transition months. We emphasize that this is only an estimate. We have223

ignored the amplitude of seed disturbances (which are also reduced in winter relative to the224

transition seasons, see PRB10) and we have assumed that transient eddy variance is only225

due to trackable disturbances.226

b. Comparisons between the two Northern Hemisphere storm tracks227

Although the primary focus of this paper is the Pacific storm track, it is instructive to228

compare results with the Atlantic storm track. Figure 2b was obtained for the variations in229

the Atlantic storm track at 300hPa following the same methods that were used to obtain230

9

Figure 2a for the Pacific storm track, with the exception that the upstream (35◦-65◦N,231

130◦-180◦W) and downstream (20◦-65◦N, 40◦W-0◦) domains have been shifted 160◦ to the232

East. For all months the correlation over the Atlantic is 0.37; for only the transition-season233

months (October, November, March, and April) it is 0.39; and for winter months (December,234

January, and February), it is 0.30 (Table 1). Collectively these data indicate that, like235

the Pacific, the relationship between upstream seeding and downstream storminess persists236

throughout the cold season. Although it is not a difference that is statistically significant at237

95%, the slight drop in correlation for the winter months relative to the transition seasons238

is a suggestion that seeding may be a less important control for Atlantic storminess in the239

depths of winter, when local conditions are exceptionally favorable for storm growth.240

While the general characteristics of Figure 2 for the Pacific and Atlantic are strikingly241

similar, there are some important differences. For example, upstream seeding and down-242

stream storminess are both strongest during winter for the Atlantic: on average, in January243

and February upstream seeding is 1% stronger (an insignificant difference), and the Atlantic244

storm track is 20% stronger (a significant difference), than in November and April (Table245

2).246

Furthermore, the Atlantic storm track is almost always more strongly seeded than the247

Pacific. For example, average upstream seeding is 58% higher in the Atlantic than the Pacific248

for the cold season average, and the biggest differences occur in winter (for Dec-Feb, average249

seeding is 70% higher in the Atlantic than in the Pacific). This may have some bearing on the250

observation that the Atlantic storm track is characterized by both less overall baroclinicity251

and also a 10%-15% stronger storm track than the Pacific (e.g., Kageyama et al. 1999). Our252

results raise the possibility that there isn’t a local or a nonlinear cause for this observation.253

Instead, it may simply be due to the fact that the Atlantic storm track is more strongly254

seeded than the Pacific. More work is needed to evaluate this suggestion.255

When the Atlantic and Pacific are considered together, it is clear that the importance256

of upstream seeding is robust in both basins and for a wide range (Fall through Spring,257

10

Atlantic and Pacific) of background states. The data for both basins are next divided into258

four regimes based on the strength of the upstream seeding: weak (< 4000m2); medium259

(4− 6000m2); strong (6− 8000m2); very strong (> 8000m2). In Figure 2, the ellipses show260

the average of the data, partitioned into the different seeding regimes; a square in the center261

of an ellipse indicates the average of the weak, medium, strong, and very strong cases for262

winter (DJF) months, and a triangle surrounded by an ellipse indicates the average for the263

transition seasons (ONMA).264

This division again underscores that the Atlantic storm track is more strongly seeded265

than the Pacific. For example, in the Atlantic only one month falls into the “low” seeding266

regime, while 81 months fall into the “very high” regime. In the Pacific, 70 months fall267

into the “low” seeding regime, and only 7 months fall into the “very high” regime. For the268

ellipses in Figure 2, we consider the “low” and “medium” regimes together for the Atlantic,269

and we consider the “high” and “very high” seeding regimes together for the Pacific. This270

division is robust to other choices of low and high seeding regimes. For example results are271

the same if we divide the data so that an equal number of months fall into low, medium,272

and high seeding regimes.273

This division by the strength of the upstream seeding also provides insight into variations274

in Pacific and Atlantic storm track intensity that occur independent of variations in upstream275

seeding. In each of its three seeding regimes (i.e., medium, strong, very strong), the Atlantic276

storm track is significantly stronger in winter than it is during fall and spring. This is277

consistent with expectation from linear theory that a stronger jet leads to increased local278

storminess. However, in each of the Pacific storm track’s three seeding regimes (i.e., weak,279

medium, and strong), the mean storminess during winter months is not significantly different280

from the transition-season months despite the increase in baroclinicity in winter compared281

to the shoulder seasons.282

The above observation helps shed light on seeming discrepancies in the literature con-283

cerning interannual, interseasaonal, and intraseasonal variations of the Pacific storm track.284

11

In particular, previous studies (e.g., Chang 2001; Zhang 1997) found evidence that nonlinear285

processes inhibit the development of baroclinic waves within the wintertime Pacific storm286

track, and this led to a conclusion that the MWM is caused by these nonlinearities. PRB10287

also found that the relationship between observed storminess and growth rates does not con-288

form to linear expectations (PRB10, their Figure 7), but their results showed that nonlinear289

processes are the dominant cause of the MWM. Chang and Guo (2011) questioned whether290

upstream seeding is indeed important for the MWM, but Penny et al. (2011) showed that an291

apparent discrepancy only arose because Chang and Guo (2011) undersampled the available292

data. While the results presented here add further support to the interpretation that reduced293

upstream seeding is the dominant cause of the MWM, but they also show that nonlinear294

processes are present and contribute.295

4. The relationship between intraseasonal jet stream296

variability and the Pacific storm track in winter (DJF)297

To this point, we have shown that upstream seeding exerts a robust and important control298

on downstream storm track intensity. This relationship is maintained through tremendous299

changes in baroclinicity, boundary conditions, and atmospheric background state. In this300

section we begin by showing that, consistent with the results from others (e.g., Nakamura301

1992; Chang 2001), over the Pacific there is a clear and counterintuitive inverse relationship302

between wintertime jet stream variability and storm track intensity. After establishing that303

this link in the intraseasonal wintertime variability does not occur because of seeding, we304

examine both variance statistics and feature-tracking statistics to diagnose why it does occur.305

Over the Pacific there is a clear relationship between jet strength and storminess that306

is additional to, and separate from, the upstream seeding relationshiop. In Figure 3a the307

20 winter (DJF) months with the strongest Pacific jet core are indicated in red (hereafter,308

the “strong-core jet” months), the 20 winter months with the weakest Pacific jet core are309

12

indicated in blue (hereafter, the “weak-core jet” months), and all other winter months (89310

out of 129) are black. For these results and all subsequent discussions, we define a metric311

of jet core strength to be the wind speed at the core of the jet, calculated as maximum312

monthly averaged zonal wind speed within a 3D spatial grid that spans the entire North313

Pacific basin (20◦ − 70◦N, 12◦E − 220◦W , 500hPa− 100hPa) for the months of December,314

January, and February. We obtain almost identical results for other reasonable choices. For315

example, results are the same if we consider two-month averages, three-month averages, or316

one-month averages as shown here; if we consider only a spatial 2D grid near jet stream317

level instead of a 3D grid that spans many vertical layers; or if we define winter as January,318

January/February, or December/January/February.319

A clear connection between jet core strength and storm track intensity is visually evident320

in Figure 3a: when the wintertime Pacific jet core is strong, the wintertime Pacific storm321

track is weak. On average over the entire Pacific basin, the storm track is 25% weaker in322

the strong-core months compared to the weak-core months, a highly significant difference323

(Table 3, hereafter we refer to this as the “inverse relationship”). On the other hand,324

upstream seeding and jet-core strength appear to vary independently for the Pacific basin;325

the correlation coefficient between monthly averaged jet core strength and monthly averaged326

seeding upstream is only -0.08 (not shown). There is some indication from Figure 3 that327

months with a strong-core jet stream also have weaker upstream seeding, but this difference328

is not statistically significant and it is not robust to other choices of seasonality (for example,329

if we only include Jan/Feb instead of Dec/Jan/Feb, then there is no relationship). Thus we330

treat any cross-correlations, for example the possibility that variability in seeding controls331

variability in the jet, as second-order effects that we do not consider.332

For comparison, the equivalent plot for the Atlantic domain is also shown in Figure 3b.333

Note that there is a hint of the same behavior for the Atlantic storm track, though it is not334

statistically significant at 95%: when the Atlantic jet is strong, the Atlantic storm track is335

weak, indicating that some of the same processes that inhibit storminess over the Pacific336

13

domain may also be relevant for the Atlantic.337

a. Variance Statistics338

The inverse relationship between jet strength and Pacific storminess is examined by339

comparing climatologies of months with a strong-core jet to those with a weak-core jet. A340

composite map of the jet stream (zonal wind speed at 300hPa, contours) and storm track341

intensity (variance of 2-10 day filtered geopotential height, (Z ′)2 at 300hPa, shading) for the342

twenty winter months with the strongest jet is shown in Figure 4a, and the twenty months343

with the weakest jet is shown in Figure 4b.344

The differences in jet stream structure between strong-core and weak-core months are345

much more extensive than just their maximum speed. The strong-core jet stream is merid-346

ionally narrow and zonally elongated, whereas the weak-core jet stream is meridionally wide347

and tilts from SW to NE. Remarkably, one of the only things that the two jet stream patterns348

share is their latitude of maximum wind speed: throughout almost the entire Pacific domain349

(90◦E - 212.5◦E), the latitude of maximum wind speed is the same for the strong-core and350

weak-core jets to within one 2.5◦ gridbox.351

Further comparison is made from the difference maps shown in Figure 5. Figure 5a shows352

the difference in zonal wind strength between the strong-core jet and weak-core jet streams,353

and Figure 5b shows the difference in storminess (Z ′)2. In both of these maps, solid lines354

denote positive values, dashed lines denote negative values, red shading indicates values355

that are positive and different from zero with 95% confidence, and blue shading values that356

are negative and different from zero with 95%. As we expected from the discussion in the357

previous subsection, upstream of the storm track there is very little difference in storminess358

between the strong-core and weak-core months, and so upstream seeding is likely not playing359

an important role within the winter season.360

Though basin-averaged storminess is reduced by 25% for the strong-core jets relative to361

the weak-core jets (Table 3), Figure 5b shows that storminess is not reduced throughout362

14

the entire Pacific domain. In fact, there is a notable similarity between the two panels in363

Figure 5: regions of enhanced storminess tend to be co-located with regions of strong jet364

stream winds, and vice versa for weak jet stream winds. Note that while the strong jet is365

over 25ms−1 stronger than the weak jet at its core, the strong-core jet is so meridionally366

narrow that the westerly flow is over 15ms−1 weaker in the midlatitudes between about367

40◦ − 65◦N . The inverse relationship that occurs in winter between jet-core strength and368

Pacific basin-average storminess occurs because the negative values in Figure 5b are larger in369

both areal extent and amplitude than the positive values, and not because there is a striking370

spatial anticorrelation. Furthermore, linear baroclinic instability theories (e.g., the Eady371

and Charney models) are all derived from relationships in the midlatitudes that are most372

relevant to an eddy-driven jet, not a subtropical jet, so it is perhaps not surprising that linear373

theory is a less successful predictor of storminess in the strong-core jet months. From the374

above observations, we suggest that it is misleading to characterize the inverse relationship375

between jet-core strength and basin-average storminess as a striking anticorrelation.376

b. Feature-tracking statistics377

The relationship between jet stream structure and storminess is further illuminated using378

feature-tracking. Here we examine the number density and average amplitude of cyclonic379

disturbances, shown in Figure 6 for (a) the strong-core and (b) the weak-core jet months.380

The basic relationship between jet strength, number density, and disturbance amplitude381

is familiar and robust. For example, in both strong-core and weak-core months, disturbance382

amplitude maximizes downstream and slightly equatorward of the maximum in number383

density, and number density is displaced slightly poleward from the core of the jet stream,384

with a maximum near the poleward flank of the jet exit region. In this section we show results385

for cyclonic disturbances only. However, we have also looked at anticyclonic disturbances386

individually and both cyclonic and anticyclonic disturbances together. All of the conclusions387

that we discuss are robust to these choices and the maps are very similar.388

15

The spatial pattern of number density shows substantial differences between the strong-389

core and weak-core jet composites (Figure 7a). In the strong-core months, there are more390

tracks in the core of the jet stream and fewer to the North and South, consistent with a391

narrower, more focused jet. Although it is striking, this difference in the pattern of number392

density differences does not account for the inverse relationship. Averaged over the Pacific393

domain (20◦−65◦N , 160◦E−160◦W ) number density is only about 1% less in the strong-core394

jet months (2% less for cyclonic and 0.5% less for anticyconic disturbances, individually), a395

fraction that is not statistically significant.396

In contrast, the strong-core jet months are characterized by significantly weaker-amplitude397

disturbances throughout the Pacific domain (Figure 7b). Within the Pacific domain, the av-398

erage central amplitude of all features is significantly weaker (9% overall, 10% for cyclonic399

and 8% for anticyclonic disturbances individually) in the strong-core compared to the weak-400

core jet months. To place this in context, from the previous section we know that the401

wintertime inverse relationship is characterized by 25% less storminess within the Pacific402

domain during strong jet-core compared to the weak-core jet months. PRB10 showed that403

storminess (measured as the variance of geopotential height fluctuations) depends approx-404

imiately on central amplitude squared (PRB10, Equation B3). Therefore, about 91% (i.e.,405

(1− 0.25))/(1− 0.09)2) of the inverse relationship can be understood by considering distur-406

bance amplitude alone.407

Other factors of the feature-tracking statistics contribute to our general impression that408

Pacific domain storms are less likely to become deep, mature disturbances in the strong-core409

jet months. For example, in strong-core months there is significantly more (18%) cyclogenesis410

within the Pacific domain, however disturbances exist for significantly fewer days (0.8 days,411

a 14% reduction) on average. From these results, we conclude that the inverse relationship412

is not due to a lack of disturbances when the jet stream is strong. Rather, it occurs because413

existing disturbances are unable to utilize locally available potential energy and therefore414

they are unlikely to become deep, mature systems.415

16

c. E-vectors416

To this point we have treated the jet stream as an independent forcing of the storm417

track, however there are feedbacks between the two. To investigate the contribution of the418

transient eddy momentum convergence in forcing mean winds, we calculate Eliasson-Palm419

flux vectors (E) in the 2-10 day filtered fields to diagnose the flow of momentum within the420

Pacific domain for both the strong- and weak-core jet streams. The horizontal component,421

EH = (v′2 − u′2,−u′v′), together with its divergence at 300hPa, are shown in Figure 8. The422

divergence field is, as expected for a spatial derivative, relatively noisy, but the individual423

maxima and minima are significantly different from zero, and so we are confident in our424

interpretation of the overall patterns. Recall that regions of E-vector divergence (negative425

contours) correspond to regions of momentum convergence from the transient eddies to the426

atmospheres background state and will accelerate the jet stream, and E-vectors (arrows)427

correspond to the direction of Rossby wave propagation (e.g., Edmon et al. 1980; Thorncroft428

et al. 1993).429

In the weak-core jet months (Figure 8b), the primary role of the eddies is to spread out430

and maintain a diffuse eddy-driven jet stream. At the storm track entrance region west of431

about 190◦E, eddies propagate downstream on a westerly trajectory. In this region, eddies432

act to reinforce the jet stream: there is momentum convergence in the core of the jet stream433

and divergence on both the poleward and equatorward flanks. There is some equatorward434

propagation, but it is small in amplitude and primarily equatorward of the storm track. At435

the downstream end of the storm track east of about 190◦E, there is momentum convergence436

over a broad region that extends meridionally from the subtropics to the pole, and the most437

substantial region of convergence is in the vicinity of the Aleutian Low. These observations438

from the E-vectors are reminiscent of a canonical picture Pacific storm track storms: eddies439

grow in the baroclinic zone off the Asian coast, propagate poleward over their lifetimes, and440

reach maturity and eventually decay near the Aleutian Low.441

In the strong-core jet months (Figure 8a), the E-vectors indicate a very different story.442

17

In these months, eddies act to maintain a strong, narrow, and elongated jet stream across443

the entire Pacific domain. At the storm track entrance west of 190◦E, there is momentum444

convergence in the core and divergence on the flanks of the jet stream, similar to the weak-445

core jet months. However, the overall direction of eddy propagation is quite weak and446

anomalously easterly compared to the weak-core jet months. At the downstream end of the447

storm track east of about 190◦E, eddies act to zonally elongate and accelerate a narrow,448

subtropical jet stream. In the vicinity of the Aleutian Low, there is momentum divergence449

(E-vector convergence), which is opposite from the weak-core months. In the strong-core jet450

months, most momentum convergence occurs on the subtropical flank of the jet stream off451

the coast of the Baja California where cut-off lows, large ridging events, and other atypical452

flow patterns, are common.453

To summarize the above discussion, the location of momentum convergence by transient454

eddies acts to reinforce the observed jet stream patterns for both the strong- and weak-455

core jet streams. There are also indications of substantial structural and life cycle changes456

between transient eddies in the weak- and strong-core jet streams, with the canonical picture457

of Pacific storminess typical of the former and a more atypical picture in the latter.458

Another striking feature of these two distinct jets, shown in Figure 9, is the presence459

of equivalent barotropic flow in the strong-core relative to the weak-core jet streams. Sev-460

eral classic simple modeling studies (e.g., Thorncroft et al. 1993; Hartmann and Zuercher461

1998) found that the addition of barotropic flow causes the lifecycles of baroclinic waves to462

transition sharply from anticyclonic (LC1-type) to cyclonic (LC2-type). There is some indi-463

cation that there are more LC1-type disturbances in the weak-core jet, for example regions464

of momentum convergence are spread meridionally over a large over a broad area in the465

Eastern Pacific. More work is needed to evaluate this possibility, as the interpretation from466

the E-vectors is not conclusive.467

18

5. Summary and Conclusions468

In this paper we have examined how variations in two mechanisms, upstream seeding469

and jet core strength, relate to storminess within the cold season Pacific storm track. We470

found that the two mechanisms vary independently and we treat them as such; any cross-471

correlations, for example the possibility that variability in seeding controls variability in jet472

core strength, are second order effects that we do not consider.473

In Section 4, we examined how intraseasonal variations in the Pacific jet core during474

winter are related to Pacific storm track intensity. To the picture that already exists in the475

literature, we add five main observations. First, upstream seeding does not explain why476

months with a strong Pacific jet core tend to be accompanied by weaker than average storm477

tracks. This is in contrast to the climatological MWM, where seeding plays the dominant478

role, and underscores that the dynamics of the MWM are not an analog for the interseasonal479

variability of the wintertime Pacific storm track. It has been common to treat the inverse480

relationship seen between the seasonal cycle of the mean jet and storminess (i.e., the MWM)481

as analogous to the inverse relationship seen in wintertime variations between the jet and482

storminess, however our results here and in PRB10 show that this is not appropriate. Second,483

the inverse relationship is not a striking anticorrelation when viewed spatially; regions with484

stronger than average jet stream winds tend to have stronger than average storm tracks,485

and vice versa for regions with weaker than average jet stream winds, and this is true for486

both the strong- and weak-core jet months. Third, the inverse relationship occurs because487

disturbances rarely become deep, mature systems when the jet core is strongest. In the488

strong-core jet months relative to the weak-core jet months, there is significantly more (18%)489

cyclogenesis; the amplitude of disturbances is significantly weaker (9%). Fourth, transient490

eddies act to reinforce the observed jet stream patterns for both the strong- and weak-core491

jet streams, and so the jet structure differences are consistent with the way that eddies are492

forcing the background flow. Finally, there are some indications that the strong- and weak-493

core jet streams are preferentially composed of LC2- and LC1-type storms, respectively,494

19

consistent with there being more barotropic shear in the strong-core jet months (Figure 9).495

In Section 3, we examined the climatological between upstream seeding and downstream496

storminess, and found that about 15%-20% of observed Pacific storminess co-varies with497

upstream seeding (the exact percentage changes with modest modifications to the locations of498

the two domains or if we consider relative vorticity or meridional wind instead of geopotential499

height, not shown). This relationship persists through the entire cold season and in both the500

Atlantic and Pacific basins. About 17% of the disturbances that cross the centerline of the501

Pacific storm-track domain also exited the upstream domain, consistent with the correlations502

that we observe.503

The framework presented in Figure 2 adds further support and clarity to the conclusion504

in PRB10 that the MWM occurs primarily because there is a notable lack of storminess505

upstream of the Pacific storm track in the heart of winter. However, nonlinearities local to506

the Pacific storm track also likely contribute to the MWM: when variability due to upstream507

seeding is regressed out of the data (the ellipses in Figure 2a), we see that Pacific storm track508

intensity is essentially constant throughout the cold season, despite greater baroclinicity in509

winter. As we discussed in the end of section 3, these results may help clarify some seeming510

discrepancies in the literature about the causes of both the MWM and the winter-to-winter511

inverse relationship.512

Lastly, the results presented here and in PRB10 lead us to consider a fundamental climate513

question: can upstream seeding be viewed as random chaotic noise? In most cases, we believe514

that the answer is likely “yes”. We have shown that upstream seeding has a measurable515

and notable affect on downstream storminess, but the relationship is noisy (e.g., Figure 2).516

Furthermore, the effect on storm track intensity is approximately linear: the observations517

do not indicate a threshold behavior wherein seeding is a more effective forcing when it518

is particularly strong or particularly weak. However, we now have observational evidence519

that the answer to the posed question is “no” in some circumstances. In this paper, we520

noted that the Atlantic storm track is more strongly seeded than the Pacific storm track521

20

on average, and hypothesized that this may help explain why the Atlantic storm track is522

stronger overall than the Pacific. PRB10 found that a wintertime reduction in upstream523

seeding, which is related to a wintertime increase in static stability over Asia, causes the524

MWM. Donohoe and Battisti (2009) also found evidence that reduced seeding of the Atlantic525

during the Last Glacial Maximum resulted in a weak storm track. These studies and others526

(e.g., Orlanski 2005; Zurita-Gotor and Chang 2005) have raised many questions about the527

role of seeding in controlling the strength of the major storm tracks. Aside from seeding’s528

demonstrated impact on the MWM, the large amount of natural variability may be masking529

its other impacts, given the length of the observational record. Hence long-term integrations530

of climate models are likely the next step.531

Acknowledgments.532

This work was funded by National Science Foundation Continental Dynamics Grants533

6312293 and EAR-0507431.534

21

535

REFERENCES536

Blackmon, M. L., 1976: Climatological spectral study of 500 mb geopotential height of537

Northern Hemisphere. Journal of the Atmospheric Sciences, 33 (8), 1607–1623.538

Blackmon, M. L., J. M. Wallace, N. C. Lau, and S. L. Mullen, 1977: Observational study of539

Northern Hemisphere wintertime circulation. Journal of the Atmospheric Sciences, 34 (7),540

1040–1053.541

Chang, E. K. M., 2001: GCM and observational diagnoses of the seasonal and interannual542

variations of the Pacific storm track during the cool season. Journal of the Atmospheric543

Sciences, 58 (13), 1784–1800.544

Chang, E. K. M., 2005: The impact of wave packets propagating across Asia on Pacific545

cyclone development. Monthly Weather Review, 133 (7), 1998–2015.546

Chang, E. K. M. and Y. Guo, 2011: Comments on “The Source of the Midwinter Suppression547

in Storminess over the North Pacific”. Journal of Climate, 24 (19), 5187–5191, doi:{10.548

1175/2011JCLI3987.1}.549

Charney, J., 1947: The dynamics of long waves in a baroclinic westerly current. Journal of550

Meteorology, 4 (5), 135–162.551

Christoph, M., U. Ulbrich, and P. Speth, 1997: Midwinter suppression of Northern Hemi-552

sphere storm track activity in the real atmosphere and in GCM experiments. Journal of553

the Atmospheric Sciences, 54 (12), 1589–1599.554

Donohoe, A. and D. S. Battisti, 2009: The true asymmetry between synoptic cyclone and555

anticyclone amplitudes: Implications for filtering methods in Lagrangrian feature tracking.556

Journal of the Atmospheric Sciences.557

22

Eady, E. T., 1949: Long waves and cyclone waves. Tellus, 1, 33–42.558

Edmon, H., B. Hoskins, and M. Mcintyre, 1980: Eliassen-Palm cross-sections for the559

troposphere. Journal of the Atmospheric Sciences, 37 (12), 2600–2616, doi:{10.1175/560

1520-0469(1980)037〈2600:EPCSFT〉2.0.CO;2}.561

Hakim, G., 2003: Developing wave packets in the North Pacific storm track. Monthly Weather562

Review, 131 (11), 2824–2837.563

Harnik, N. and E. K. M. Chang, 2004: The effects of variations in jet width on the growth564

of baroclinic waves: Implications for midwinter Pacific storm track variability. Journal of565

the Atmospheric Sciences, 61 (1), 23–40.566

Hartmann, D. and P. Zuercher, 1998: Response of baroclinic life cycles to barotropic shear.567

Journal of the Atmospheric Sciences, 55 (3), 297–313, doi:{10.1175/1520-0469(1998)568

055〈0297:ROBLCT〉2.0.CO;2}.569

Hodges, K. I., 1994: A general-method for tracking analysis and its application to meteoro-570

logical data. Monthly Weather Review, 122 (11), 2573–2586.571

Hodges, K. I., 1995: Feature tracking on the unit-sphere. Monthly Weather Review, 123 (12),572

3458–3465.573

Hodges, K. I., 1999: Adaptive constraints for feature tracking. Monthly Weather Review,574

127 (6), 1362–1373.575

Hoskins, B. J. and K. I. Hodges, 2002: New perspectives on the Northern Hemisphere winter576

storm tracks. Journal of the Atmospheric Sciences, 59 (6), 1041–1061.577

Kageyama, M., P. Valdes, G. Ramstein, C. Hewitt, and U. Wyputta, 1999: Northern hemi-578

sphere storm tracks in present day and last glacial maximum climate simulations: A579

comparison of the European PMIP models. Journal of Climate, 12 (3), 742–760, doi:580

{10.1175/1520-0442(1999)012〈0742:NHSTIP〉2.0.CO;2}.581

23

Li, C. and D. S. Battisti, 2008: Reduced Atlantic storminess during last glacial maximum:582

Evidence from a coupled climate model. Journal of Climate, 21 (14), 3561–3579.583

Nakamura, H., 1992: Midwinter suppression of baroclinic wave activity in the Pacific. Journal584

of the Atmospheric Sciences, 49 (17), 1629–1642.585

Nakamura, H., T. Izumi, and T. Sampe, 2002: Interannual and decadal modulations recently586

observed in the Pacific storm track activity and east Asian winter monsoon. Journal of587

Climate, 15 (14), 1855–1874.588

Nakamura, H. and T. Sampe, 2002: Trapping of synoptic-scale disturbances into the North-589

Pacific subtropical jet core in midwinter. Geophysical Research Letters, 29 (16).590

Nakamura, H. and A. Shimpo, 2004: Seasonal variations in the Southern Hemisphere storm591

tracks and jet streams as revealed in a reanalysis dataset. Journal of Climate, 17 (9),592

1828–1844.593

Orlanski, I., 2005: A new look at the pacific storm track variability: Sensitivity to tropical594

SSTs and to upstream seeding. Journal of the Atmospheric Sciences, 62 (5), 1367–1390.595

Penny, S., G. H. Roe, and D. Battisti, 2010: The source of the midwinter suppression in596

storminess over the North Pacific. Journal of Climate, 23 (3), 634648.597

Penny, S., G. H. Roe, and D. Battisti, 2011: Reply to comments on the source of the598

midwinter suppression in storminess over the North Pacific. Journal of Climate, 24 (19),599

5192–5194.600

Thorncroft, C., B. Hoskins, and M. Mcintyre, 1993: 2 paradigms of baroclinic-wave life-601

cycle behavior. Quarterly Journal of the Royal Meteorological Society, 119 (509, Part602

a), 17–55, doi:{10.1002/qj.49711950903}.603

Uppala, S. M., et al., 2005: The ERA-40 re-analysis. Quarterly Journal of the Royal Meteo-604

rological Society, 131 (612), 2961–3012, part B.605

24

Wallace, J. M., G. H. Lim, and M. L. Blackmon, 1988: Relationship between cyclone606

tracks, anticyclone tracks and baroclinic wave-guides. Journal of the Atmospheric Sci-607

ences, 45 (3), 439–462.608

Yin, J., 2002: The peculiar behavior of baroclinic waves during the midwinter suppression609

of the Pacific storm track. Ph.D. thesis, University of Washington, 137 pp.610

Zhang, Y. Q., 1997: On the mechanisms of the mid-winter suppression of the Pacific storm-611

track. Ph.D. thesis, Princeton University.612

Zhang, Y. Q. and I. M. Held, 1999: A linear stochastic model of a GCM’s midlatitude storm613

tracks. Journal of the Atmospheric Sciences, 56 (19), 3416–3435.614

Zurita-Gotor, P. and E. K. M. Chang, 2005: The impact of zonal propagation and seeding615

on the eddy-mean flow equilibrium of a zonally varying two-layer model. Journal of the616

Atmospheric Sciences, 62 (7), 2261–2273.617

25

List of Tables618

1 Correlation between monthly averaged storminess ((Z ′)2, 300hPa) in the up-619

stream and downstream domains for the Pacific and Atlantic storm tracks.620

Bold indicates a correlation that is different from zero with 95% confidence621

(determined by t-test). 27622

2 Fractional difference between storminess ((Z ′)2) in Jan/Feb relative to Nov/Apr.623

Bold indicates that JF and NA storminess are different with 95% confidence624

(determined by t-test for difference between means). 28625

3 Fractional difference in storminess for the strong jet core DJF months relative626

to the weak jet core months for the Pacific storm track. Bold indicates that627

the differences are different from zero with 95% confidence (determined by628

t-test for difference between means). 29629

26

Table 1. Correlation between monthly averaged storminess ((Z ′)2, 300hPa) in the up-stream and downstream domains for the Pacific and Atlantic storm tracks. Bold indicates acorrelation that is different from zero with 95% confidence (determined by t-test).

Pacific AtlanticOct-Apr 0.41 0.37Dec, Jan, Feb 0.39 0.30Oct, Nov, Mar, Apr 0.38 0.39

27

Table 2. Fractional difference between storminess ((Z ′)2) in Jan/Feb relative to Nov/Apr.Bold indicates that JF and NA storminess are different with 95% confidence (determined byt-test for difference between means).

Pacific AtlanticUpstream Domain −14% +6%Downstream Domain −8% +17%

28

Table 3. Fractional difference in storminess for the strong jet core DJF months relative tothe weak jet core months for the Pacific storm track. Bold indicates that the differences aredifferent from zero with 95% confidence (determined by t-test for difference between means).

(Z ′)2 Number dens Amplitude Cyclogenesis Track durationTotal −25% −1% −9% +18% −14%Cyclones N/A −2% −10%Antiyclones N/A 0.5% −8%

29

List of Figures630

1 Sample of 100 random cyclonic disturbances that propagate through the up-631

stream domain. Black dots represent cyclogenesis location, and gray lines632

represent the path of disturbance center. The upstream and downstream do-633

mains are also indicated. 33634

2 Storminess ((Z ′)2, 300hPa) averaged in the upstream domain on the x-axis635

compared to the downstream domain on the y-axis for the seven cold season636

months between October 1958 and April 2001 within (a) Pacific storm track637

and (b) Atlantic storm track. In addition to the raw monthly data (see key),638

both plots show means for weak, medium, strong, and very strong seeding639

cases for winter months (Dec/Jan/Feb, squares, dark gray shaded ellipses for640

95% confidence) and transition months (Oct/Nov/Mar/Apr, triangles, light641

gray shaded ellipses for 95% confidence). Confidence intervals for the esti-642

mate of the mean are determined with a t-test. For this and all subsequent643

figures, “cold season” months are Oct - Apr, “winter” months are DJF, and644

“transition” months are ON, MA. 34645

3 Parallel to Figure 2, except for winter (DJF) months only in (a) the Pacific646

domain and (b) the Atlantic domain. Winter months with the top 20 jet647

core strengths are indicated with a red upward-pointing triangle, the bottom648

20 are indicated with a blue downward-pointing triangle, and all others are649

indicated with a black square. Shaded ellipses indicate 95% confidence for the650

estimate of the mean (filled triangles) as determined from a t-test. CHANGE:651

a and b labels 35652

4 Zonal winds (contours every 10m/s, zero contour thick) and storminess ((Z ′)2,653

shading every 1500m2 above 2000m2) at 300hPa for the (a) strong-core jet654

stream and (b) weak-core jet stream months. 36655

30

5 Difference between (a) zonal wind strength, contours every 5m/s, zero contour656

thick and dashed, and (b) storminess ((Z ′)2), contours every 1000m2. Both657

plots are strong-core jet months minus weak-core jet months, and shading658

where values are different from zero with 95% confidence (blue for negative,659

red for positive, determined from a t-test). CHANGE: red/blue matching,660

zero contour dashed. 37661

6 Number density (contours every 1 disturbance per month, bolded contours are662

1 and 6 per month) and feature amplitude (shading, interval every 2dm over663

14dm) in DJF for (a) the strong-core jet and (b) the weak-core jet months.664

The 30ms−1 zonal wind contour is indicated as a dashed white/black line for665

reference. 38666

7 Difference between (a) number density, contours every 0.5 disturbances per667

month and (b) feature amplitude, contours every 1dm. Both plots are strong-668

core jet months minus weak-core jet months, and shading where values are669

different from zero with 95% confidence (blue for negative, red for positive,670

determined from a t-test). Regions with less than 1 disturbance per month671

on average in either the strong-core or weak-core jet months are blocked from672

view. CHANGE: Red/blue matching 39673

8 Horizontal component of E-vectors at 300hPa (arrows) and divergence of E-674

vectors at the same level (contours every 3ms−1day−1 and shading) for the675

(a) strong-core and (b) weak-core jet months. Numbers that are different676

from zero with 95% confidence are shaded (blue for negative, red for positive,677

determined from a t-test). The zero contour is a long dashed line, positive678

contours are solid, and negative contours are dashed. E-vector arrows are679

plotted every 5◦. 40680

31

9 Vertical structure of zonal winds for the (a) strong-core jet stream and (b)681

weak-core jet stream at 180◦E. Shading interval is every 10ms−1, zero contour682

is dashed. Difference between the two jet streams (strong minus weak) are683

indicated on both plots with contours every 5ms−1, zero line omitted. 41684

32

Fig. 1. Sample of 100 random cyclonic disturbances that propagate through the upstreamdomain. Black dots represent cyclogenesis location, and gray lines represent the path ofdisturbance center. The upstream and downstream domains are also indicated.

33

2 4 6 8 10 12

4

6

8

10

12

14

16

octnovdecjanfebmaraprdjfonma

dow

nstre

am, 1

0 m3

2(a)

2 4 6 8 10 12

4

6

8

10

12

14

16

upstream, 10 m3 2

dow

nstre

am, 1

0 m3

2

(b)

Fig. 2. Storminess ((Z ′)2, 300hPa) averaged in the upstream domain on the x-axis comparedto the downstream domain on the y-axis for the seven cold season months between October1958 and April 2001 within (a) Pacific storm track and (b) Atlantic storm track. In additionto the raw monthly data (see key), both plots show means for weak, medium, strong, andvery strong seeding cases for winter months (Dec/Jan/Feb, squares, dark gray shaded ellipsesfor 95% confidence) and transition months (Oct/Nov/Mar/Apr, triangles, light gray shadedellipses for 95% confidence). Confidence intervals for the estimate of the mean are determinedwith a t-test. For this and all subsequent figures, “cold season” months are Oct - Apr,“winter” months are DJF, and “transition” months are ON, MA.

34

2 4 6 8 10 12

4

6

8

10

12

14

16

upstream, 10 m3 2

dow

nstre

am, 1

0 m3

2

djfmin djf jetmax djf jet

(a)

2 4 6 8 10 12

4

6

8

10

12

14

16

dow

nstre

am, 1

0 m3

2

djfmin djf jetmax djf jet

(a)

Fig. 3. Parallel to Figure 2, except for winter (DJF) months only in (a) the Pacific domainand (b) the Atlantic domain. Winter months with the top 20 jet core strengths are indicatedwith a red upward-pointing triangle, the bottom 20 are indicated with a blue downward-pointing triangle, and all others are indicated with a black square. Shaded ellipses indicate95% confidence for the estimate of the mean (filled triangles) as determined from a t-test.

35

(a) (b)

Fig. 4. Zonal winds (contours every 10m/s, zero contour thick) and storminess ((Z ′)2,shading every 1500m2 above 2000m2) at 300hPa for the (a) strong-core jet stream and (b)weak-core jet stream months.

36

(a) (b)

Fig. 5. Difference between (a) zonal wind strength, contours every 5m/s, zero contour thickand dashed, and (b) storminess ((Z ′)2), contours every 1000m2. Both plots are strong-corejet months minus weak-core jet months, and shading where values are different from zerowith 95% confidence (blue for negative, red for positive, determined from a t-test).

37

(a) (b)

Fig. 6. Number density (contours every 1 disturbance per month, bolded contours are 1and 6 per month) and feature amplitude (shading, interval every 2dm over 14dm) in DJF for(a) the strong-core jet and (b) the weak-core jet months. The 30ms−1 zonal wind contouris indicated as a dashed white/black line for reference.

38

(a) (b)

Fig. 7. Difference between (a) number density, contours every 0.5 disturbances per monthand (b) feature amplitude, contours every 1dm. Both plots are strong-core jet months minusweak-core jet months, and shading where values are different from zero with 95% confidence(blue for negative, red for positive, determined from a t-test). Regions with less than 1disturbance per month on average in either the strong-core or weak-core jet months areblocked from view.

39

100 m2 s−2

100 m2 s−2

(a)

(b)

Fig. 8. Horizontal component of E-vectors at 300hPa (arrows) and divergence of E-vectorsat the same level (contours every 3ms−1day−1 and shading) for the (a) strong-core and (b)weak-core jet months. Numbers that are different from zero with 95% confidence are shaded(blue for negative, red for positive, determined from a t-test). The zero contour is a longdashed line, positive contours are solid, and negative contours are dashed. E-vector arrowsare plotted every 5◦.

40

−1001020304050607080

20 30 40 50 60 70

200

300

400

500

600

700

800

900

Hei

ght (

hPa)

200

300

400

500

600

700

800

900

Hei

ght (

hPa)

20 30 40 50 60 70Latitude ( N)o

m/s:

(b)

(a)

Fig. 9. Vertical structure of zonal winds for the (a) strong-core jet stream and (b) weak-corejet stream at 180◦E. Shading interval is every 10ms−1, zero contour is dashed. Differencebetween the two jet streams (strong minus weak) are indicated on both plots with contoursevery 5ms−1, zero line omitted.

41