• Chan, J. C.-L., 1985: Tropical cyclone activity in the northwest Pacific in relation to the El Niño/Southern Oscillation phenomenon. Mon. Wea. Rev.,113, 599–606.

  • Chen, T.-C., and M.-C. Yen, 1993: Interannual variation of summertime stationary eddies. J. Climate,6, 2263–2277.

  • ——, and S.-P. Weng, 1998: Interannual variation of the summer synoptic-scale disturbance activity in the western tropical Pacific. Mon. Wea. Rev., in press.

  • Dong, K.-Q., 1988: El Niño and tropical cyclone frequency in the Australian region and the northwest Pacific. Aust. Meteor. Mag.,28, 219–225.

  • Frank, W. M., 1987: Tropical cyclone formation. A Global View of Tropical Cyclones, Office of Naval Research. 53–90.

  • Gray, W. M., 1968: Global view of the origin of tropical disturbances and storms. Mon. Wea. Rev.,96, 669–700.

  • ——, 1979: Hurricanes: Their formation, structure and likely role in the tropical circulation. Meteorology over the Tropical Oceans, D. B. Shaw, Ed., Roy. Meteor. Soc., 155–218.

  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-year reanalysis project. Bull. Amer. Meteor. Soc.,77, 437–471.

  • Lander, M. A., 1993: Comments on “A GCM simulation of the relationship between tropical storm formation and ENSO.” Mon. Wea. Rev.,121, 2137–2143.

  • ——, 1994: An exploratory analysis of the relationship between tropical storm formation in the western North Pacific and ENSO. Mon. Wea. Rev.,122, 636–651.

  • Lau, K.-M., and L. Peng, 1992: Dynamics of atmospheric teleconnections during the northern summer. J. Climate,5, 140–158.

  • Lighthill, J., G. Holland, W. Gray, C., Landsea, G. Craig, J. Evans, Y. Kurihara, and C. Guard, 1994: Global climate change and tropical cyclones. Bull. Amer. Meteor. Soc.,75, 2147–2157.

  • McBride, J. L., 1995: Tropical cyclone formation. Global Perspective on Tropical Cyclones, WMO/TD-No. 693, World Meteorological Organization, 63–105.

  • Neumann, C. J., 1993: Global overview. Global Guide to Tropical Cyclone Forecasting, World Meteor. Org., 1.1–1.56.

  • Reynolds, R. W., 1988: A real-time global sea surface temperature analysis. J. Climate,1, 75–86.

  • ——, and D. C. Marsico, 1993: An improved real-time global sea surface temperature analysis. J. Climate,6, 114–119.

  • Sadler, J. C., 1967: On the origin of tropical vortices. Proc. Working Panel on Tropical Dynamic Meteorology, Norfolk, VA, Naval Weather Research Facility, 39–75.

  • Wright, P. B., 1985: The Southern Oscillation: An ocean–atmosphere feedback system. Bull. Amer. Meteor. Soc.,66, 398–412.

  • Wu, G., and N.-C. Lau, 1992: A GCM simulation of the relationship between tropical-storm formation and ENSO. Mon. Wea. Rev.,120, 958–977.

  • View in gallery

    Time series of (a) the summer-mean sea surface temperature (SST) departure, ΔSST, over the NINO3 region and (b) the summer-mean 850-mb streamfunction departure of ΔΨ(850 mb) averaged over a part of region (1 + 3) (15°–30°N, 120°–160°E) from their corresponding multiple summer-mean values. Letters of W (C) indicate summers with ΔSST(NINO3) ≥ 0.5°C (⩽−0.5°C).

  • View in gallery

    Tropical cyclone genesis frequency of every summer (June–August) during 1979–94 over (a) total domain (0°–30°N, 120°E–180°), (b) region (1 + 3) (15°–30°N, 120°E–180°), (c) region (2 + 4) (0°–15°N, 120°E–180°), and (d) the difference of tropical cyclone genesis frequency between region (2 + 4) and (1 + 3). Each region of the analysis domain is marked in Figs. 6 and 7. The tropical cyclone genesis frequency of each warm (cold) summer is heavily (lightly) stippled.

  • View in gallery

    Tropical cyclone genesis frequency of every summer during 1979–94 over (a) region (1 + 2) (0°–30°N, 120°–150°E), (b) region (3 + 4) (0°–30°N, 150°E–180°), (c) the difference of tropical cyclone genesis frequency between regions (1 + 2) and (3 + 4), and (d) the longitudinal extent of the monsoon trough. The tropical cyclone genesis frequencies of warm (cold) summers are heavily (lightly) stippled in histograms of (a)–(c). The east end of the monsoon trough located east (west) of 150°E is marked by a heavily stippled (doubly hatched) block in (d).

  • View in gallery

    Same as Fig. 1a except for fall. Letters of W (C) indicate falls with ΔSST(NINO3) > 0.5°C (⩽−0.5°C) over 1979–94.

  • View in gallery

    Same as Fig. 3 except for fall.

  • View in gallery

    The 850-mb streamline charts superimposed with outgoing longwave radiation (OLR) for (a) summer climatology averaged over 1979–94, (c) summer climatology averaged over warm summers (1982, 1983, 1987, 1991), (d) summer climatology averaged over six cold summers (1981, 1984, 1985, 1988, 1989, 1994), and (b) the difference of the 850-mb flow between Figs. 2d and 2c. The solid black dots represent the locations of tropical cyclones identified at the beginnings of their life cycles. Note that areas of OLR ⩽ 210 W m−2 and ΔOLR ⩽ −8 W m−2 are heavily stippled, but areas of 210 W m−2 ⩽ OLR ⩽ 240 W m−2 and ΔOLR ≥ 8 W m−2 are lightly stippled.

  • View in gallery

    Latitudinal locations of the monsoon trough (solid line) and the minimum OLR value (dashed line) averaged over 130°–140°E. The two latitudes marked on the right ordinate are the averaged latitudinal locations of the monsoon trough and minimum OLR value within the longitudinal zone (130°–140°E).

  • View in gallery

    Same as Fig. 5 except for fall.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 224 224 16
PDF Downloads 160 160 12

Interannual Variation in the Tropical Cyclone Formation over the Western North Pacific

View More View Less
  • 1 Atmospheric Science Program, Department of Geological and Atmospheric Sciences, Iowa State University, Ames, Iowa
  • | 2 Typhoon Research Department, Meteorological Research Institute, Tsukuba, Japan
  • | 3 Atmospheric Science Program, Department of Geological and Atmospheric Sciences, Iowa State University, Ames, Iowa
© Get Permissions
Full access

Abstract

The interannual variation in tropical cyclone genesis frequency over the western North Pacific was examined for the active tropical cyclone (including summer and fall) during 1979–94. An emphasis was put on the possible effect of the interannual variation of atmospheric circulation and monsoon trough on tropical cyclone occurrence. The major findings of this study are the following.

  1. A distinct increase (decrease) of tropical cyclone genesis frequency occurs north of the climatological location of the monsoon trough in the Philippine Sea during summers (June–August) with anomalous cold (warm) sea surface temperature (SST) over the NINO3 region. The interannual variation of tropical cyclone genesis in this region results from the appearance of an anomalous cyclonic (anticyclonic) cell situated in a summer teleconnection wave train emanating from the western tropical Pacific and progressing along the rim of the North Pacific. In addition to the north–south interannual variation, there is also a longitudinal interannual variation in the summer tropical cyclone genesis frequency over this region. The contrast of tropical cyclone genesis between the regions west and east of 150°E is reduced (enhanced) when the monsoon trough extends (retreats) eastward (westward) across this longitude during warm (cold) summers.
  2. For fall (September–November), there is no clear relationship between the north–south interannual variation in the tropical cyclone genesis over the western North Pacific and SST (NINO3). However, there is a perceptible tendency of the longitudinal interannual variation in tropical cyclone genesis frequency to follow the eastward extension/westward retreat of the monsoon trough in a way such as it does during the summer season.

Corresponding author address: Tsing-Chang Chen, Atmospheric Science Program, 3010 Agronomy Hall, Department of Geological and Atmospheric Sciences, Iowa State University, Ames, IA 50011.

Email: zntcc@climate1.agron.iastate.edu

Abstract

The interannual variation in tropical cyclone genesis frequency over the western North Pacific was examined for the active tropical cyclone (including summer and fall) during 1979–94. An emphasis was put on the possible effect of the interannual variation of atmospheric circulation and monsoon trough on tropical cyclone occurrence. The major findings of this study are the following.

  1. A distinct increase (decrease) of tropical cyclone genesis frequency occurs north of the climatological location of the monsoon trough in the Philippine Sea during summers (June–August) with anomalous cold (warm) sea surface temperature (SST) over the NINO3 region. The interannual variation of tropical cyclone genesis in this region results from the appearance of an anomalous cyclonic (anticyclonic) cell situated in a summer teleconnection wave train emanating from the western tropical Pacific and progressing along the rim of the North Pacific. In addition to the north–south interannual variation, there is also a longitudinal interannual variation in the summer tropical cyclone genesis frequency over this region. The contrast of tropical cyclone genesis between the regions west and east of 150°E is reduced (enhanced) when the monsoon trough extends (retreats) eastward (westward) across this longitude during warm (cold) summers.
  2. For fall (September–November), there is no clear relationship between the north–south interannual variation in the tropical cyclone genesis over the western North Pacific and SST (NINO3). However, there is a perceptible tendency of the longitudinal interannual variation in tropical cyclone genesis frequency to follow the eastward extension/westward retreat of the monsoon trough in a way such as it does during the summer season.

Corresponding author address: Tsing-Chang Chen, Atmospheric Science Program, 3010 Agronomy Hall, Department of Geological and Atmospheric Sciences, Iowa State University, Ames, IA 50011.

Email: zntcc@climate1.agron.iastate.edu

1. Introduction

The interannual variation of tropical cyclone genesis in the western North Pacific has been the captivating subject of several previous studies (e.g., Chan 1985; Dong 1988; Wu and Lau 1992; Lander 1994). The consensus reached by these studies as to the cause is an interannual oscillation in the frequency of tropical cyclone genesis between the western and central North Pacific. During El Niño years, fewer tropical cyclones were formed over west of around 150°–160°E, while more were generated east of these longitudes. The reversed situation occurred during La Niña years. Chan(1985) suggested that the frequency contrast of tropical cyclone genesis between the two regions may be caused by the development of an anomalous east–west Walker circulation, which would lead to an interannual alternation of enhancement and suppression of cumulus convection between the western and central North Pacific. This east–west alternation of cumulus convection may affect the interannual variation of tropical cyclone genesis. Chan’s suggestion was reinforced by Wu and Lau (1992) who linked the interannual variation in tropical cyclone genesis frequencies to that in low-level divergence between the western and central tropical Pacific.

In his recent review of tropical cyclone genesis, McBride (1995) pointed out that the seasonal distribution of tropical cyclone genesis location was determined primarily by two factors: sea surface temperatures (SST) greater than 26.5°C and the location of the monsoon trough. Analyzing the SST compiled by the National Centers of Environmental Prediction (NCEP)(Reynolds 1988; Reynolds and Marsico 1993), we found that during the active tropical cyclone season the interannual variation of SSTs in the northwestern Pacific tropical cyclone genesis region is not significant in the distribution of tropical cyclone genesis location. Thus, the second factor may be of importance to the interannual variation in tropical cyclone genesis frequencies, because “most cyclones form in the shear zone between monsoonal (usually cross-equatorial) westerlies and the trade easterlies (Sadler 1967). Enhancement of the flow on either side of this monsoon trough increases the low-level relative vorticity and makes conditions more favorable for genesis” (Frank 1987). During the active months of tropical cyclone genesis, the monsoon trough always exists in the western tropical Pacific (McBride 1995). As inferred from Frank’s observation, it is likely that any mechanism causing the interannual variation of the atmospheric circulation around the monsoon trough may affect the tropical cyclone genesis frequencies over the western North Pacific (Lighthill et al. 1994).

As shown by the low-level flow in the western North Pacific (which will be shown later in this paper), the monsoon trough during the active tropical cyclone seasons (which include summer and fall) is situated between the North Pacific anticyclone and the Borneo–New Guinea high. Because these two high systems are asymmetric components of the summer–fall atmospheric circulation, the interannual variation of the monsoon trough should be linked to that of these two asymmetric circulation elements. Although the interannual variation of summer circulation has been scrutinized by recent studies (e.g., Lau and Pang 1992; Chen and Yen 1993), the major focus of these studies was either the global-scale circulation in the Tropics or the link between the downstream teleconnectivity and the drought/flood of North America. Actually, the interannual variation of summer circulation in the western North Pacific has not been extensively examined. An El Niño–Southern Oscillation (ENSO) episode generally emerges in fall and reaches its peak intensity in winter. The depiction of the anomalous circulation associated with an ENSO event is often focused on the winter season, while the development of the anomalous circulation during fall is often neglected. Because of the insufficient research on the interannual variation in the atmospheric circulation during summer and fall, we may ask that if tropical cyclone genesis occurs frequently in the western North Pacific, is there a possible interannual variation of the atmospheric circulation over the western North Pacific that would result in interannual variation of the monsoon trough, and in turn interannual variation of tropical cyclone genesis frequencies? We endeavor in this study to search for this possibility.

The data used in this study included the 6-h tropical cyclone tracks of a 16-yr period (1979–94) collected by the Regional Specialized Meteorological Center within the Forecast Division of the Japan Meteorological Agency, Tokyo, Japan, and outgoing longwave radiation(OLR) and SST compiled by NCEP and the NCEP–National Center for Atmospheric Research (NCAR) reanalysis data for 1979–94.

2. Interannual variation of tropical cyclone occurrence

Various statistical schemes have been applied in previous studies in search of the possible relationship between the interannual variation in the tropical cyclone genesis1 frequency over the western North Pacific and the ENSO activity. However, for simplicity of statistical analysis, we shall examine in this study this relationship in terms of basic statistics. The peak tropical cyclone season over the western North Pacific covers summer and fall (Neumann 1993). Thus, we shall focus our analysis only on the time period ranging from June to November for the years examined, rather than the annual tropical cyclone occurrence as done by other studies. The statistical analysis of this study consists of the following two parts.

  1. Division of analysis domain: Tropical cyclone frequencies east of the date line are much lower than west of this longitude (Fig. 4 of Lander 1994). Also, as indicated by the xt diagram of OLR (not shown), interannual variation of cumulus convection inside the South China Sea is opposite in phase to that in the central tropical Pacific. For these two reasons, we concentrate our analysis on the region of 0°–30°N, 120°E–180°. As will be shown later, the main portion of the seasonal-mean monsoon trough east of the Philippines extends eastward in summer to about 150°E and farther eastward in fall. This trough is located latitudinally during both seasons at about 10°–15°N within the longitudinal zone of 130°–140°E. In order to obtain a sense of the spatial variation in the interannual variation of tropical cyclone genesis frequencies, the aforementioned analysis domain is divided into four subregions: region 1 (15°–30°N, 120°–150°E), region 2 (0°–15°N, 120°–150°E), region 3 (15°–30°N, 150°E–180°), and region 4 (0°–15°N, 150°E–180°). This domain division is presented in some synoptic charts in the next section.
  2. Seasonal-mean tropical cyclone occurrence: The major theme of this study is to seek the effect of a possible interannual variation of atmospheric circulation over the western North Pacific on the tropical cyclone genesis frequencies in this region. As is well known, the most pronounced interannual variation of atmospheric circulation is perhaps associated with the ENSO activity. For the following reasons, however, we shall split the active tropical cyclone period analyzed in this study into summer (June–August) and fall (September–November) to illustrate the effect sought in this study.
    • It was pointed out by Wright (1985) that some characteristics of the Southern Oscillation (SO) exhibit a distinct seasonal variation. The maximum development of the SO characteristics occurs during September–February and the minimum around April–June. Thus, the fall characteristics of the SO may differ from those in the summer. In contrast, spring (March–May) is a transition season during which the ENSO may switch either from a warm winter phase to a cold summer phase or vice versa.
    • The South China Sea–western tropical Pacific summer monsoon decays after August. The monsoon trough, which radiates out during summer from northern Indochina to the Philippine Sea, is detached from Indochina in fall, because of the development of a minor high over East Asia and the equatorward migration of its western portion in the South China Sea. Thus, the monsoon trough in fall becomes more east–west oriented.

a. Summer

Shown in Fig. 1a is the time series of the summer-mean SST departure in the NINO3 region, ΔSST(NINO3), from its multiple-summer average [generated from the SST compiled by Reynolds (1988) and Reynolds and Marsico (1993) at NCEP]. Summers with ΔSST(NINO3) larger (smaller) than 0.5°C (−0.5°C) are marked with W (C). If the interannual variation of the summer circulation containing the monsoon trough is induced by (or related to) the tropical summertime SST anomalies, there should be a correlation between them. It will be shown later in the next section that a northward (southward) migration of the summer-mean monsoon trough follows the appearance of anomalous cyclonic (anticyclonic) flow in region (1 + 3) during cold (warm) summer. Since the atmospheric circulation in both the Tropics and higher latitudes can be portrayed well by streamfunction Ψ, we shall apply the summer-mean streamfunction departure at 850 mb, ΔΨ(850 mb), averaged over a major part of region (1 + 3) from its multiple-summer mean as a quantitative index to indicate the anomalous cyclonic [ΔΨ(850 mb) < 0] and anticyclonic [ΔΨ(850 mb) > 0] condition in this region. Streamfunction can be obtained by solving the Poisson equation with the global vorticity field. Following the ΔSST(NINO3) time series, we also mark W and C on the ΔΨ(850 mb) time series in Fig. 1b. As we can see, interannual variation of both indices are relatively coincident. Thus, we adopt the following criteria to define

  1. cold summer
    • ΔSST(NINO3) ⩽ −0.5°C and,
    • ΔΨ(850 mb) < 0 (anomalous cyclonic flow) in region (1 + 3) accompanying a northward migration of the monsoon trough
  2. warm summer
    • ΔSST (NINO3) ≥ 0.5°C
    • ΔΨ(850 mb) > 0 (anomalous anticyclonic flow) in region (1 + 3) accompanying a southward migration of the monsoon trough.
The summers satisfying marginally the ΔSST (NINO3) criterion may be dropped out from our selection if more summers are included to evaluate the multiple-summer averaged SST(NINO3). However, in order to avoid any possible obscurity in our analysis caused by the improper selection of summers, we introduce the ΔΨ(850 mb) criterion that those summers selected exhibit a similar response of their circulation to ΔSST(NINO3) anomalies in the eastern tropical Pacific. Based upon the above two criteria, summers of (1982, 1983, 1987, 1991) and (1981, 1984, 1985, 1988, 1989, 1994) are selected as warm and cold ones, respectively.

The histogram of tropical cyclone genesis frequency occurring every summer in the analysis domain is displayed in Fig. 2a. The contrast between ΔSST(NINO3) and the tropical cyclone genesis frequencies reveals an increasing (decreasing) tendency of tropical cyclone genesis during cold (warm) summers. A sense of spatial variation in the interannual variation of tropical cyclone genesis frequencies within the analysis domain can be obtained from the statistics shown in Table 1. Displayedin this table are the multiple summer-mean tropical cyclone genesis frequencies (n), the seasonal tropical cyclone genesis frequencies averaged over six cold summers (nc), the seasonal tropical cyclone genesis frequencies averaged over four warm summers (nw), the difference ncnw, and the total number of tropical cyclone occurrences over 16 summers (Σ n), respectively. The major results of this table are highlighted as follows.

  1. For the total domain, the difference between summer tropical cyclone genesis frequencies averaged over six cold and four warm summers, ncnw ≃ 5, is about four-tenths of the averaged summer genesis frequencies, (n ≃ 12).
  2. The total number of tropical cyclones formed Σ n in region 1 is comparable to that formed in region (1 + 3), and likewise, Σ n in region 2 is comparable to Σ n in region (2 + 4). This comparison indicates that the total tropical cyclone genesis frequencies in region 3 and region 4 are much smaller than in region 1 and region 2. Thus, in the latter two regions, the contrast of tropical cyclone genesis between cold and warm summers may be a good indication for the impact of the summer anomalous SST in the tropical Pacific and the corresponding anomalous circulation on tropical cyclone genesis in the western North Pacific.
  3. Although the summer-mean tropical cyclone genesis frequencies in both regions (1 + 3) and (2 + 4) are comparable, the difference of tropical cyclone genesis frequency between warm and cold summers behaves in these two regions in opposite ways. It is revealed from the contrast between nw and nc that the interannual variation of tropical cyclone genesis frequency in region (1 + 3) is much larger than that in region (2 + 4) and is in an opposite way. Quantitatively, ncnw is larger than n; cold summers have more tropical cyclone genesis in region (1 + 3). In contrast, ncnw in region (2 + 4) is negative and smaller than n; cold summer have fewer tropical cyclone geneses.
As indicated by the comparisons between Figs. 2b and 2c, the contrast of interannual variation in tropical cyclone genesis frequency between regions (1 + 3) and (2 + 4) is not only pronounced, but also persistent over all cold and warm summers. To be more quantitative,n(region 1 + 3) − n(region 2 + 4) of every summer is shown in the bottom histogram of Fig. 2: |n(region 1 + 3) − n(region 2 + 4)| for all cold and warm summers analyzed in this study is larger than/equal to 3.

Previous studies examining the interannual variation of tropical cyclone genesis frequencies in the western North Pacific region focused primarily on the interannual oscillation between the western and eastern tropical Pacific (e.g., Chan 1985; Wu and Lau 1992; Lander 1994). The interannual oscillation of anomalous east–west Walker circulation between the western and eastern tropical Pacific in accordance with the ENSO activity has been suggested as a possible forcing mechanism causing the aforementioned interannual oscillation of tropical cyclone genesis frequencies. As illustrated by Table 1, a pronounced contrast exists between interannual variations in regions north and south of the climatological monsoon trough. Lander (1994, his Fig. 5) observed annual total significant changes in the latitude of tropical cyclone formation. However, the contribution from the pronounced contrast between interannual variations in tropical cyclone genesis frequencies north and south of the climatological location of the monsoon trough has not been recognized and analyzed in past studies. Thus, some questions concerning this interannual variation are raised:

  1. Why does a pronounced interannual variation in tropical cyclone genesis frequencies occur north of the climatological monsoon trough during the summer?
  2. How is this interannual variation in tropical cyclone formation related to the interannual variation of the monsoon trough east of the Philippines?
  3. Is the interannual variation of the monsoon trough linked to that of the large-scale circulation in the North Pacific basin?
To answer these questions, we shall examine interannual variations of the following summer circulation elements in section 3: 1) latitudinal location of the monsoon trough east of the Philippines, and 2) summertime circulation in the North Pacific basin and its relationship with the monsoon trough.

In addition to the effect of the monsoon trough’s meridional migration, Lander (1994) pointed out that the interannual variation of tropical cyclone genesis may also be affected by the longitudinal variation of the monsoon trough. Although the majority of tropical cyclonegenesis takes place in regions 1 and 2, it is inferred from Table 1 that approximately a quarter of tropical cyclone genesis occurred in region (3 + 4) during the summers of 1979–94. It is likely that the effect of the monsoon trough’s east–west extension on tropical cyclone genesis can be revealed from the contrast of tropical cyclone genesis between regions (1 + 2) and (3 + 4). It is shown in Table 2 that this contrast increases (decreases) during cold (warm) summers. To clarify further the effect of the monsoon trough’s longitudinal variation on tropical cyclone genesis frequency, the histograms of three quantities in each summer are shown in Fig. 3: tropical cyclone genesis frequency in regions (1 + 2) [n(region 1 + 2)] and (3 + 4) [n(region 3 + 4)], and the difference between them Δn [=n(region 1 + 2) − n(region 3 + 4)]. As judged subjectively from the summer-mean 850-mb streamline charts in every summer (not shown), dark shading (double hatching) is used in Fig. 3d when the monsoon trough extends (retreats) eastward (westward) across 150°E. The contrast between Figs. 3c and 3d provides a strong indication that Δn increases noticeably during cold summers in accompaniment with a westward retreat of the monsoon trough as Lander (1994) suggested.

b. Fall

As shown previously, the interannual variation of the summer tropical cyclone genesis frequency in the western North Pacific may be caused by the north–south meridional migration and by the longitudinal variation of the monsoon trough. Because the peak tropical cyclone season includes fall and the monsoon trough still exists in this season over the Philippine Sea, we may question whether there is interannual variation of the fall circulation over the western North Pacific to affect the tropical cyclone genesis frequency in this region in the same way as in summer.

It was revealed in Fig. 1 that a coherent interannual variation exists between the ΔSST(NINO3) and ΔΨ(850 mb) indices in summer, and from Fig. 2 that the tropical cyclone genesis frequency has an increasing tendency during cold summers. Shown in Fig. 4 is the time series of the fall ΔSST(NINO3) index. Following the summer criterion, we may declare those falls with ΔSST(NINO3) ≥0.5°C (⩽−0.5°C) as warm (cold) and mark them W (C) accordingly on the fall ΔSST(NINO3) index. Unlike the summer season, there is no coherent interannual variation in fall between the ΔSST(NINO3) and ΔΨ(850 mb) (not shown) indices. In other words, there is no systematic correlation between interannual variations of ΔSST(NINO3) and the atmospheric circulation in region (1 + 3). Actually, we do not find any indication that the fall tropical cyclone genesis frequency over the western North Pacific shows any interannual variation in accordance with SST(NINO3). The monsoon trough neither undergoes a north–southmeridional migration nor affects the contrast of tropical cyclone genesis frequency between regions (1 + 3) and (2 + 4) following the interannual variation of ΔSST(NINO3) index.

Since there is no systematic meridional migration of the fall monsoon trough, can this monsoon trough undergo interannually a longitudinal variation to cause the interannual variation of tropical cyclone genesis frequency? To explore the answer to the question posed here, we display in Table 3 n(region 1 + 2), n(region 3 + 4), and the difference between them, Δn. Although n(region 3 + 4) is less than a half of n(region 1 + 2), Δn shows a clear increase (decrease) during cold (warm) summers. This coherent relationship may be explored further with histograms of these three quantities in Fig. 5. The east–west extension of the fall monsoon trough across 150°E is also marked at the bottom of this figure, following the convention of Fig. 3. As revealed from Fig. 5, Δn and ΔSST(NINO3) are out of phase. Plausibly, it is expected that the interannual variation of Δn is related to the interannual variation of the atmospheric circulation in connection with the east–west extension of the fall monsoon trough in response to the interannual variation of fall ΔSST(NINO3). This issue will be addressed in further detail later.

Our assertion of the relationship between Δn and ΔSST(NINO3) in fall is complicated by large Δn during falls of 1979, 1983, and 1993. The reason for this unusual enhancement of tropical cyclone genesis frequency in Δn, and in turn region (1 + 2), is unclear to us.

3. Interannual variation of summer circulation

a. Summer

Shown in Fig. 6a is the summer-mean 850-mb streamline chart constructed with the NCEP–NCAR reanalysis data (Kalnay et al. 1996) for 1979–94, superimposed by locations of tropical cyclone genesis (denoted by dots), OLR, and regions used in our analysis (encircled by boxes). As indicated by the low-value OLR (210 W m−2) and cyclonic streamlines, the monsoon trough east of the Philippines is situated between the North Pacific anticyclone and the Borneo–New Guinea high with monsoon westerlies and trade easterlies located south and north of this trough, respectively. The first of Gray’s (1968, 1979) six environmental factors determining thedistribution of tropical cyclone genesis is a large value of low-level relative vorticity. Due to the strong horizontal shear (which reflects, in large part, relative vorticity) associated with the monsoon trough, the western North Pacific is a major region of tropical cyclone genesis. Since regions (1 + 3) (Fig. 2) and (1 + 2) (Fig. 3) exhibit more tropical cyclone genesis during cold summers than warm summers, it is likely that the monsoon trough embedded in the large-scale summer circulation may undergo some type of interannual variation between these two climate regimes. To explore this possibility, composite charts of 850-mb streamline for four warm (1982, 1983, 1987, 1991) and six cold (1981, 1984, 1985, 1988, 1989, 1994) summers are displayed in Figs. 6c and 6d, respectively. Superimposed on these two composite charts are the locations of tropical cyclone genesis for corresponding summers and OLR.

Some striking contrasts of tropical cyclone genesis locations between warm and cold summers emerge from the comparison between Figs. 6c and 6d: Tropical cyclone genesis occurs more often 1) north (south) of 15°N during cold (warm) summers and 2) west of 150°E during cold summers. The tropical cyclone genesis frequency in the South China Sea is not a concern of this study but it is still of interest to note that the tropical cyclone activity increases (decreases) during cold (warm) summers. As inferred from the distribution of tropical cyclone genesis around the monsoon trough and from this trough’s location, the interannual variations of summer tropical cyclone genesis over the western North Pacific may result from the interannual variation of the monsoon trough in two ways.

1) North–south migration: The monsoon trough migrates from the latitudinal location (east of the Philippines) south of 15°N during warm summers to the latitudinal location around 15°N during cold summers. To be more quantitative, the latitudinal locations of the monsoon trough between 130° and 140°E are determined with the 850-mb streamline charts and average over this longitudinal zone. The averaged monsoon trough location and the latitudinal location of minimum OLR value averaged over 130°–140°E are displayed in Fig. 7. As shown clearly in this figure, there is a northward shift of the monsoon trough and the associated minimum OLR during cold summers (1981, 1984, 1985, 1988, 1989, 1994) and a southward shift of both the monsoon trough and associated minimum OLR during warm summers (1982, 1983, 1987, 1991).

2) East–west extension: Based upon our observations of the summer-mean 850-mb streamline charts (not shown), the second type of the monsoon trough’s interannual variation, that is, the eastward extension/westward retreat across 150°E, is marked in Fig. 3d. As revealed from this figure, the monsoon trough retreats (extends) westward (eastward) during cold (warm) summers. This assertion may be summarized by the contrast between the composite summertime 850-mb streamline charts and low-valued OLR of warm (Fig. 6c) and cold (Fig. 6d) summers. Since the monsoon trough is situated between the North Pacific anticyclone and the Borneo–New Guinea high, the aforementioned north–south meridional shift and east–west extension of the monsoon trough should be a part of the interannual variation of large-scale summer circulation. In other words, the interannual variation of the monsoon trough between cold and warm summers is an indicator of the interannual variation of large-scale summer circulation in which this trough embeds.

In his comment on Wu and Lau’s (1992) study for the relationship between tropical cyclone formation frequency and the ENSO activity, Lander (1993) pointed out that “the main low-level circulation associated with suppression of formation of tropical cyclones in the western North Pacific is a persistent subtropical ridge anchored from just east of the Philippines and extending northeastward toward the date line that is accompanied by higher than normal sea level pressure and low-level easterly wind anomalies throughout the low latitudes of the western North Pacific.” Lander (1993) commentedfurther that “anomalous high pressure and easterly winds in the low latitudes of the western North Pacific are often associated with the transition to cold conditions following an El Niño.” Based on preliminary synoptic investigations, Lander’s comments indicate the possible effect of the interannual variation of large-scale summer circulation on tropical cyclone genesis frequencies. The teleconnection wave train presented in Fig. 6b and our argument concerning the possible effectof this wave train on the summer tropical cyclone genesis frequencies over the western North Pacific are consistent with Lander’s observation.

The contrast between the North Pacific anticyclone structure during cold (Fig. 6d) and warm (Fig. 6c) summers reveals that the ridge line of this anticyclone shifts northward in the former summer condition. A better view of the circulation change from warm to cold summers can be obtained through differences of 850-mb flow fields and OLR between the two extreme circulation conditions. The differences in 850-mb flow fields and OLR, depicted with Δ[V (850 mb), OLR], are shown in Fig. 6b. Prominent features of this figure are highlighted as follows.

  1. A short-wave train emanates from the Philippine Sea along the rim of the North Pacific with an anomalous anticyclonic cell in region (2 + 4) and an anomalous cyclonic cell in region (1 + 3). The former anomalous cell and the anomalous tropical easterlies associated with this anticyclonic cell are consistent with Lander’s remark. During cold summers, the east–west elongated anomalous cyclonic cell covering region (1 + 3) east of Taiwan enhances low-level relative vorticity and in turn increases tropical cyclone genesis frequencies during cold summers. Because of this cyclonic cell’s appearance, the monsoon trough retreats westward and broadens its north–south extent. For this reason, the tropical cyclone genesis frequency is enhanced in region (1 + 2). The reverse situation occurs during warm summers.
  2. In addition to the east–west differentiation of OLR anomalies in the Tropics, there is a north–south stratification of OLR anomalies associated with the juxtaposition of anomalous anticyclonic and cyclonic cells. As indicated by OLR anomalies, cumulus convection is intensified in region (1 + 3) and the northern part of region 2 during cold summers. Conceivably, tropical cyclone genesis frequencies should be increased accordingly. The opposite situation should take place in warm summers in these regions.

It becomes clear from our analysis that the increase (decrease) in tropical cyclone genesis frequencies in regions (1 + 3) and (1 + 2) results from the interannual summer circulation change shown in Fig. 6b. Apparently, there is a close link between the interannual variation of tropical cyclone genesis frequencies and the teleconnection wave train of the summertime circulation in the western North Pacific following the alternation between warm and cold summers.

b. Fall

Following the summer case (section 3a), the fall-mean 850-mb streamline chart with locations of tropical cyclone genesis, OLR and analysis domain are shown in Fig. 8a. The salient features of the fall circulation in contrast to the summer circulation are 1) along the east coast of Asia, a trough appears over the eastern seaboard of northeast Asia and an anticycloneforms over east China; 2) the west side of the North Pacific anticyclone retreats eastward to form a low pressure corridor between this anticyclone and the east China anticyclone; 3) being detached over Indochina, the western part of the monsoon trough moves equatorward in such a way that this trough becomes more east–west oriented; and 4) the Aleutian low deepens and splits into two low centers: a major one centered in eastern Alaska and a minor one in the Skelekhov Gulf.

The locations of fall tropical cyclone genesis over the 16-yr period are dotted in Fig. 8a. Compared to the summer season (Fig. 6a), the occurrence of tropical cyclone genesis in fall concentrates more along the fall-mean monsoon trough and rarely spreads across 25°N. The larger meridional spread of tropical cyclone genesis in summer (Fig. 6a) is caused by the interannual meridional migration of the monsoon trough. In contrast, the smaller meridional spread of tropical cyclone genesis is possible because of smaller meridional migration of the monsoon trough. According to the statistics of fall tropical cyclone genesis, there is a pronounced interannual variation in the difference of fall tropical cyclone genesis between regions (1 + 2) and (3 + 4), Δn, during cold falls. If the monsoon trough does not undergo a significant meridional migration, what then may be the mechanism causing the pronounced interannual variation in Δn? To seek the answer to this question, we construct the composite 850-mb streamline charts with OLR for cold (1984, 1985, 1988, 1989) and warm (1982, 1986, 1987, 1991) falls. Interesting features revealed from these two composite charts are as follows.

  1. The contrast of the fall 850-mb circulation in the western tropical Pacific between cold and warm does not show significant meridional shift of the monsoon trough. The major change of the 850-mb circulation between the two climate regimes is the disappearance (eastward extension) of the monsoon trough in (to) region 4. The increase of tropical cyclone genesis frequency in region (1 + 2) results from the deepening of the monsoon trough in this region. As revealed from feature 1, this minor meridional shift of the monsoon trough does not seem to exert a profound effect on the locations of tropical cyclone genesis as in the summer case.
  2. Other 850-mb circulation contrasts between the cold and warm falls include the slight eastward shift of the east China anticyclone during cold falls, the deepening of the east coast trough of northeast Asia and the eastward shift of the western center of the North Pacific anticyclone during cold falls, and the filling of the two low centers at about 60°N during cold falls.

The contrast of tropical cyclone genesis locations and 850-mb circulation between the two different circulation regimes suggests that the pronounced difference in tropical cyclone genesis frequency between regions (1 + 2) and (3 + 4) during cold falls results from the circulation pattern change over the analysis domain. To substantiate this suggestion, the differences in composite streamline charts and OLR between the two circulation regimes, that is, Figs. 8d and 8c, are shown in Fig. 8b. The salient features of Fig. 8b related to the interannual variation of the fall tropical cyclone genesis frequency between regions (1 + 2) and (3 + 4) are as follows.

  1. A well-organized teleconnection wave train emanating from the analysis domain emerges from the circulation difference between cold and warm falls. Regardless of its structure difference from the summer one (Fig. 6b), the fall teleconnection wave train may well be a response of the fall atmospheric circulation to the tropical Pacific SST anomalies, as indicated by ΔSST(NINO3).
  2. The anomalous circulation in each region of the analysis domain is characterized by one type of anomalous flow: Regions 1 and 4 by anomalous anticyclonic flow and regions 2 and 3 by anomalous cyclonic flow. The enhancement of tropical cyclone genesis frequency in region (1 + 2) is caused primarily by the occurrence of anomalous cyclonic flow in Region 2 and is related to the deepening of the monsoon trough there.
Since the anomalous flows in regions 1, 2, and 3 are parts of the fall teleconnection wave train, it is likely that the effect of the tropical Pacific SST anomalies [as indicated by ΔSST(NINO3)] on the fall tropical cyclone genesis over the western North Pacific occurs through the generation of this wave train.

4. Concluding remarks

Previous studies examining the interannual variation of tropical cyclone genesis frequencies in the western North Pacific focused primarily on the east–west oscillation of this frequency between the western and central North Pacific. In accordance with the ENSO activity, the east–west oscillation of the anomalous Walker circulation between these two territories of the North Pacific was suggested as a possible cause of the interannual variation of tropical cyclone genesis frequencies. According to Gray (1968, 1979), the relative vorticity associated with the monsoon trough in the western North Pacific is a factor vital to the tropical cyclone formation in this region. Conceivably, any mechanism possibly causing the interannual variation of the atmospheric circulation in which the monsoon trough embeds may result in the interannual variation of the relative vorticity associated with this trough. In turn, this mechanism may affect interannually the tropical cyclone genesis frequency over the region surrounding the monsoon trough. The most likely candidate for this mechanism is the interannual variation component of the atmospheric circulation containing the monsoon trough related to the ENSO activity. Based upon this hypothesis, an effort was made in this study to explore 1) the interannual variation of tropical cyclone genesis frequency in the western North Pacific and 2) the possible mechanism causing this interannual variation.

The active tropical cyclone season covers summer and fall. Since the characteristics of atmospheric circulation and Southern Oscillation in these two seasons differ in some ways, we split our analysis into summer and fall. Our major findings for the interannual variation of tropical cyclone genesis frequency and the monsoon trough are summarized as follows.

  1. For summer, the monsoon trough undergoes a north–south migration as well as a longitudinal variation: the monsoon trough exhibits a northward (southward) migration across 15°N (the climatological latitudinal location of the monsoon trough) and westward retreat [eastward extension across 150°E (the east end of the climatological monsoon trough)] during cold [ΔSST(NINO3) ⩽ −0.5°C] {warm [ΔSST(NINO3) ≥ 0.5°C]} summers. These interannual variations in the monsoon trough result in the enhancement (reduction) of tropical cyclone genesis frequency north of 15°N and west of 150°E during cold (warm) summers.
  2. During fall, the South China Sea–western tropical Pacific monsoon diminishes, but the monsoon trough still exists between 10° and 15°N. Unlike the summer season, the fall monsoon trough does not show a pronounced north–south migration. However, the monsoon trough undergoes a longitudinal variation in accordance with the interannual variation of ΔSST(NINO3); the monsoon trough extends (retreats) eastward (westward) across 150°E during warm[ΔSST(NINO3) ≥ 0.5°C] {cold [ΔSST(NINO3) ⩽ −0.5°C]} falls and the tropical cyclone genesis frequency west of 150°E is accordingly suppressed (enhanced) in warm (cold) falls.

What may be the possible mechanism responsible for the interannual variation of tropical cyclone genesis frequency over the western North Pacific found in this study? Since the anomalous circulation associated with the ENSO activity may be responsible for the interannual variation of the monsoon trough, we thus constructed the difference in streamline charts between cold and warm summers/falls. Emerging from these streamline charts at 850 mb is a teleconnection wave train emanating from our analysis domain. The interannual variation of the monsoon trough, which causes the interannual variation of tropical cyclone genesis over the western North Pacific, is a part of the teleconnection wave train. The anomalous teleconnection wave train is a possible response of the summer–fall atmospheric circulation to the tropical Pacific SST anomalies, as indicated by ΔSST(NINO3). Thus, we may argue that the interannual variation of tropical cyclone genesis frequency over the western North Pacific is caused by the tropical Pacific SST anomalies through the formation of the anomalous teleconnection wave train. In this study, as we analyzed only 16 years, which include four warm and six cold summers, and four warm and four cold falls, the conclusion drawn from our result should be considered tentative. An effort along the line of this study is urged when the 40-yr NCEP–NCAR reanalysis data become available in the near future.

The teleconnection wave train of the North Pacific in summer and fall have not been well examined in the past. The effect of this teleconnection wave train on the tropical cyclone genesis frequency in the western North Pacific is certainly an intriguing topic. However, the possible effect of this wave train on the activity of weather disturbances and monsoon in Southeast–East Asia and the North American climate, for example, drought and flood, are important subjects that should be addressed. Although some studies (e.g., Lau and Peng 1992) examined the summer teleconnectivity, their main focus was its downstream effect on the interannual summer climate change of North America, rather than the upstream effect on the circulation over the western North Pacific. Recently, Chen and Weng (1998) examined the relationship between the interannual variation of the equatorial wave activity in the western tropical Pacific and the summer teleconnection wave train. How this wave train affects the Southeast–East Asian monsoon is unclear. In light of our depiction of the summer/fall teleconnection wave train, the teleconnectivity of large-scale circulation in the North Pacific basin warrants further investigation.

Acknowledgments

This study is supported by the NSF Grant ATM-9416954. We thank Dr. M. Murakami (the Typhoon Research Department of Meteorological Research Institute at Tsukuba, Japan) for his assistance in accessing the tropical cyclone track data. Comments made by Dr. Chris Landsea and an anonymous reviewer were helpful in clarifying some issues discussed in this paper. The typing support by Mrs. Reatha Diedrichs and editorial assistance by Mrs. Susan Carr are highly appreciated.

REFERENCES

  • Chan, J. C.-L., 1985: Tropical cyclone activity in the northwest Pacific in relation to the El Niño/Southern Oscillation phenomenon. Mon. Wea. Rev.,113, 599–606.

  • Chen, T.-C., and M.-C. Yen, 1993: Interannual variation of summertime stationary eddies. J. Climate,6, 2263–2277.

  • ——, and S.-P. Weng, 1998: Interannual variation of the summer synoptic-scale disturbance activity in the western tropical Pacific. Mon. Wea. Rev., in press.

  • Dong, K.-Q., 1988: El Niño and tropical cyclone frequency in the Australian region and the northwest Pacific. Aust. Meteor. Mag.,28, 219–225.

  • Frank, W. M., 1987: Tropical cyclone formation. A Global View of Tropical Cyclones, Office of Naval Research. 53–90.

  • Gray, W. M., 1968: Global view of the origin of tropical disturbances and storms. Mon. Wea. Rev.,96, 669–700.

  • ——, 1979: Hurricanes: Their formation, structure and likely role in the tropical circulation. Meteorology over the Tropical Oceans, D. B. Shaw, Ed., Roy. Meteor. Soc., 155–218.

  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-year reanalysis project. Bull. Amer. Meteor. Soc.,77, 437–471.

  • Lander, M. A., 1993: Comments on “A GCM simulation of the relationship between tropical storm formation and ENSO.” Mon. Wea. Rev.,121, 2137–2143.

  • ——, 1994: An exploratory analysis of the relationship between tropical storm formation in the western North Pacific and ENSO. Mon. Wea. Rev.,122, 636–651.

  • Lau, K.-M., and L. Peng, 1992: Dynamics of atmospheric teleconnections during the northern summer. J. Climate,5, 140–158.

  • Lighthill, J., G. Holland, W. Gray, C., Landsea, G. Craig, J. Evans, Y. Kurihara, and C. Guard, 1994: Global climate change and tropical cyclones. Bull. Amer. Meteor. Soc.,75, 2147–2157.

  • McBride, J. L., 1995: Tropical cyclone formation. Global Perspective on Tropical Cyclones, WMO/TD-No. 693, World Meteorological Organization, 63–105.

  • Neumann, C. J., 1993: Global overview. Global Guide to Tropical Cyclone Forecasting, World Meteor. Org., 1.1–1.56.

  • Reynolds, R. W., 1988: A real-time global sea surface temperature analysis. J. Climate,1, 75–86.

  • ——, and D. C. Marsico, 1993: An improved real-time global sea surface temperature analysis. J. Climate,6, 114–119.

  • Sadler, J. C., 1967: On the origin of tropical vortices. Proc. Working Panel on Tropical Dynamic Meteorology, Norfolk, VA, Naval Weather Research Facility, 39–75.

  • Wright, P. B., 1985: The Southern Oscillation: An ocean–atmosphere feedback system. Bull. Amer. Meteor. Soc.,66, 398–412.

  • Wu, G., and N.-C. Lau, 1992: A GCM simulation of the relationship between tropical-storm formation and ENSO. Mon. Wea. Rev.,120, 958–977.

Fig. 1.
Fig. 1.

Time series of (a) the summer-mean sea surface temperature (SST) departure, ΔSST, over the NINO3 region and (b) the summer-mean 850-mb streamfunction departure of ΔΨ(850 mb) averaged over a part of region (1 + 3) (15°–30°N, 120°–160°E) from their corresponding multiple summer-mean values. Letters of W (C) indicate summers with ΔSST(NINO3) ≥ 0.5°C (⩽−0.5°C).

Citation: Monthly Weather Review 126, 4; 10.1175/1520-0493(1998)126<1080:IVITTC>2.0.CO;2

Fig. 2.
Fig. 2.

Tropical cyclone genesis frequency of every summer (June–August) during 1979–94 over (a) total domain (0°–30°N, 120°E–180°), (b) region (1 + 3) (15°–30°N, 120°E–180°), (c) region (2 + 4) (0°–15°N, 120°E–180°), and (d) the difference of tropical cyclone genesis frequency between region (2 + 4) and (1 + 3). Each region of the analysis domain is marked in Figs. 6 and 7. The tropical cyclone genesis frequency of each warm (cold) summer is heavily (lightly) stippled.

Citation: Monthly Weather Review 126, 4; 10.1175/1520-0493(1998)126<1080:IVITTC>2.0.CO;2

Fig. 3.
Fig. 3.

Tropical cyclone genesis frequency of every summer during 1979–94 over (a) region (1 + 2) (0°–30°N, 120°–150°E), (b) region (3 + 4) (0°–30°N, 150°E–180°), (c) the difference of tropical cyclone genesis frequency between regions (1 + 2) and (3 + 4), and (d) the longitudinal extent of the monsoon trough. The tropical cyclone genesis frequencies of warm (cold) summers are heavily (lightly) stippled in histograms of (a)–(c). The east end of the monsoon trough located east (west) of 150°E is marked by a heavily stippled (doubly hatched) block in (d).

Citation: Monthly Weather Review 126, 4; 10.1175/1520-0493(1998)126<1080:IVITTC>2.0.CO;2

Fig. 4.
Fig. 4.

Same as Fig. 1a except for fall. Letters of W (C) indicate falls with ΔSST(NINO3) > 0.5°C (⩽−0.5°C) over 1979–94.

Citation: Monthly Weather Review 126, 4; 10.1175/1520-0493(1998)126<1080:IVITTC>2.0.CO;2

Fig. 5.
Fig. 5.

Same as Fig. 3 except for fall.

Citation: Monthly Weather Review 126, 4; 10.1175/1520-0493(1998)126<1080:IVITTC>2.0.CO;2

Fig. 6.
Fig. 6.

The 850-mb streamline charts superimposed with outgoing longwave radiation (OLR) for (a) summer climatology averaged over 1979–94, (c) summer climatology averaged over warm summers (1982, 1983, 1987, 1991), (d) summer climatology averaged over six cold summers (1981, 1984, 1985, 1988, 1989, 1994), and (b) the difference of the 850-mb flow between Figs. 2d and 2c. The solid black dots represent the locations of tropical cyclones identified at the beginnings of their life cycles. Note that areas of OLR ⩽ 210 W m−2 and ΔOLR ⩽ −8 W m−2 are heavily stippled, but areas of 210 W m−2 ⩽ OLR ⩽ 240 W m−2 and ΔOLR ≥ 8 W m−2 are lightly stippled.

Citation: Monthly Weather Review 126, 4; 10.1175/1520-0493(1998)126<1080:IVITTC>2.0.CO;2

Fig. 7.
Fig. 7.

Latitudinal locations of the monsoon trough (solid line) and the minimum OLR value (dashed line) averaged over 130°–140°E. The two latitudes marked on the right ordinate are the averaged latitudinal locations of the monsoon trough and minimum OLR value within the longitudinal zone (130°–140°E).

Citation: Monthly Weather Review 126, 4; 10.1175/1520-0493(1998)126<1080:IVITTC>2.0.CO;2

Fig. 8.
Fig. 8.

Same as Fig. 5 except for fall.

Citation: Monthly Weather Review 126, 4; 10.1175/1520-0493(1998)126<1080:IVITTC>2.0.CO;2

Table 1.

Summer (June–August) tropical cyclone genesis frequency averaged over 1979–94 (n), over four warm summers (1982, 1983, 1987, 1991) (nw), over six cold summers (1981, 1984, 1985, 1988, 1989, 1994), the difference between the two latter quantities (ncnw), and the accumulation of tropical cyclone genesis frequency from 1979 to 1994 (Σn).

Table 1.
Table 2.

Same as Table 1 except for the contrast of tropical cyclone genesis frequency between regions west (1 + 2) and east (3 + 4) of 150°E.

Table 2.
Table 3.

Same as Table 2 except for fall, the cold and warm falls include those of (1984, 1985, 1988, 1989) and (1982, 1986, 1987, 1991), respectively.

Table 3.

1

The tropical cyclone genesis of this study refers to the first appearance of any identified tropical cyclone whose maximum sustained wind speed exceeds 17.5 m s−1.

Save