Abstract

Although prior studies have established that the extratropical flow pattern often amplifies downstream of recurving tropical cyclones (TCs), the extratropical flow response to recurving TCs has not to the authors' knowledge been systematically examined from a climatological perspective. In this study, a climatology of the extratropical flow response to recurving western North Pacific TCs is constructed from 292 cases of TC recurvature during 1979–2009. The extratropical flow response to TC recurvature is evaluated based on a time-lagged composite time series of an index of the North Pacific meridional flow surrounding TC recurvature. Similar time series are constructed for recurving TCs stratified by characteristics of the large-scale flow pattern, the TC, and the phasing between the TC and the extratropical flow to assess factors influencing the extratropical flow response to TC recurvature. Results reveal that following TC recurvature, significantly amplified flow develops over the North Pacific and persists for ~4 days. The tendency for significantly amplified North Pacific flow to develop following TC recurvature is sensitive to the strength of the TC–extratropical flow interaction (the phasing between the TC and the extratropical flow), which is based on the negative potential vorticity advection by the divergent outflow of the TC. In contrast, the tendency for significantly amplified North Pacific flow to develop following TC recurvature is relatively insensitive to the intensity or size of the recurving TC, or whether it subsequently reintensifies after becoming extratropical.

1. Introduction

As a tropical cyclone (TC) recurves into the midlatitudes and undergoes extratropical transition (ET; e.g., Jones et al. 2003), its interaction with the extratropical flow may be associated with the amplification of the extratropical flow pattern and the onset of high-impact weather events far downstream (e.g., Namias 1963; Agustí-Panareda et al. 2004, 2005; McTaggart-Cowan et al. 2007; Harr and Dea 2009; Anwender et al. 2010; Cordeira and Bosart 2010; Grams et al. 2011; Pantillon et al. 2013). For example, the recurvature of TC Oscar over the western North Pacific (WNP; September 1995) was followed by the amplification of a high-latitude ridge over western North America and an early season cold-air outbreak over the central United States (Fig. 1; Archambault et al. 2007). The cold-air outbreak on 20–21 September 1995 heavily damaged crops in the central plains and led to the earliest freeze on record in Des Moines, Iowa; Chicago, Illinois; and Wichita, Kansas [(National Oceanic and Atmospheric Administration/U.S. Department of Agriculture) NOAA/USDA 1995].

Fig. 1.

Schematic diagram illustrating key processes that link the recurvature and extratropical transition of WNP TC Oscar during 15–18 Sep 1995 to a cold-air outbreak over the central United States on 20–21 Sep 1995. Contours and color shading are schematic representations of the upper-tropospheric streamfunction pattern and wind speed maxima, respectively, for 15–21 Sep. The “L” and “H” labels denote surface cyclone and anticyclone centers, respectively.

Fig. 1.

Schematic diagram illustrating key processes that link the recurvature and extratropical transition of WNP TC Oscar during 15–18 Sep 1995 to a cold-air outbreak over the central United States on 20–21 Sep 1995. Contours and color shading are schematic representations of the upper-tropospheric streamfunction pattern and wind speed maxima, respectively, for 15–21 Sep. The “L” and “H” labels denote surface cyclone and anticyclone centers, respectively.

The recurvature of WNP TC Oscar is linked to the onset of an amplified flow pattern and high-impact weather downstream (Fig. 1) as described in the following stages:

  1. Stage 1: As the TC recurves ahead of a trough, its interaction with the extratropical flow contributes to ridge amplification and jet streak intensification (e.g., Bosart and Lackmann 1995; Agustí-Panareda et al. 2004; Grams et al. 2013b).

  2. Stage 2: In conjunction with Rossby wave dispersion (e.g., Orlanski and Sheldon 1995; Harr and Dea 2009), extratropical cyclogenesis occurs in the poleward-exit region of the jet streak, which, in turn, amplifies a downstream ridge (e.g., Riemer et al. 2008; Riemer and Jones 2010).

  3. Stage 3: This ridge amplification, which occurs over high-latitude western North America, provides dynamical forcing for a >1040-hPa surface anticyclone over west-central Canada that migrates equatorward along the eastern side of the Rockies (e.g., Colle and Mass 1995), resulting in a cold-air outbreak over the central United States.

Stages 1 and 2 are typical of cases of recurving TCs associated with Rossby wave amplification and dispersion (e.g., Riemer et al. 2008; Harr et al. 2008; Harr and Dea 2009; Anwender et al. 2010; Riemer and Jones 2010). Stage 3, on the other hand, is specific to this particular recurving TC case (Archambault et al. 2007).

Although recurving TCs are often linked to extratropical flow amplification (i.e., an increase in wave amplitude), they may be associated with strengthening of the jet stream in the absence of flow amplification [e.g., TC Jangmi (2008); Grams et al. 2013b], or with little to no discernible change to the extratropical flow pattern [e.g., TC Opal (1997); Harr and Elsberry 2000]. Since high-impact weather events may be associated with extratropical flow amplification (e.g., Bosart et al. 1996; Martius et al. 2008; Archambault et al. 2010), understanding which recurving TCs favor extratropical flow amplification is critical from a forecasting standpoint. The tendency for model forecast error and uncertainty associated with TC recurvature and ET to grow and spread downstream in conjunction with an amplifying extratropical flow pattern (e.g., Henderson et al. 1999; Harr et al. 2008; Anwender et al. 2008; Reynolds et al. 2009; Anwender et al. 2010; Pantillon et al. 2013) underscores the need for situational awareness regarding the extratropical flow response to TC recurvature.

The extratropical flow response to TC recurvature is known to be modulated by factors relating to the characteristics of the large-scale flow pattern (e.g., Riemer et al. 2008; Harr and Dea 2009; Reynolds et al. 2009), the TC (e.g., Riemer et al. 2008; Davis et al. 2008), and the phasing between the TC and the extratropical flow (e.g., Klein et al. 2002; Ritchie and Elsberry 2007; Riemer and Jones 2010; Grams et al. 2013a,b; Pantillon et al. 2013). However, to the authors' knowledge, the relative importance of these factors in modulating the extratropical flow response to TC recurvature has yet to be examined comprehensively. To this end, we first conduct a climatology of WNP TC recurvature to provide context for subsequent findings. This climatology is also intended to update and extend previous climatologies of WNP TC recurvature that were conducted several decades ago (i.e., Burroughs and Brand 1973), or as a part of larger studies and thus with a limited scope (i.e., Klein et al. 2000; Jones et al. 2003). We then perform a climatology of the extratropical flow response to WNP TC recurvature that, to the authors' knowledge, represents the first of its kind. This climatology is used to explore how characteristics of the large-scale flow pattern, the TC, and the phasing between the TC and the extratropical flow influence the amplification of the flow downstream of the TC. The geographical focus of this study on the WNP is motivated by the climatological tendency for more recurving TCs to occur over the WNP than over any other ocean basin (e.g., Jones et al. 2003), and by the potential for recurving TCs to impact North America by inducing Rossby wave amplification and dispersion along the strong waveguide associated with the North Pacific jet stream (e.g., Hakim 2003).

The remainder of the paper is organized as follows. In section 2, the data and methodology are described. Section 3 contains a climatology of recurving WNP TCs, whereas section 4 provides a climatology of the extratropical flow response to recurving WNP TCs. A discussion of key findings and conclusions are contained in section 5.

2. Data and methodology

a. Identification of recurving WNP TCs

All WNP TCs during 1979–2009 are identified using 6-h (0000, 0600, 1200, and 1800 UTC) best-track data compiled by the Regional Specialized Meteorological Center Tokyo-Typhoon Center, which is operated by the Japan Meteorological Agency (JMA) under the jurisdiction of the World Meteorological Organization. The JMA best-track dataset includes information on TCs occurring within the domain bounded by 0°–60°N, 100°E–180°, including the South China Sea. For a given tropical system, the JMA best-track dataset catalogues its latitude and longitude, minimum sea level pressure (MSLP), 10 m above ground level (AGL) maximum sustained wind (MSW), storm category [i.e., tropical depression (TD), tropical storm (TS), typhoon (TY), or extratropical cyclone (EC)], and orientation and length of the minimum and maximum radii of 10 m AGL 15.4 and 25.7 m s−1 (30 and 50 kt) winds. The JMA best-track dataset is well suited for this study because, in contrast to the Joint Typhoon Warning Center best-track dataset, the JMA dataset contains wind radius information for each TC during the entire period of interest (1979–2009), as well as the central latitude, central longitude, and MSLP for systems even after they have become ECs.

Since the routine aircraft reconnaissance of WNP TCs ended in 1987 (e.g., Barcikowska et al. 2012), the JMA best-track dataset relies mainly on empirically derived information obtained from estimates of TC intensity using the Dvorak current intensity technique (e.g., Velden et al. 2006) of cloud-pattern recognition applied to satellite imagery rather than in situ observations. Thus, there is some inherent uncertainty in the JMA best-track dataset. However, the data are expected to be of sufficient quality to identify paired groupings of recurving TCs with disparate characteristics (e.g., strong and weak TCs; section 2b).

In this study, TC recurvature (i.e., T + 0 h) is defined as a change in TC heading from westward to eastward while the TC moves poleward. The recurvature point is designated as the most westward position attained by the TC. If a recurving TC is at its most westward position for consecutive 6-h intervals, the first occurrence of its most westward position is designated the recurvature point. If a TC undergoes multiple recurvatures, the most westward position of the TC during its last recurvature is designated as the recurvature point. The TCs that track in a loop are not included in the study. Additional criteria used to identify recurving TCs are that a TC 1) must be at TS intensity or above (MSW of ≥17 m s−1) at recurvature, and 2) for consistency with previous studies of WNP ET (e.g., Klein et al. 2000, 2002; Jones et al. 2003), eventually become ECs (i.e., complete ET). The TC recurvature criteria used in this study are the same as those used in a climatology of WNP TC recurvature between May and December by Burroughs and Brand (1973), except that Burroughs and Brand (1973) did not require recurving TCs to be at least at TS intensity at recurvature or to become ECs.

In the JMA best-track dataset, 36.5% of the WNP TCs (292 of 801) that reach TS intensity or greater for 1979–2009 meet all the recurvature criteria stipulated above (Fig. 2a). If the WNP TCs that fail to complete ET but otherwise meet the TC recurvature criteria (i.e., non-ET TCs, N = 56; Fig. 2b) are included, the WNP TC recurvature frequency increases to 43.4% (348 out of 801). These frequencies compare favorably to the 40.3% WNP TC recurvature frequency (236 out of 586 WNP TCs) documented by Burroughs and Brand (1973) in their 1945–69 climatology.

Fig. 2.

Tracks of 1979–2009 (a) recurving WNP TCs (colored lines) and the mean recurving WNP TC track (black line; T − 48 h–T + 96 h), and (b) WNP TCs that fail to complete ET after recurvature (non-ET TCs). Track color corresponds to TC category (blue to TD, green to TS, red to TY, and purple to EC). Surface elevation is shaded according to the grayscale (km). The star in (a) denotes the mean recurvature point (24.9°N, 134.0°E) of recurving WNP TCs.

Fig. 2.

Tracks of 1979–2009 (a) recurving WNP TCs (colored lines) and the mean recurving WNP TC track (black line; T − 48 h–T + 96 h), and (b) WNP TCs that fail to complete ET after recurvature (non-ET TCs). Track color corresponds to TC category (blue to TD, green to TS, red to TY, and purple to EC). Surface elevation is shaded according to the grayscale (km). The star in (a) denotes the mean recurvature point (24.9°N, 134.0°E) of recurving WNP TCs.

b. Climatology categories

To examine factors influencing the extratropical flow response to recurving WNP TCs, recurving WNP TCs are grouped based on characteristics of (i) the large-scale flow pattern, (ii) the TC, and (iii) the phasing of the TC with the extratropical flow. For the “characteristics of the large-scale flow pattern” category, recurving TCs are stratified by time of year [i.e., by month for May–December, the time of year containing nearly all cases of TC recurvature (284 out of 292; 97.3%)] and by recurvature latitude. As will be shown in section 3c, both the time of year and latitude of recurvature are closely related to the large-scale flow pattern over the WNP. Given that most recurving WNP TCs (270 out of 292; 92.5%) recurve between 15° and 35°N, TC recurvatures are subdivided into four 5° latitude bins within this latitude band.

For the “characteristics of the TC” category, recurving TCs are stratified by their intensity and size at recurvature. Specifically, recurving TCs are categorized as strong or weak if, at recurvature, they feature an MSLP within the bottom or top quintile (N = 58), respectively, of all recurving TCs. Similarly, recurving TCs are categorized as large or small if, at recurvature, they feature a maximum 15.4 m s−1 (30 kt) wind radius within the top or bottom quintile, respectively, of all recurving TCs. The mean MSLP of recurving TCs categorized as strong is nearly 70 hPa lower than the mean MSLP of recurving TCs categorized as weak (925 vs 994 hPa). A similar disparity is seen between large and small recurving TCs, with large TCs featuring a mean length of the maximum 15.4 m s−1 (30 kt) wind radius that is nearly 4 times longer than that of small TCs [601 km (324 n mi) vs 153 km (83 n mi)].

The MSLP and the length of the maximum 15.4 m s−1 (30 kt) wind radius of recurving TCs are moderately anticorrelated (r = −0.61). Of note is that such a relationship between TC intensity and size is not observed for WNP TCs overall (Weatherford and Gray 1988; Chan and Chan 2012).

For the “characteristics of the phasing of the TC with the extratropical flow” category, recurving TCs are stratified by whether they subsequently reintensify as an EC (i.e., JMA best-track MSLP decreases after the system completes ET), by whether they become strong or weak ECs [i.e., JMA best-track MSLP at the completion of ET is within the bottom or top quintile (N = 58) of all recurving TCs], and by the strength of their TC–extratropical flow interaction. A recurving TC that subsequently reintensifies as an EC is considered to have undergone favorable phasing with the extratropical flow (e.g., with an upstream trough or downstream jet streak; Klein et al. 2002). A recurving TC that becomes a strong EC is similarly influenced by a favorable extratropical flow pattern into which it is moving. The basis for considering the strength of the TC–extratropical flow interaction as a measure of phasing, as well as how the interaction strength is assessed, will be described in section 2d.

c. Evaluation of the extratropical flow response to recurving WNP TCs

The extratropical flow response to recurving WNP TCs is evaluated by constructing time-lagged composite time series of an index of the meridional flow over the North Pacific. The North Pacific meridional flow index is computed for every 6-h period for 1979–2009 from an area-weighted average of the absolute value of meridional wind on the dynamic tropopause (DT) [1.5-potential vorticity unit (PVU; 1 PVU = 10−6 K m2 kg−1 s−1) surface] for 25°–55°N, 140°E–120°W (Fig. 3) using 2.5° National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis data (Kalnay et al. 1996; Kistler et al. 2001).

Fig. 3.

The 25°–55°N, 140°E–120°W domain used to compute the North Pacific meridional flow index.

Fig. 3.

The 25°–55°N, 140°E–120°W domain used to compute the North Pacific meridional flow index.

A two-sided Student's t test (e.g., Wilks 2006, see section 5.2.1) is used to establish whether the departure from climatology of the composite North Pacific meridional flow index is significant (i.e., whether the North Pacific flow pattern is significantly amplified relative to climatology). North Pacific meridional flow index climatologies for all recurving WNP TCs and for each category of recurving WNP TCs (see section 2b) are constructed by weighting the monthly means and standard deviations of the North Pacific meridional flow index (Table 1) according to the monthly distribution of recurving TCs in the category. The North Pacific meridional flow index tends to be higher (i.e., the North Pacific flow pattern tends to be more amplified) and more variable in October–May than in June–September (Table 1).

Table 1.

Long-term (1979–2009) monthly means and standard deviations of the North Pacific meridional flow index used to compute statistical significance of the North Pacific meridional flow index anomalies associated with recurving WNP TCs categorized by month.

Long-term (1979–2009) monthly means and standard deviations of the North Pacific meridional flow index used to compute statistical significance of the North Pacific meridional flow index anomalies associated with recurving WNP TCs categorized by month.
Long-term (1979–2009) monthly means and standard deviations of the North Pacific meridional flow index used to compute statistical significance of the North Pacific meridional flow index anomalies associated with recurving WNP TCs categorized by month.

d. Evaluation of the TC–extratropical flow interaction

Although several features of a TC contribute to its interaction with the extratropical flow (e.g., the cyclonic TC core and anticyclonic TC outflow anomaly; Riemer et al. 2008; Riemer and Jones 2010), the divergent outflow of a TC impinging on the jet stream has been identified as a key signature of Rossby wave amplification occurring in conjunction with TC recurvature and ET (e.g., Riemer et al. 2008; Anwender et al. 2010; Hodyss and Hendricks 2010; Pantillon et al. 2013; Grams et al. 2013b,a). Specifically, ridge amplification and jet streak intensification occur as negative PV advection by the divergent outflow associated with a TC acts to deform PV contours poleward and strengthen the meridional PV gradient, respectively (Fig. 4). Note that this framework includes the role of diabatic heating: low upper-tropospheric PV air that arises from the vertical redistribution of PV by diabatic heating is then advected poleward by the diabatically driven divergent outflow.

Fig. 4.

Schematic representation of ridge amplification and jet streak intensification associated with the divergent outflow of a TC impinging upon an upper-tropospheric jet stream/waveguide. Vectors represent the upper-tropospheric irrotational wind (i.e., divergent outflow) associated with the TC. Shading denotes negative PV advection by the irrotational wind.

Fig. 4.

Schematic representation of ridge amplification and jet streak intensification associated with the divergent outflow of a TC impinging upon an upper-tropospheric jet stream/waveguide. Vectors represent the upper-tropospheric irrotational wind (i.e., divergent outflow) associated with the TC. Shading denotes negative PV advection by the irrotational wind.

In this study, the negative PV advection by the divergent outflow associated with a TC is used as a metric for the strength of the TC–extratropical flow interaction during TC recurvature. The TC–extratropical flow interaction metric is computed as follows:

  1. The point and time of the maximum TC–extratropical flow interaction (i.e., the maximum 250–150-hPa layer-averaged negative PV advection by the irrotational wind)1 surrounding TC recurvature (i.e., T − 48 h through T + 144 h) are identified from the 6-h 2.5° NCEP–NCAR reanalysis. A point of maximum interaction cannot be identified during 20 cases of recurving TCs, which are referred to as no-interaction cases, because the negative PV advection by the irrotational wind associated with the TC never exceeds an arbitrary threshold of 1 PVU day−1.

  2. Six-hour geostationary full-disk infrared satellite images obtained from NOAA/National Climatic Data Center (NCDC)/Global International Satellite Cloud Climatology Project (ISCCP) B1 Browse System (GIBBS; http://www.ncdc.noaa.gov/gibbs/) for T − 48 h through T + 144 h are visually inspected to confirm that the point of maximum interaction is in proximity to the TC cirrus shield and thus directly associated with the TC divergent outflow.

  3. A spatial average of the 250–150-hPa layer-averaged PV advection by the irrotational wind is computed for a 15° × 15° domain2 centered on the point of maximum interaction.

  4. A temporal average of the spatially averaged PV advection by the irrotational wind is computed for a 48-h period centered on the time of maximum interaction.

The outcome of the above procedure is that a value representing the strength of the TC–extratropical flow interaction is obtained for each case of TC recurvature featuring an interaction (272 out of 292 recurving TCs). The strength of the interaction is weakly correlated to the intensity and size of the recurving TC: the correlations of the interaction metric with the MSLP and with the length of the maximum 15.4 m s−1 (30 kt) wind radius of all recurving TCs are r = 0.23 and r = −0.27, respectively.

An example of the analysis used to compute the TC–extratropical flow interaction metric is provided for recurving WNP TC Oscar (1995; see Fig. 1) at the time of maximum interaction (0600 UTC 16 September, 6 h after recurvature; Fig. 5a). At this time, TC Oscar is moving northeastward toward an upper-tropospheric PV trough–ridge couplet. Strong negative PV advection by divergent outflow associated with ascent over and northeastward of the TC extends along the eastern flank of the trough and western flank of the ridge. The TC–extratropical flow interaction is thus characterized by favorable phasing between the TC and ridge (e.g., Riemer et al. 2008; Grams et al. 2013b), with the negative PV advection by the divergent outflow of the TC acting to anchor the ridge and facilitate its amplification (not shown). In conjunction with ridge amplification, an anticyclonically curved jet streak is undergoing intensification (not shown) along and downstream of the region of negative PV advection by the divergent outflow associated with the TC (Fig. 5a).

Fig. 5.

The time of maximum TC–extratropical flow interaction associated with TC Oscar (0600 UTC 16 Sep 1995) as represented in (a) the NCEP–NCAR reanalysis and (b) the Climate Forecast System Reanalysis. Analyses show 500-hPa ascent (dashed green, every 2 × 10−3 hPa s−1), and 250–150-hPa layer-averaged wind speed (light gray denotes 50 m s−1; dark gray denotes 70 m s−1), PV (solid blue, every 3 PVU), irrotational wind (vectors, >3 m s−1 only), and PV advection by the irrotational wind (dashed red, every 3 PVU day−1, negative values only). The TC symbol denotes the JMA best-track position of TC Oscar. The star denotes the point of maximum interaction, and the black rectangle in (a) shows the 15° × 15° domain used to compute the interaction metric. The inset in (a) shows the coinciding infrared satellite image for approximately the same domain as the main figure. The star and rectangle in this inset have the same meaning as in the main figure.

Fig. 5.

The time of maximum TC–extratropical flow interaction associated with TC Oscar (0600 UTC 16 Sep 1995) as represented in (a) the NCEP–NCAR reanalysis and (b) the Climate Forecast System Reanalysis. Analyses show 500-hPa ascent (dashed green, every 2 × 10−3 hPa s−1), and 250–150-hPa layer-averaged wind speed (light gray denotes 50 m s−1; dark gray denotes 70 m s−1), PV (solid blue, every 3 PVU), irrotational wind (vectors, >3 m s−1 only), and PV advection by the irrotational wind (dashed red, every 3 PVU day−1, negative values only). The TC symbol denotes the JMA best-track position of TC Oscar. The star denotes the point of maximum interaction, and the black rectangle in (a) shows the 15° × 15° domain used to compute the interaction metric. The inset in (a) shows the coinciding infrared satellite image for approximately the same domain as the main figure. The star and rectangle in this inset have the same meaning as in the main figure.

The 2.5° NCEP–NCAR reanalysis is used to create the analyses of TC–extratropical flow interactions because of its availability and its application in previous studies of ET (e.g., Sinclair 2002, 2004). Although its relatively coarse resolution limits its ability to capture the structure of TCs in the tropics, it is comparable to other reanalyses in its ability to represent the synoptic aspects of TC–extratropical flow interactions.3

To illustrate this point, the analysis of TC Oscar at the time of maximum interaction constructed from the NCEP–NCAR reanalysis (Fig. 5a) is compared to a smoothed version of the analysis constructed from the recently available 0.5° NCEP Climate Forecast System Reanalysis (CFSR; Saha et al. 2010) (Fig. 5b).4 In each analysis, the synoptic features, including the PV trough–ridge couplet and divergent outflow of the TC, are quite similar, and, crucially for this study, the fields of negative PV advection by the irrotational wind are nearly identical.

In addition to the comparable representation of the TC–extratropical flow interaction in the reanalyses, the representation of the interactions in these reanalyses is consistent with satellite imagery. The region of negative PV advection by the irrotational wind (Figs. 5a,b) aligns closely with the upstream edge of the cirrus shield emanating from TC Oscar (inset in Fig. 5a) indicative of outflow from the TC (e.g., Klein et al. 2000).

To rank the strength of the TC–extratropical flow interaction for each case of TC recurvature featuring an interaction, the magnitudes of the interaction metric are ordered from largest to smallest. The TCs with an interaction metric in the top and bottom quintile (N = 54) are categorized as strong and weak interactions, respectively (Tables 2 and 3). Based on the interaction ranking, the recurvature of TC Oscar is associated with the second strongest interaction during the 1979–2009 period of study (Table 2).

Table 2.

Recurving WNP TCs associated with strong TC–extratropical flow interactions.

Recurving WNP TCs associated with strong TC–extratropical flow interactions.
Recurving WNP TCs associated with strong TC–extratropical flow interactions.
Table 3.

Recurving WNP TCs associated with weak TC–extratropical flow interactions. Note that recurving TCs with no interaction point (N = 20) are not categorized as weak interaction cases but rather as “no interaction” cases.

Recurving WNP TCs associated with weak TC–extratropical flow interactions. Note that recurving TCs with no interaction point (N = 20) are not categorized as weak interaction cases but rather as “no interaction” cases.
Recurving WNP TCs associated with weak TC–extratropical flow interactions. Note that recurving TCs with no interaction point (N = 20) are not categorized as weak interaction cases but rather as “no interaction” cases.

3. Climatology of recurving WNP TCs

a. Characteristics of typical and extreme recurving WNP TCs

Characteristics of recurving WNP TCs based on the JMA best-track dataset are compiled in Table 4. The mean latitude and longitude of TC recurvature are 24.9°N and 134.0°E, respectively, although the latitude and longitude of recurvature ranges from 9.5° to 42.3°N and from 109.8° to 169.8°E, respectively. On average, TCs are at minimal TY intensity at recurvature, with a mean MSLP of 964 hPa and a mean MWS of 34.9 m s−1 (67.9 kt).5 However, recurving TCs can be much more intense: at recurvature, TC Flo (1990) featured an 890-hPa MSLP and a 61.7 m s−1 (120 kt) MSW.

Table 4.

Summary of recurving WNP TC characteristics.

Summary of recurving WNP TC characteristics.
Summary of recurving WNP TC characteristics.

On average, the length of the maximum 15.4 m s−1 (30 kt) wind radius of a recurving WNP TC is 359 km (194 nm), and the length of the maximum 25.7 m s−1 (50 kt) wind radius is 119 km (64 n mi). The largest TCs at recurvature in terms of wind extent are TC Keith (1997) [a 1111-km (600 n mi) maximum 15.4 m s−1 (30 kt) wind radius] and TC Holly (1984) [a 370-km (200 n mi) maximum 25.7 m s−1 (50 kt) wind radius]. The smallest TC at recurvature was TC Vipa (2001), which had a 74-km (40 n mi) maximum 15.4 m s−1 (30 kt) wind radius.

b. Interannual variability of WNP TC recurvature frequency

During 1979–2009, 292 out of 801 WNP TCs (36.5%) meet the recurvature criteria used in this study, a mean of 9.4 yr−1. The annual number of recurving TCs ranges from 4 in 1983 and 1998 to 15 in 2004 (Fig. 6a). Only approximately one-fifth of the variance in the annual number of recurving TCs (21%) is accounted for by the variance in the annual number of all TCs (Fig. 6a), likely reflecting the influence of the substantial interannual variability in steering flow over the WNP on TC motion (e.g., Harr and Elsberry 1991; Wang and Chan 2002; Camargo et al. 2007).

Fig. 6.

Yearly time series of (a) all WNP TCs (open diamonds) and recurving WNP TCs (filled diamonds), and (b) recurving WNP TCs (filled diamonds) and recurving WNP TCs that reintensify as ECs (filled circles) for 1979–2009. The square of the linear correlation between time series (r2) is given in the top-right corner of each panel.

Fig. 6.

Yearly time series of (a) all WNP TCs (open diamonds) and recurving WNP TCs (filled diamonds), and (b) recurving WNP TCs (filled diamonds) and recurving WNP TCs that reintensify as ECs (filled circles) for 1979–2009. The square of the linear correlation between time series (r2) is given in the top-right corner of each panel.

During 1979–2009, 124 recurving WNP TCs subsequently reintensify after becoming ECs, a mean of 4 yr−1. Recurving TCs that reintensify thus comprise 15.5% of all TCs and 42.5% of all recurving TCs. The finding that 15.5% of all TCs reintensify as ECs is somewhat lower than the 27% (30 out of 112) found by Klein et al. (2000) for 1994–98 (June–October only).

Out of 292 recurving WNP TCs, 124 TCs subsequently reintensify after becoming ECs (42.5%), a mean of 4 yr−1. The annual number of recurving TCs that reintensify as ECs varies from 1 in 1998 and 2008 to 8 in 1994 (Fig. 6b). Approximately two-fifths of the variance in the annual number of recurving TCs that reintensify as ECs (39%) is accounted for by the variance in the annual number of all recurving TCs (Fig. 6b), suggesting that the favorability of the extratropical flow for phasing with an ex-TC varies from year to year.

c. Monthly variability of recurving WNP TC frequency

Recurving WNP TCs generally occur during May–December and are most frequent during August–October (Fig. 7a), findings that corroborate the time-of-year distributions of recurving TCs documented by Burroughs and Brand (1973) and ET cases documented by Klein et al. (2000) and Jones et al. (2003). The monthly distribution of recurving TCs generally mirrors that of all TCs, although the September peak in the number of recurving TCs lags the August peak in the number of all TCs. The monthly distribution of non-ET TCs is relatively flat (Fig. 7a), indicating that these cases are not strongly favored at a particular time of the year.

Fig. 7.

Monthly distributions of (a) all WNP TCs (unshaded), recurving WNP TCs (light shading), and WNP TCs that fail to complete ET after recurvature (non-ET TCs; dark shading), and (b) recurving WNP TCs (light shading) and recurving WNP TCs that reintensify as ECs (dark shading). Numbers above each bar in (a) and (b) indicate percentages of TCs that undergo recurvature and percentages of recurving TCs that reintensify as ECs for each month, respectively.

Fig. 7.

Monthly distributions of (a) all WNP TCs (unshaded), recurving WNP TCs (light shading), and WNP TCs that fail to complete ET after recurvature (non-ET TCs; dark shading), and (b) recurving WNP TCs (light shading) and recurving WNP TCs that reintensify as ECs (dark shading). Numbers above each bar in (a) and (b) indicate percentages of TCs that undergo recurvature and percentages of recurving TCs that reintensify as ECs for each month, respectively.

August–October is the preferred time of year for recurving WNP TCs to subsequently reintensify as ECs (Fig. 7b). September features the most recurving TCs and most recurving TCs that reintensify as ECs (Fig. 7b). September also features the highest percentage of TCs that recurve (53%; Fig. 7a) over the entire year, and the highest percentage that reintensify as ECs after recurvature (54%; Fig. 7b) when considering the primary recurving TC season (i.e., May–December). These results corroborate the finding by Klein et al. (2000) that September is the most favorable month for recurving TCs to subsequently reintensify as ECs.

January–April is the most unfavorable time of year for recurving WNP TCs to reintensify as ECs (Fig. 7b). In general, November–July have a lower percentage of recurving TCs that reintensify as ECs (Fig. 7b)6 compared to August–October, which suggests that a favorable phasing of an ex-TC with an extratropical disturbance is less likely during November–July than during the remainder of the year.

d. Monthly variability of recurving WNP TC tracks

Inspection of WNP tracks stratified by month for the primary recurving TC season (Fig. 8) reveals a time-of-year dependence on TC track. Specifically, TCs tend to recurve more sharply in May and October–December (Figs. 8a,f–h) than in June–September (Figs. 8b–e), corroborating previous findings by Burroughs and Brand (1973). A related finding is that TCs that recurve in May, November, and December tend to reach high latitudes (i.e., track poleward of 50°N) less frequently than those in June–October (cf. Figs. 8a,g,h and 8b–f).

Fig. 8.

As in Fig. 2, but for recurving TC tracks during (a) May, (b) June, (c) July, (d) August, (e) September, (f) October, (g) November, and (h) December. The black line denotes the mean recurving TC track for T − 48 h–T + 72 h.

Fig. 8.

As in Fig. 2, but for recurving TC tracks during (a) May, (b) June, (c) July, (d) August, (e) September, (f) October, (g) November, and (h) December. The black line denotes the mean recurving TC track for T − 48 h–T + 72 h.

e. Monthly variability of WNP TC recurvature points

Consistent with earlier climatologies of WNP TC recurvature (i.e., Riehl 1972; Burroughs and Brand 1973), the latitude of TC recurvature is found to vary strongly with time of year (Fig. 9a). Of all the months during the primary recurving TC season, August generally features the highest-latitude TC recurvature points (Fig. 9a; also cf. Figs. 8d and 8a–c,e–h) and May and December generally feature the lowest-latitude TC recurvature points (Fig. 9a; also cf. Figs. 8a,h and 8b–g). Recurvature latitude varies more with time of year than does longitude (cf. Figs. 9a,b), although TCs exhibit a slight tendency to recurve farther west in May–July than in August–December (Fig. 9b). Consistent with the seasonal distribution of TC recurvature latitude (Fig. 9a), relative to other months low-latitude (i.e., 15°–20°N) recurvatures are most common in May and October–December, whereas high-latitude (i.e., 30°–35°N) recurvatures are most common in July–September (Fig. 10).

Fig. 9.

Box plots showing (a) latitudinal and (b) longitudinal distributions of WNP TC recurvature points by month (May–December). The lower and upper bounds of each box are drawn at the 25th and 75th percentiles, respectively; the bar and asterisk inside each box denote the median and mean, respectively; and the whiskers extend from the 25th and 75th percentiles to the minimum and maximum values, respectively.

Fig. 9.

Box plots showing (a) latitudinal and (b) longitudinal distributions of WNP TC recurvature points by month (May–December). The lower and upper bounds of each box are drawn at the 25th and 75th percentiles, respectively; the bar and asterisk inside each box denote the median and mean, respectively; and the whiskers extend from the 25th and 75th percentiles to the minimum and maximum values, respectively.

Fig. 10.

Monthly frequency distributions of recurving WNP TCs stratified by recurvature latitude.

Fig. 10.

Monthly frequency distributions of recurving WNP TCs stratified by recurvature latitude.

The time-of-year dependence of the latitude of WNP TC recurvature is closely related to the relationship between the time of year and the strength and latitudinal position of the North Pacific jet stream evident in analyses of monthly TC recurvature points overlain on the corresponding monthly mean 250-hPa jet stream (Fig. 11). The 1000–500-hPa thickness field is also plotted to depict thickness troughs and ridges, as well as the baroclinicity associated with the jet stream. Note that these analyses are displayed for every other month of the primary recurving TC season.

Fig. 11.

WNP TC recurvature points by month and corresponding long-term (1979–2009) monthly mean 250-hPa wind speed (shaded, every 10 m s−1 starting at 20 m s−1) and 1000–500-hPa thickness (dashed, every 6 dam) over eastern Asia and the west–central North Pacific for (a) May, (b) July, (c) September, and (d) November. Thick solid line denotes axis of thickness trough.

Fig. 11.

WNP TC recurvature points by month and corresponding long-term (1979–2009) monthly mean 250-hPa wind speed (shaded, every 10 m s−1 starting at 20 m s−1) and 1000–500-hPa thickness (dashed, every 6 dam) over eastern Asia and the west–central North Pacific for (a) May, (b) July, (c) September, and (d) November. Thick solid line denotes axis of thickness trough.

These analyses illustrate that months with a stronger, more equatorward jet stream (e.g., November; Fig. 11d) tend to feature lower-latitude TC recurvatures than months with a weaker, more poleward jet stream (e.g., July; Fig. 11b). The seasonal variation of the mean characteristics of the North Pacific jet stream has implications for the extratropical flow response associated with TC recurvature (discussed subsequently in section 4b).

In May (Fig. 11a), WNP TCs tend to recurve south and east of a positively tilted thickness trough near the eastern Asian coast and mainly east of the equatorward entrance region of a relatively straight jet stream. The TCs in July (Fig. 11b) also tend to recurve south and east of a positively tilted thickness trough. The jet stream in July is weak in comparison to May (cf. Figs. 11a,b).

In September (Fig. 11c), WNP TCs exhibit a tendency to recurve just east of a slightly positively tilted thickness trough into a low-amplitude trough–ridge thickness pattern extending from northeastern China to just east of Japan. Furthermore, September TCs tend to recurve near the equatorward entrance region of an anticyclonically curved jet stream. The tendency for September TCs to recurve just downstream of a trough and near the equatorward entrance region of an anticyclonically curved jet stream indicates that the September extratropical flow pattern favors the phasing of recurving TCs with the extratropical flow. The jet stream in September is stronger than in July (Fig. 11b) but of comparable strength to the jet stream in May (Fig. 11a).

In November (Fig. 11d), WNP TCs generally recurve east of the equatorward entrance region of a jet stream and well to the south and east of a broad, neutrally tilted thickness trough situated along the eastern Asian coast. The jet stream and thickness gradient are strong in November relative to May, July, and September (cf. Fig. 11d and 11a–c).

f. Monthly variability of recurving WNP TC intensity and size

Recurving WNP TCs tend to be more intense and larger in September–November relative to recurving TCs that occur during the rest of the primary recurving TC season (Figs. 12a and 12b, respectively). To quantify the time-of-year difference in the intensity of recurving TCs (Fig. 12a), the mean MSLP of recurving TCs in September–November is compared to that of recurving TCs in June–August. Similarly, to quantify the time-of-year difference in the size of recurving TCs (Fig. 12b), the mean length of the maximum 15.4 m s−1 (30 kt) wind radius of recurving TCs in September–November is compared to that of recurving TCs in June–August. The mean MSLP of recurving TCs in September–November is 959 hPa, significantly lower than the mean of 971 hPa in June–August, and the mean length of the maximum 15.4 m s−1 (30 kt) wind radius of recurving TCs in September–November is 404 km (218 n mi), significantly longer than the mean of 306 km (165 n mi) in June–August. These time-of-year differences are significant at the 99.9% confidence level based on a two-sided Student's t test, and are in line with previous findings that recurving WNP TCs tend to be more intense and larger in fall relative to other times of the year (e.g., Burroughs and Brand 1973).

Fig. 12.

As in Fig. 9, but for the (a) MSLP (hPa) and (b) length of the maximum 15.4 m s−1 (30 kt) wind radius (km) at recurvature.

Fig. 12.

As in Fig. 9, but for the (a) MSLP (hPa) and (b) length of the maximum 15.4 m s−1 (30 kt) wind radius (km) at recurvature.

4. Climatology of extratropical flow response to recurving WNP TCs

a. Characteristic extratropical flow response

The composite evolution of the North Pacific meridional flow index surrounding WNP TC recurvature during 1979–2009 (Fig. 13) provides the basis for evaluating the extratropical flow response to WNP TC recurvature. The mean North Pacific meridional flow index increases from the climatological level at T + 0 h to nearly 10% above the climatological level at T + 54 h. The median and the 25th and 75th percentiles of the index also increase during this period. These findings indicate that the North Pacific flow typically becomes amplified following TC recurvature. Furthermore, the mean index is significantly above normal at the 95% confidence level during T + 12 h–T + 114 h, indicating that the North Pacific flow pattern tends to be significantly amplified for a ~4-day period following TC recurvature.

Fig. 13.

Time-lagged composite time series of the North Pacific meridional flow index (m s−1) (left ordinate) and departure from climatology (right ordinate) surrounding WNP TC recurvature (T + 0 h; thick vertical line). Circles correspond to the mean value, with a statistically significant anomaly at the 95% confidence level indicated by the blue dot, and statistically significant anomalies at the 99% confidence level shown by the red dots. The solid gray curve represents the median, and the bounding dashed gray curves denote the 25th and 75th percentiles.

Fig. 13.

Time-lagged composite time series of the North Pacific meridional flow index (m s−1) (left ordinate) and departure from climatology (right ordinate) surrounding WNP TC recurvature (T + 0 h; thick vertical line). Circles correspond to the mean value, with a statistically significant anomaly at the 95% confidence level indicated by the blue dot, and statistically significant anomalies at the 99% confidence level shown by the red dots. The solid gray curve represents the median, and the bounding dashed gray curves denote the 25th and 75th percentiles.

A somewhat counterintuitive finding is that following the recurvature of non-ET WNP TCs, the North Pacific flow also tends to amplify (Fig. 14). The mean North Pacific meridional flow index associated with non-ET TCs is significantly above normal at the 95% confidence level during T + 18 h–T + 114 h. This result suggests that a recurving TC may still interact with the extratropical flow without completing ET. Given that a consistent definition of ET has not been applied operationally (e.g., Jones et al. 2003), this result also may partially be an artifact of the inadequacy of the best-track dataset for use in discriminating between recurving TCs that do and do not complete ET.

Fig. 14.

Time-lagged composite time series of the North Pacific meridional flow index departure from climatology (%) for recurving TCs and non-ET TCs. Significant departures at the 95% (99%) confidence level are shown by the open (filled) dots.

Fig. 14.

Time-lagged composite time series of the North Pacific meridional flow index departure from climatology (%) for recurving TCs and non-ET TCs. Significant departures at the 95% (99%) confidence level are shown by the open (filled) dots.

The findings presented in Figs. 13 and 14 indicate that the North Pacific flow pattern tends to be amplified for ~4 days following WNP TC recurvature. It is thus of interest to determine whether this association is sensitive to characteristics of the large-scale flow pattern, the TC, and the phasing of the TC with the extratropical flow.

b. Stratification by characteristics of the large-scale flow pattern

To determine whether the North Pacific flow amplification following WNP TC recurvature is sensitive to characteristics of the large-scale flow pattern, composite time series of the North Pacific meridional flow index are constructed for recurving TCs stratified by month (Fig. 15) and recurvature latitude (Fig. 16). Recall from Fig. 11 that time of year and recurvature latitude both are closely related to the strength and latitudinal position of the North Pacific jet stream.

Fig. 15.

As in Fig. 14, but for recurving TCs in May–December.

Fig. 15.

As in Fig. 14, but for recurving TCs in May–December.

Fig. 16.

As in Fig. 14, but for TCs recurving within 15°–20°, 20°–25°, 25°–30°, and 30°–35°N latitude bands.

Fig. 16.

As in Fig. 14, but for TCs recurving within 15°–20°, 20°–25°, 25°–30°, and 30°–35°N latitude bands.

Stratifying recurving TCs by month (Fig. 15) indicates that the North Pacific flow is significantly amplified at the 95% confidence level for at least a 36-h period following TC recurvature in all months except June and December. Following TC recurvature in August–November, however, the North Pacific flow is significantly amplified at the 99% confidence level for at least a 24-h period. The TC recurvatures in August and September feature the most prolonged period of significantly amplified flow at the 95% confidence level (i.e., at least 90 h; T + 18 h–T + 108 h and T + 30 h–T + 132 h, respectively).

These findings are consistent with the favorable conditions for the phasing of a recurving TC with the extratropical flow at this time of year indicated by the tendency for TCs in September to recurve just downstream of a climatological trough and near the climatological equatorward entrance region of the North Pacific jet stream (Fig. 11c). The result that the association between TC recurvature and amplified North Pacific flow is relatively weak in June relative to August–November is consistent with the tendency for the North Pacific jet stream, and thus the Rossby waveguide, to be weak in June–July relative to August–November (e.g., cf. Figs. 11b and 11c–d).

Stratifying recurving WNP TCs by recurvature latitude (Fig. 16) reveals that the statistical significance of the amplified North Pacific flow that develops following 15°–20°N TC recurvatures is generally lower than that following 20°–25°, 25°–30°, and 30°–35°N TC recurvatures. These results indicate that TC recurvature at lower latitudes (i.e., 15°–20°N), which is characteristic of TCs that recurve in December when the association between TC recurvature and amplified North Pacific flow is relatively weak (Fig. 15), is slightly less favorable for North Pacific flow amplification than TC recurvature at higher latitudes (i.e., 20°–35°N). This finding may relate to the tendency for low-latitude TC recurvature to occur in the presence of a strong jet stream and associated strong vertical wind shear: strong vertical wind shear may hamper Rossby wave amplification and dispersion by low-latitude recurving TCs because of its more detrimental effect on TCs at lower latitudes compared to those at higher latitudes (e.g., Jones 1995; DeMaria 1996).

In addition, 20°–25°N TC recurvatures tend to feature a relatively long period of significantly amplified North Pacific flow (108 h; T + 6 h–T + 114 h) relative to 15°–20°N and 25°–30°N TC recurvatures (66 h; T + 30 h–T + 96 h) (Fig. 16). In contrast, 30°–35°N TC recurvatures tend to feature a relatively short period of significantly amplified North Pacific flow (54 h; T + 12 h–T + 66 h) (Fig. 16).

c. Stratification by characteristics of the TC

To determine whether the North Pacific flow amplification following WNP TC recurvature is sensitive to characteristics of the TC, composite time series of the North Pacific meridional flow index are constructed for recurving TCs stratified by their intensity and size (Fig. 17). Based on these time series, significantly amplified North Pacific flow tends to develop following the recurvature of both strong and weak TCs, and large and small TCs, respectively. Strong and weak recurving TCs (Fig. 17, top panels) are associated with significantly amplified North Pacific flow for at least a 78-h period (T + 24 h–T + 114 h and T + 18 h–T + 96 h, respectively). Similarly, large and small recurving TCs (Fig. 17, bottom panels) are associated with significantly amplified North Pacific flow for at least a 72-h period (T + 30 h–T + 102 h and T + 6 h–T + 90 h, respectively). However, the raw departure of the North Pacific meridional flow index from climatology is less following the recurvature of weak and small TCs compared to strong and large TCs (Fig. 17, top and bottom panels, respectively). The latter result is consistent with the minimum in both magnitude and standard deviation in the North Pacific meridional flow index in June–August (Table 1) when recurving TCs tend to be relatively weak and small (section 3f). That is, relative to other times of the year, relatively little amplification of the North Pacific flow is required to yield significant departures in the North Pacific meridional flow index in June–August, when relatively weak and small recurving TCs are favored.

Fig. 17.

As in Fig. 14, but for strong and weak, and large and small recurving TCs.

Fig. 17.

As in Fig. 14, but for strong and weak, and large and small recurving TCs.

d. Stratification by characteristics of the phasing of the TC with the extratropical flow

To determine whether the North Pacific flow amplification following WNP TC recurvature is sensitive to characteristics of the phasing of the TC with the extratropical flow, composite time series of the North Pacific meridional flow index are constructed for recurving TCs stratified based on whether they subsequently reintensify as ECs, based on whether they become strong or weak ECs upon completion of ET, and based on the strength of the TC–extratropical flow interaction (Fig. 18). Stratifying recurving TCs based on whether they reintensify as ECs (Fig. 18, top panels) reveals that both reintensifying and nonreintensifying WNP TCs are associated with significantly amplified North Pacific flow at the 99% confidence level for at least a 78-h period (T + 18 h–T + 96 h and T + 18 h–T + 114 h, respectively), although the raw departure of the North Pacific meridional flow index from climatology is greater for reintensifying TCs than for nonreintensifying TCs.

Fig. 18.

As in Fig. 14, but for reintensifying and nonreintensifying recurving TCs, recurving TCs that become strong and weak ECs, and recurving TCs associated with strong and weak TC–extratropical flow interactions.

Fig. 18.

As in Fig. 14, but for reintensifying and nonreintensifying recurving TCs, recurving TCs that become strong and weak ECs, and recurving TCs associated with strong and weak TC–extratropical flow interactions.

Stratifying recurving WNP TCs based on whether they become strong or weak ECs (Fig. 18, middle panels) indicates that strong EC cases tend to be associated with significantly amplified North Pacific flow following recurvature, while weak EC cases do not. Following the recurvature of TCs that become strong ECs, the North Pacific flow pattern is significantly amplified at the 99% confidence level for 108 h (T + 18 h–T + 126 h). On the other hand, following the recurvature of TCs that become weak ECs, the North Pacific flow pattern never becomes significantly amplified at the 99% confidence level, and is significantly amplified at the 95% confidence level for only 6 h (T + 36 h–T + 42 h).

Stratifying recurving TCs based on the strength of the TC–extratropical flow interaction (Fig. 18, bottom panels) indicates that strong interactions tend to favor the development of significantly amplified flow over the North Pacific following TC recurvature in comparison to weak TC–extratropical flow interactions. The North Pacific flow is significantly amplified for an 84-h period (T + 18 h–T + 102 h) following strong interactions in contrast to a 30-h period (T + 54 h–T + 84 h) following weak interactions. Furthermore, following strong interactions, the flow pattern is significantly amplified at the 99% confidence level for a 72-h period (T + 24 h–T + 96 h), whereas the flow pattern is never significantly amplified at the 99% confidence level following weak interactions. Therefore, the tendency for significantly amplified North Pacific flow to develop following TC recurvature is sensitive to the strength of the TC–extratropical flow interaction.

e. Characteristics of TC–extratropical flow interactions

Given the statistical relationship between the strength of the TC–extratropical flow interaction and the tendency for amplified flow to develop following TC recurvature, it is of interest to explore the characteristics of strong and weak interactions (Figs. 1922). Strong interactions tend to occur closer in time to TC recurvature compared to weak interactions (Fig. 19). For strong interactions, the mean time of maximum interaction is T + 31 h, which is significantly different at the 99.9% confidence level based on a two-sided Student's t test than the T + 56 h mean time of maximum interaction for weak interactions.

Fig. 19.

Frequency distribution of the timing of the maximum TC–extratropical flow interaction relative to WNP TC recurvature for strong (solid curve) and weak (dashed curve) TC–extratropical flow interactions.

Fig. 19.

Frequency distribution of the timing of the maximum TC–extratropical flow interaction relative to WNP TC recurvature for strong (solid curve) and weak (dashed curve) TC–extratropical flow interactions.

Fig. 20.

Monthly frequency distribution of all recurving WNP TCs (gray fill), and of recurving WNP TCs associated with strong (solid curve) and weak (dashed curve) TC–extratropical flow interactions.

Fig. 20.

Monthly frequency distribution of all recurving WNP TCs (gray fill), and of recurving WNP TCs associated with strong (solid curve) and weak (dashed curve) TC–extratropical flow interactions.

Fig. 21.

As in Fig. 9, but for the (a) MSLP (hPa) and (b) length of the maximum 15.4 m s−1 (30 kt) wind radius (km) at recurvature for recurving TCs associated with strong and weak TC–extratropical flow interactions.

Fig. 21.

As in Fig. 9, but for the (a) MSLP (hPa) and (b) length of the maximum 15.4 m s−1 (30 kt) wind radius (km) at recurvature for recurving TCs associated with strong and weak TC–extratropical flow interactions.

Fig. 22.

As in Fig. 9, but for the longitude of the maximum TC–extratropical flow interaction for recurving TCs associated with strong and weak interactions.

Fig. 22.

As in Fig. 9, but for the longitude of the maximum TC–extratropical flow interaction for recurving TCs associated with strong and weak interactions.

Recurving WNP TCs associated with strong interactions occur more frequently in September–December and less frequently in April–August relative to the climatological distribution of recurving TCs (Fig. 20). Recurving TCs associated with weak interactions occur more frequently in May, July, and August and less frequently in June and September–December relative to the climatological distribution of recurving TCs. The mean recurvature date of recurving TCs associated with strong interactions is 24 September, which is significantly later in the year at the 99.9% confidence level based on a two-sided Student's t test than the 18 August mean recurvature date of recurving TCs associated with weak interactions.

Recurving WNP TCs associated with strong interactions tend to be more intense (Fig. 21a) and larger (Fig. 21b) relative to recurving TCs associated with weak interactions, although the intensity and size differences are much less than the differences between recurving TCs categorized as strong or weak, and large or small, respectively (see section 2b). The mean MSLP of recurving TCs associated with strong interactions is 953 hPa, which is significantly lower at the 99% confidence level based on a two-sided Student's t test than the mean 969-hPa MSLP of recurving TCs associated with weak interactions. The mean length of the maximum 15.4 m s−1 (30 kt) wind radius of recurving TCs associated with strong interactions is 437 km (236 n mi), which is significantly larger at the 99.9% confidence level than the 341-km (184 n mi) mean length of the maximum 15.4 m s−1 (30 kt) radius of recurving TCs associated with weak interactions.

Strong interactions tend to occur farther west than weak interactions (Fig. 22), which is consistent with the tendency for strong interactions to occur closer in time to TC recurvature than weak interactions (Fig. 19). A comparison of the mean longitude of interaction for strong and weak interactions reveals that the mean longitude of the maximum interaction associated with strong interactions is 136°E, which is significantly different at the 99.9% confidence level based on a two-sided Student's t test than the 148°E mean longitude of the maximum interaction associated with weak interactions.

5. Discussion and conclusions

Building upon previous studies (e.g., Harr and Dea 2009; Anwender et al. 2010) that document specific cases of recurving TCs associated with extratropical flow amplification, a 31-yr climatological analysis for 1979–2009 illustrates that WNP TC recurvature tends to be associated with an amplification of the North Pacific flow pattern. The typical persistence of amplified North Pacific flow for a ~4-day period following TC recurvature, combined with the frequent occurrence of recurving WNP TCs (9.4 yr−1), suggests that recurving WNP TCs may be important modulators of the intraseasonal climate variability over the North Pacific, and by inference, downstream over North America.

During the primary recurving WNP TC season of May–December, the months of August–November are found to have the strongest association between TC recurvature and amplified North Pacific flow. In contrast, June and December are found to have the weakest associations between TC recurvature and amplified North Pacific flow. These results imply that August–November may be particularly favorable, and December and June particularly unfavorable, for recurving WNP TCs to be associated with Rossby wave amplification and dispersion.

The tendency for the North Pacific flow to amplify significantly in association with recurving WNP TCs is relatively insensitive to TC intensity and size: both strong and weak, and large and small recurving TCs are associated with a significant amplification of the North Pacific flow pattern. However, the North Pacific flow pattern tends to be more amplified (based on the raw departure of the North Pacific meridional flow index from climatology) for strong and large TCs than for weak and small TCs. Overall, these findings suggest that the characteristics of the recurving TC are not primary factors governing the extratropical flow response to recurving TCs.

In addition, the tendency for the North Pacific flow to amplify significantly in association with recurving WNP TCs is relatively insensitive to whether the TC reintensifies as an EC, or even whether it completes ET (becomes an EC). The finding that the North Pacific flow tends to amplify following TC recurvature regardless of whether the TC subsequently reintensifies as an EC is consistent with prior studies (e.g., Harr and Dea 2009) that link Rossby wave amplification and dispersion to recurving TCs that do not subsequently reintensify as ECs. These results suggest that the completion of ET and subsequent reintensification as an EC may not be primary factors modulating the extratropical flow response to recurving TCs.

On the other hand, a relationship is noted between the tendency for the North Pacific flow to amplify following TC recurvature and whether TCs subsequently become strong or weak ECs. However, since the North Pacific flow tends to amplify within 24 h of recurvature, prior to when recurving TCs typically become ECs (e.g., Klein et al. 2000), this relationship may be associative rather than causal (i.e., strong ECs tend to go hand in hand with highly amplified flow).

A key finding of this study is that the relationship between North Pacific flow amplification and WNP TC recurvature is sensitive to the strength of the TC–extratropical flow interaction (based on the negative potential vorticity advection by the divergent outflow of the TC). This finding is likely a result of the important role of the divergent outflow of the TC in Rossby wave amplification noted in previous studies (e.g., Davis et al. 2008; Riemer et al. 2008; Riemer and Jones 2010; Hodyss and Hendricks 2010; Pantillon et al. 2013; Grams et al. 2013a). However, a composite analysis study of strong and weak TC–extratropical flow interactions is under way to help establish the physical mechanisms underpinning the statistical relationship between the strength of the interaction and North Pacific flow amplification identified in this study.

Recurving TCs associated with strong TC–extratropical flow interactions tend to be somewhat more intense and larger on average than those associated with weak interactions. These findings are consistent with previous findings that relatively intense and large TCs feature relatively strong and spatially expansive divergent circulations that may increase the resilience of the TCs to vertical wind shear (e.g., Jones 1995; DeMaria 1996; Davis et al. 2008; Riemer et al. 2008). These findings are also in line with findings by Davis et al. (2008) that for vertical wind shear that increases up to 20–25 m s−1, relatively intense and large TCs tend to be associated with an increase in total vertical mass flux, which is proportional to diabatic heating, precipitation, and upper-level divergence.

Davis et al. (2008) further note that sustained vertical mass flux associated with a recurving TC may perturb the jet stream (i.e., provide Rossby wave forcing). Therefore, given that strong TC–extratropical flow interactions favor the amplification of the flow pattern, it is proposed that the divergent outflow of a TC impinging upon the PV gradient associated with the jet stream is a key synoptic signature of Rossby wave forcing by a TC. It is thus suggested that the strength of the TC–extratropical flow interaction may be a physically meaningful indicator for flow amplification following TC recurvature.

Based on these findings, evaluating forecasts of the extratropical flow evolution surrounding TC recurvature in the TC–extratropical flow interaction framework may prove useful in an operational setting. For example, ensemble prediction system members could be stratified based on the strength of the TC–extratropical flow interaction to illustrate the potential impact of the variation in the strength of the TC–extratropical flow on the extratropical flow evolution following TC recurvature. Whether the strength of the TC–extratropical flow interaction influences the predictability of the downstream extratropical flow evolution following TC recurvature should be assessed in future work.

Acknowledgments

The authors gratefully acknowledge Dr. Chris Davis (NCAR) for his invaluable contributions to the direction of this research while hosting the first author as a graduate student visitor in the NCAR Advanced Study Program. Research support was provided by NSF Grant AGS-0935830 and NOAA Grant NA09OAR4310192. A portion of this research was completed while the first author held a National Research Council Research Associateship Award at the Naval Postgraduate School. The authors wish to thank Dr. Michael Riemer (University of Mainz) and an anonymous reviewer for their thoughtful and detailed suggestions that substantially improved the quality of this manuscript.

REFERENCES

REFERENCES
Agustí-Panareda
,
A.
,
C. D.
Thorncroft
,
G. C.
Craig
, and
S. L.
Gray
,
2004
:
The extratropical transition of Hurricane Irene (1999): A potential vorticity perspective
.
Quart. J. Roy. Meteor. Soc.
,
130
,
1047
1074
.
Agustí-Panareda
,
A.
,
S. L.
Gray
,
G. C.
Craig
, and
C.
Thorncroft
,
2005
:
The extratropical transition of Tropical Cyclone Lili (1996) and its crucial contribution to a moderate extratropical development
.
Mon. Wea. Rev.
,
133
,
1562
1573
.
Anwender
,
D.
,
P. A.
Harr
, and
S. C.
Jones
,
2008
:
Predictability associated with the downstream impacts of the extratropical transition of tropical cyclones: Case studies
.
Mon. Wea. Rev.
,
136
,
3226
3247
.
Anwender
,
D.
,
S. C.
Jones
,
M.
Leutbecher
, and
P. A.
Harr
,
2010
:
Sensitivity experiments for ensemble forecasts of the extratropical transition of Typhoon Tokage (2004)
.
Quart. J. Roy. Meteor. Soc.
,
136
,
183
200
.
Archambault
,
H. M.
,
L. F.
Bosart
, and
D.
Keyser
,
2007
: Recurving typhoons as precursors to an early season arctic outbreak over the continental U.S. Preprints, Ninth Northeast Regional Operational Workshop, Albany, NY, National Weather Service/Amer. Meteor. Soc., 1–55. [Available online at http://cstar.cestm.albany.edu/nrow/nrow9/Archambault/archambault.ppt.]
Archambault
,
H. M.
,
D.
Keyser
, and
L. F.
Bosart
,
2010
:
Relationships between large-scale regime transitions and major cool-season precipitation events in the northeastern United States
.
Mon. Wea. Rev.
,
138
,
3454
3473
.
Barcikowska
,
M.
,
F.
Feser
, and
H.
von Storch
,
2012
:
Usability of best track data in climate statistics in the western North Pacific
.
Mon. Wea. Rev.
,
140
,
2818
2830
.
Bosart
,
L. F.
, and
G. M.
Lackmann
,
1995
:
Postlandfall tropical cyclone reintensification in a weakly baroclinic environment: A case study of Hurricane David (September 1979)
.
Mon. Wea. Rev.
,
123
,
3268
3291
.
Bosart
,
L. F.
,
G. J.
Hakim
,
K. R.
Tyle
,
M. A.
Bedrick
,
W. E.
Bracken
,
M. J.
Dickinson
, and
D. M.
Schultz
,
1996
:
Large-scale antecedent conditions associated with the 12–14 March 1993 cyclone (“Superstorm ‘93”) over eastern North America
.
Mon. Wea. Rev.
,
124
,
1865
1891
.
Burroughs
,
L. D.
, and
S.
Brand
,
1973
:
Speed of tropical storms and typhoons after recurvature in the western North Pacific Ocean
.
J. Appl. Meteor.
,
12
,
452
458
.
Camargo
,
S. J.
,
A. W.
Robertson
,
S. J.
Gaffney
,
P.
Smyth
, and
M.
Ghil
,
2007
:
Cluster analysis of typhoon tracks. Part II: Large-scale circulation and ENSO
.
J. Climate
,
20
,
3654
3676
.
Chan
,
K. T. F.
, and
J. C. L.
Chan
,
2012
:
Size and strength of tropical cyclones as inferred from QuikSCAT data
.
Mon. Wea. Rev.
,
140
,
811
824
.
Colle
,
B. A.
, and
C. F.
Mass
,
1995
:
The structure and evolution of cold surges east of the Rocky Mountains
.
Mon. Wea. Rev.
,
123
,
2577
2610
.
Cordeira
,
J. M.
, and
L. F.
Bosart
,
2010
:
The antecedent large-scale conditions of the “Perfect Storms” of late October and early November 1991
.
Mon. Wea. Rev.
,
138
,
2546
2569
.
Davis
,
C. A.
,
S. C.
Jones
, and
M.
Riemer
,
2008
:
Hurricane vortex dynamics during Atlantic extratropical transition
.
J. Atmos. Sci.
,
65
,
714
736
.
DeMaria
,
M.
,
1996
:
The effect of vertical shear on tropical cyclone intensity change
.
J. Atmos. Sci.
,
53
,
2076
2088
.
Evans
,
C.
, and
R.
Hart
,
2008
:
Analysis of the wind field evolution associated with the extratropical transition of Bonnie (1998)
.
Mon. Wea. Rev.
,
136
,
2047
2065
.
Grams
,
C. M.
, and
Coauthors
,
2011
:
The key role of diabatic processes in modifying the upper-tropospheric wave guide: A North Atlantic case study
.
Quart. J. Roy. Meteor. Soc.
,
137
,
2174
2193
.
Grams
,
C. M.
,
S. C.
Jones
,
C. A.
Davis
,
P. A.
Harr
, and
M.
Weissmann
,
2013a
:
The impact of Typhoon Jangmi (2008) on the midlatitude flow. Part I: Upper-level ridge building and modification of the jet
.
Quart. J. Roy. Meteor. Soc.
, doi:10.1002/qj.2091,
in press
.
Grams
,
C. M.
,
S. C.
Jones
, and
C. A.
Davis
,
2013b
:
The impact of Typhoon Jangmi (2008) on the midlatitude flow. Part II: Downstream evolution
.
Quart. J. Roy. Meteor. Soc.
,
doi:10.1002/qj.2119, in press
.
Hakim
,
G. J.
,
2003
:
Developing wave packets in the North Pacific storm track
.
Mon. Wea. Rev.
,
131
,
2824
2837
.
Harr
,
P. A.
, and
R. L.
Elsberry
,
1991
:
Tropical cyclone track characteristics as a function of large-scale circulation anomalies
.
Mon. Wea. Rev.
,
119
,
1448
1468
.
Harr
,
P. A.
, and
R. L.
Elsberry
,
2000
:
Extratropical transition of tropical cyclones over the western North Pacific. Part I: Evolution of structural characteristics during the transition process
.
Mon. Wea. Rev.
,
128
,
2613
2633
.
Harr
,
P. A.
, and
J. M.
Dea
,
2009
:
Downstream development associated with the extratropical transition of tropical cyclones over the western North Pacific
.
Mon. Wea. Rev.
,
137
,
1295
1319
.
Harr
,
P. A.
,
D.
Anwender
, and
S. C.
Jones
,
2008
:
Predictability associated with the downstream impacts of the extratropical transition of tropical cyclones: Methodology and a case study of Typhoon Nabi (2005)
.
Mon. Wea. Rev.
,
136
,
3205
3225
.
Henderson
,
J. M.
,
G. M.
Lackmann
, and
J. R.
Gyakum
,
1999
:
An analysis of Hurricane Opal's forecast track errors using quasigeostrophic potential vorticity inversion
.
Mon. Wea. Rev.
,
127
,
292
307
.
Hodyss
,
D.
, and
E.
Hendricks
,
2010
:
The resonant excitation of baroclinic waves by the divergent circulation of recurving tropical cyclones
.
J. Atmos. Sci.
,
67
,
3600
3616
.
Jones
,
S. C.
,
1995
:
The evolution of vortices in vertical shear. I: Initially barotropic vortices
.
Quart. J. Roy. Meteor. Soc.
,
121
,
821
851
.
Jones
,
S. C.
, and
Coauthors
,
2003
:
The extratropical transition of tropical cyclones: Forecast challenges, current understanding, and future directions
.
Wea. Forecasting
,
18
,
1052
1092
.
Kalnay
,
E.
, and
Coauthors
,
1996
:
The NCEP/NCAR 40-Year Reanalysis Project
.
Bull. Amer. Meteor. Soc.
,
77
,
437
471
.
Kistler
,
R.
, and
Coauthors
,
2001
:
The NCEP–NCAR 50-Year Reanalysis: Monthly means CD-ROM and documentation
.
Bull. Amer. Meteor. Soc.
,
82
,
247
267
.
Klein
,
P. M.
,
P. A.
Harr
, and
R. L.
Elsberry
,
2000
:
Extratropical transition of western North Pacific tropical cyclones: An overview and conceptual model of the transformation stage
.
Wea. Forecasting
,
15
,
373
395
.
Klein
,
P. M.
,
P. A.
Harr
, and
R. L.
Elsberry
,
2002
:
Extratropical transition of western North Pacific tropical cyclones: Midlatitude and tropical cyclone contributions to reintensification
.
Mon. Wea. Rev.
,
130
,
2240
2259
.
Martius
,
O.
,
C.
Schwierz
, and
H. C.
Davies
,
2008
:
Far-upstream precursors of heavy precipitation events on the Alpine south-side
.
Quart. J. Roy. Meteor. Soc.
,
134
,
417
428
.
McTaggart-Cowan
,
R.
,
L. F.
Bosart
,
J. R.
Gyakum
, and
E. H.
Atallah
,
2007
:
Hurricane Katrina (2005). Part II: Evolution and hemispheric impacts of a diabatically generated warm pool
.
Mon. Wea. Rev.
,
135
,
3927
3949
.
Namias
,
J.
,
1963
:
Large-scale air-sea interactions over the North Pacific from summer 1962 through the subsequent winter
.
J. Geophys. Res.
,
68
,
6171
6186
.
NOAA/USDA
,
1995
: Weekly Weather and Crop Bulletin, No. 39, 20 pp. [Available online at http://usda01.library.cornell.edu/usda/waob/weather_weekly//1990s/1995/weather_weekly-09-26-1995.pdf.]
Orlanski
,
I.
, and
J. P.
Sheldon
,
1995
:
Stages in the energetics of baroclinic systems
.
Tellus
,
47
,
605
628
.
Pantillon
,
F.
,
J.-P.
Chaboureau
,
C.
Lac
, and
P.
Mascart
,
2013
:
On the role of a Rossby wave train during the extratropical transition of Hurricane Helene (2006)
.
Quart. J. Roy. Meteor. Soc.
,
139
,
370
386
,
doi:10.1002/qj.1974
.
Reynolds
,
C. A.
,
M. S.
Peng
, and
J. H.
Chen
,
2009
:
Recurving tropical cyclones: Singular vector sensitivity and downstream impacts
.
Mon. Wea. Rev.
,
137
,
1320
1337
.
Riehl
,
H.
,
1972
:
Intensity of recurved typhoons
.
J. Appl. Meteor.
,
11
,
613
615
.
Riemer
,
M.
, and
S. C.
Jones
,
2010
:
The downstream impact of tropical cyclones on a developing baroclinic wave in idealized scenarios of extratropical transition
.
Quart. J. Roy. Meteor. Soc.
,
136
,
617
637
.
Riemer
,
M.
,
S. C.
Jones
, and
C. A.
Davis
,
2008
:
The impact of extratropical transition on the downstream flow: An idealized modeling study with a straight jet
.
Quart. J. Roy. Meteor. Soc.
,
134
,
69
91
.
Ritchie
,
E. A.
, and
R. L.
Elsberry
,
2007
:
Simulations of the extratropical transition of tropical cyclones: Phasing between the upper-level trough and tropical cyclone
.
Mon. Wea. Rev.
,
135
,
862
876
.
Saha
,
S.
, and
Coauthors
,
2010
:
The NCEP Climate Forecast System Reanalysis
.
Bull. Amer. Meteor. Soc.
,
91
,
1015
1057
.
Schenkel
,
B. A.
, and
R. E.
Hart
,
2012
:
An examination of tropical cyclone position, intensity, and intensity life cycle within atmospheric reanalysis datasets
.
J. Climate
,
25
,
3453
3475
.
Sinclair
,
M. R.
,
2002
:
Extratropical transition of southwest Pacific tropical cyclones. Part I: Climatology and mean structure changes
.
Mon. Wea. Rev.
,
130
,
590
609
.
Sinclair
,
M. R.
,
2004
:
Extratropical transition of southwest Pacific tropical cyclones. Part II: Midlatitude circulation characteristics
.
Mon. Wea. Rev.
,
132
,
2145
2168
.
Velden
,
C.
, and
Coauthors
,
2006
:
The Dvorak tropical cyclone intensity estimation technique: A satellite-based method that has endured for over 30 years
.
Bull. Amer. Meteor. Soc.
,
87
,
1195
1210
.
Wang
,
B.
, and
J. C. L.
Chan
,
2002
:
How strong ENSO events affect tropical storm activity over the western North Pacific
.
J. Climate
,
15
,
1643
1658
.
Weatherford
,
C. L.
, and
W. M.
Gray
,
1988
:
Typhoon structure as revealed by aircraft reconnaissance. Part I: Data analysis and climatology
.
Mon. Wea. Rev.
,
116
,
1032
1043
.
Wilks
,
D. S.
,
2006
: Statistical Methods in the Atmospheric Sciences. 2nd ed. Academic Press, 627 pp.

Footnotes

1

The 250–150-hPa layer is the layer in which negative PV advection by the irrotational wind is typically maximized in association with recurving WNP TCs.

2

This domain size is selected to match the approximate spatial scale of the field of negative PV advection by the irrotational wind typically associated with recurving TCs. The values of the TC–extratropical flow interaction metric are relatively insensitive to domain size.

3

Schenkel and Hart (2012) note that reanalyses are better able to resolve TCs at higher latitudes than near the equator, which they speculate may be attributable in part to the increased horizontal resolution of reanalyses with latitude and to the wind field expansion of TCs undergoing ET (e.g., Evans and Hart 2008).

4

With the exception of the irrotational wind, which is unsmoothed, the CFSR analysis fields in Fig. 5b are smoothed to facilitate comparison to Fig. 5a without considering differences in resolution.

5

Western North Pacific TCs that fail to complete ET after recurving (non-ET TCs) tend to be less intense at recurvature, featuring a mean MSLP and MSW of 982 hPa and 27.7 m s−1 (53.9 kt), respectively. These means are significantly different from the respective MSLP and MSW means for recurving TCs that complete ET at the 99.9% confidence level based on a two-sided Student's t test.

6

The finding that the majority (67%) of January recurving TCs reintensify as ECs is likely an artifact of a small sample size.