• Arpe, K., 1989: Impacts of changes in the ECMWF analysis-forecasting scheme on the systematic error of the model. Proc. Ten Years of Medium-Range Weather Forecasting, Vol. 1, Shinfield Park, Reading, United Kingdom, European Centre for Medium-Range Weather Forecasts, 466–474.

  • Challa, M., and R. L. Pfeffer, 1990: Formation of Atlantic hurricanes from cloud clusters and depressions. J. Atmos. Sci.,47, 909–927.

  • Chen, S. S., and W. M. Frank, 1993. A numerical study of the genesis of extratropical convective mesovortices. Part I: Evolution and dynamics. J. Atmos. Sci.,50, 2401–2426.

  • Emanuel, K. A., N. Renno, L. R. Schade, M. Bister, M. Morgan, D. J. Raymond, and R. Rotunno, 1993: Tropical cyclogenesis over the eastern North Pacific: Some results of TEXMEX. Preprints, 20th Conf. on Hurricanes and Tropical Meteorology, San Antonio, TX, Amer. Meteor. Soc., 110–113.

  • Frank, W. M., 1982: Large-scale characteristics of tropical cyclones. Mon. Wea. Rev.,110, 572–586.

  • ——, 1988: Tropical cyclone formation. A Global View of Tropical Cyclones, R. L. Elsberry, Ed., 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.

  • ——, 1975: Tropical cyclone genesis. Atmospheric Science Paper 234, 121 pp. [Available from Dept. of Atmos. Sci., Colorado State University, Fort Collins, CO 80523.].

  • ——, 1988: Environmental influences on tropical cyclones. Aust. Meteor. Mag.,36, 127–139.

  • Joint Typhoon Warning Center (JTWC), 1988: Annual tropical cyclone report. U.S. Naval Oceanography Command Center, 216 pp. [Available from U.S. Naval Oceanography Command Center, COMNAVMARIANAS Box 12, FPO, San Francisco, CA 96630.].

  • ——, 1989: Annual tropical cyclone report. U.S. Naval Oceanography Command Center, 254 pp. [Available from U.S. Naval Oceanography Command Center, COMNAVMARIANAS Box 12, FPO, San Francisco, CA 96630.].

  • Landsea, C. W., 1993: A climatology of intense (or major) Atlantic hurricanes. Mon. Wea. Rev.,121, 1703–1713.

  • Lee, C.-S., R. Edson, and W. M. Gray, 1989: Characteristics of north Indian Ocean tropical cyclone activity. Mon. Wea. Rev.,117, 407–426.

  • McBride, J. L., 1981: Observational analysis of tropical cyclone formation. Part I: Basic definition of data sets. J. Atmos. Sci.,38, 1117–1131.

  • ——, and R. M. Zehr, 1981: Observational analysis of tropical cyclone formation. Part II: Comparison of non-developing versus developing systems. J. Atmos. Sci.,38, 1132–1151.

  • ——, and T. D. Keenan, 1982: Climatology of tropical cyclone genesis in the Australian region. J. Climatol.,2, 13–33.

  • Montgomery, M. T., and B. F. Farrell, 1993: Tropical cyclone formation. J. Atmos. Sci.,50, 285–310.

  • Ramage, C. S., 1974: Monsoonal influences on the annual variation of tropical cyclone development over the Indian and Pacific Oceans. Mon. Wea. Rev.,102, 745–753.

  • Raymond, D. J., and H. Jiang, 1990: A theory for long-lived mesoscale convective systems. J. Atmos. Sci.,47, 3067–3077.

  • Reed, R. J., and E. E. Recker, 1971: Structure and properties of synoptic-scale wave disturbances in the equatorial western Pacific. J. Atmos. Sci.,28, 1117–1133.

  • Ritchie, E. A., 1995: Mesoscale aspects of tropical cyclone formation. Ph.D. dissertation, Monash University, 167 pp. [Available from Monash University, Wellington Rd., Clayton, Victoria 3168, Australia.].

  • Sadler, J. C., 1976: A role of the tropical upper troposphere in early season typhoon development. Mon. Wea. Rev.,104, 1266–1278.

  • ——, 1978: Mid-season typhoon development and intensity changes and the tropical upper tropospheric trough. Mon. Wea. Rev.,106, 1137–1152.

  • Shapiro, L. J., 1977: Tropical storm formation from easterly waves: A criterion for development. J. Atmos. Sci.,34, 1007–1021.

  • Unden, P., 1989: Tropical data assimilation and analysis of divergence. Mon. Wea. Rev.,117, 2495–2517.

  • Velasco, I., and J. M. Fritsch, 1987: Mesoscale convective complexes in the Americas. J. Geophys. Res.,92, 9591–9613.

  • Wallace, J. M., 1971: Spectral studies of tropospheric wave disturbances in the tropical western Pacific. Rev. Geophys. Space Phys.,9, 557–612.

  • Zehnder, J. A., 1991: The interaction of planetary-scale tropical easterly waves with topography: A mechanism for the initiation of tropical cyclones. J. Atmos. Sci.,48, 1217–1230.

  • Zehr, R. M., 1992: Tropical cyclogenesis in the western North Pacific. NOAA Tech. Rep. NESDIS 61, 181 pp. [Available from U.S. Department of Commerce, NOAA/NESDIS, 5200 Auth Rd., Washington, DC 20233.].

  • View in gallery
    Fig. 1.

    Locations relative to an idealized monsoon trough of tropical cyclones at the time of genesis for 1988–89.

  • View in gallery
    Fig. 2.

    Composites of the 200-hPa winds for the 1988 storm cases at (a) −72 h, (b) −48 h, (c) −24 h, and (d) time of tropical cyclogenesis. Maximum wind vector is 40 m s−1.

  • View in gallery
    Fig. 3.

    Same as in Figure 2 but for 1989 storm cases.

  • View in gallery
    Fig. 4.

    Composites of the 850-hPa winds for the 1988 storm cases at (a) −120 h, (b) −48 h, (c) −24 h, and (d) time of tropical cyclogenesis. Maximum wind vector is 10 m s−1.

  • View in gallery
    Fig. 5.

    Same as in Fig. 4 but for 1989 storm cases.

  • View in gallery
    Fig. 6.

    Wind anomaly fields for the 1988 850-hPa composites: (a) −48 h, (b) −24 h, and (c) time of genesis. Maximum wind vector is 10 m s−1.

  • View in gallery
    Fig. 7.

    Same as in Figure 6 but for 1989 850-hPa composites.

  • View in gallery
    Fig. 8.

    The 200–850-hPa vertical shear of the zonal wind at the time of genesis for the 1988 composite.

  • View in gallery
    Fig. 9.

    Composite of the 850-hPa wind for the 1989 nongenesis cases of the monsoon trough circulation. The center point is the confluence region of the monsoon trough. Maximum wind vector is 10 m s−1.

  • View in gallery
    Fig. 10.

    Same as in Fig. 9 but for the 200-hPa wind. Maximum wind vector is 40 m s−1.

  • View in gallery
    Fig. 11.

    Locations relative to the genesis point of axes of upper-level troughs present at the time of tropical cyclogenesis.

  • View in gallery
    Fig. 12.

    Locations relative to the confluence region of the monsoon trough of the closest point of upper-level troughs for the nongenesis cases.

  • View in gallery
    Fig. 13.

    Locations of preexisting tropical cyclones relative to the positions of new storms at the time of tropical cyclogenesis.

  • View in gallery
    Fig. 14.

    Time–longitude Hovmöller diagram of the zonal component of the 850-hPa wind along the genesis latitude for Typhoon Ruby (1988). Times are 1200 UTC.

  • View in gallery
    Fig. 15.

    Track of Typhoon Pat relative to the genesis location of Typhoon Ruby. Genesis occurred at 1200 UTC 20 October 1988.

  • View in gallery
    Fig. 16.

    Time–longitude Hovmöller diagram of the meridional component of the 850-hPa wind along the genesis latitude for Typhoon Pat (1988). Times are 1200 UTC.

  • View in gallery
    Fig. 17.

    Time–latitude Hovmöller diagram of the zonal component of the 850-hPa wind along the genesis longitude for Typhoon Skip (1988). Times are 0000 UTC.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 439 194 7
PDF Downloads 384 159 8

Large-Scale Influences on Tropical Cyclogenesis in the Western North Pacific

Lisa M. BriegelDepartment of Meteorology, The Pennsylvania State University, University Park, Pennsylvania

Search for other papers by Lisa M. Briegel in
Current site
Google Scholar
PubMed
Close
and
William M. FrankDepartment of Meteorology, The Pennsylvania State University, University Park, Pennsylvania

Search for other papers by William M. Frank in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

Objectively analyzed data from the European Centre for Medium-Range Weather Forecasts are used to examine the large-scale aspects of the formation of tropical cyclones. It is hypothesized that tropical cyclogenesis occurs when external atmospheric forcing on the synoptic or larger scale provides uplift through a deep layer, enhancing convection, in a region with environmental conditions favorable for genesis. Emphasis is placed on the roles of upper-level troughs, low-level wind surges, preexisting tropical cyclones, and propagating wave disturbances in triggering tropical cyclogenesis. Composites of the 200-hPa and 850-hPa flows reveal the presence of both upper-level troughs and low-level wind surges, respectively, prior to genesis. In the composites, the wind surges also appear to be related to the presence of a prior circulation located approximately 2000 km to the west of the genesis location. An examination of the individual cases demonstrates that approximately 85% of all storms had either an upper-level trough or an identifiable low-level feature, while 49% had both an upper- and a lower-level feature. Given the limitations of the objective analyses over the tropical oceans, this provides strong evidence for the role of large-scale external forcing in triggering tropical cyclogenesis.

Corresponding author address: Ms. Lisa M. Briegel, Department of Meteorology, The Pennsylvania State University, 503 Walker Building, University Park, PA 16802-5013.

Email: briegel@essc.psu.edu

Abstract

Objectively analyzed data from the European Centre for Medium-Range Weather Forecasts are used to examine the large-scale aspects of the formation of tropical cyclones. It is hypothesized that tropical cyclogenesis occurs when external atmospheric forcing on the synoptic or larger scale provides uplift through a deep layer, enhancing convection, in a region with environmental conditions favorable for genesis. Emphasis is placed on the roles of upper-level troughs, low-level wind surges, preexisting tropical cyclones, and propagating wave disturbances in triggering tropical cyclogenesis. Composites of the 200-hPa and 850-hPa flows reveal the presence of both upper-level troughs and low-level wind surges, respectively, prior to genesis. In the composites, the wind surges also appear to be related to the presence of a prior circulation located approximately 2000 km to the west of the genesis location. An examination of the individual cases demonstrates that approximately 85% of all storms had either an upper-level trough or an identifiable low-level feature, while 49% had both an upper- and a lower-level feature. Given the limitations of the objective analyses over the tropical oceans, this provides strong evidence for the role of large-scale external forcing in triggering tropical cyclogenesis.

Corresponding author address: Ms. Lisa M. Briegel, Department of Meteorology, The Pennsylvania State University, 503 Walker Building, University Park, PA 16802-5013.

Email: briegel@essc.psu.edu

1. Introduction

Due to the sparse observational network over the tropical oceans, progress toward understanding the physical processes involved in tropical cyclone formation has been difficult. Several theories have been presented to explain the way in which a loosely organized area of convection first becomes a tropical depression and then intensifies into a tropical storm or typhoon. However, the historical observational database generally has not permitted detailed analyses of the large-scale flow characteristics associated with the formation of individual storms. A few cases of tropical cyclogenesis have been sampled by aircraft, but most of those were observed intermittently, and it is difficult to resolve the evolution of the large-scale environment of the storm using measurements from a single platform.

With the introduction of global objective analyses that combine model output with observations, researchers have been provided with an increased ability to explore the physical processes involved in the formation of individual tropical cyclones. This study uses objectively analyzed data from the European Centre for Medium-Range Weather Forecasts (ECMWF) to examine the large-scale aspects of the development of tropical cyclones.

Section 2 presents background discussion of the nature of tropical cyclone genesis and reviews previous research in the area. The dataset and methodology are described in section 3, results are presented in section 4, and section 5 contains the conclusions.

2. Background

a. Tropical cyclogenesis

In the first global climatology of tropical cyclogenesis, Gray (1968) established several environmental conditions as favorable for the formation of tropical cyclones: a warm ocean layer of sufficient depth, conditional instability through a deep atmospheric layer, higher than normal midtropospheric relative humidity, above-normal low-level vorticity, weak vertical wind shear over the center of circulation, and location sufficiently far from the equator (significant value of the Coriolis parameter f). The first three parameters are considered to be thermodynamic variables, while the last three are dynamic. While the necessary thermodynamic and Coriolis conditions exist over a considerable portion of the tropical oceans for long periods of time, the low-level vorticity and vertical shear parameters can change significantly on much smaller timescales and space scales (Gray 1975; McBride 1981). Thus, it has been hypothesized that tropical cyclogenesis occurs when above-normal low-level vorticity and locally weak vertical wind shear occur within a thermodynamically favorable environment (Gray 1975). More recent studies have indicated that cyclogenesis also tends to occur with a period of enhanced large-scale uplifting with concurrent deep convection (Zehr 1992).

Frank (1988) proposed that tropical cyclone development incorporates two distinct phases, genesis and intensification, that are dominated by different dynamical and physical processes. The first consists of the formation of a mesoscale vortex within a loosely organized tropical cloud cluster. This is a relatively rapid process in which external forcing is thought to be required to produce convection by perturbing the dynamic variables over a relatively large area sufficiently above their climatological values. Within the stratiform cloud deck associated with this deep convection, a mesoscale vortex then forms. If the vortex reaches a threshold intensity, it becomes capable of intensifying into a mature tropical cyclone due to its own interactions with the ocean without further external forcing. External forcing may facilitate the process, however. We refer to this latter stage as intensification, and we define tropical cyclogenesis as the process of developing a vortex capable of self-intensification.

During tropical cyclogenesis, the vortex formation has been observed to occur in conjunction with the formation of a mesoscale convective system (MCS) within a preexisting tropical disturbance (Zehr 1992; Ritchie 1995). The initial vortex forms within the stratiform rain region of the MCS, where the radius of deformation is reduced from ambient values (Chen and Frank 1993; Emanuel et al. 1993; Velasco and Fritsch 1987; Raymond and Jiang 1990). In order for a large convective area (on the order of a few hundred kilometers) such as an MCS to develop and be maintained, low-level convergence and deep uplifting on the synoptic or larger scales must occur, presumably due to external forcing (Gray 1988).

The above scenario is in agreement with a recent observational study by Zehr (1992), who examined 50 individual tropical storms and typhoons that occurred in the western North Pacific during 1983–84. He proposed a conceptual model of tropical cyclogenesis that consists of two distinct stages: the initial development of a mesoscale convective vortex (MCV) within a tropical MCS, followed by the transformation of this vortex into a tropical cyclone with a subsequent lowering of the central surface pressure. He found that each stage was preceded by a significant increase in convection concurrent with deep uplifting.

It is difficult to determine exactly when a disturbance has crossed the boundary between genesis and intensification. When a tropical disturbance develops a rotation center at the surface, forecasters usually classify it as a tropical depression. Once the vortex achieves maximum winds of 17 m s−1 or greater, it is then classified as a tropical storm. There is nearly universal agreement that when a storm reaches tropical storm intensity, it will intensify into a mature cyclone without external forcing unless it is inhibited by an unfavorable environment (principally strong vertical wind shear and/or moving over land or cold water). However, a storm of tropical-depression strength also tends to be a relatively stable system, and it appears to be capable of intensifying without obvious external forcing, albeit very slowly. We therefore choose to define tropical cyclogenesis as the sequence of events leading to the formation of a tropical depression. While there may be other suitable definitions, this one seems to fit observations adequately, and it has the advantage that it is relatively well defined in the historical data records (e.g., best track data). It should be noted, however, that the time of declared tropical depression and the time at which a vortex becomes self-sustaining may differ by as much as 12–24 h. Unfortunately, there is no way currently to identify the exact moment of genesis, but the time of tropical depression is quite adequate for the purposes of this research. We view tropical cyclogenesis as an impulsive, multistage process that requires external forcing on scales significantly larger than the convective disturbance. The goal of this paper is to document the climatology and nature of the larger-scale circulations associated with the formation of tropical cyclones.

b. Large-scale forcing of genesis

A review of the literature on tropical cyclogenesis reveals a number of different proposals for atmospheric mechanisms that might provide the external forcing (e.g., enhanced large-scale upward motion) necessary to cause an MCS, and thus an MCV, to form. Some researchers have proposed that the environmental forcing may occur in the lower levels. Lee et al. (1989) and Zehr (1992) argued that inward-propagating regions of enhanced low-level inflow, which they referred to as wind surges, might trigger tropical cyclogenesis. Zehr (1992) observed that most of the wind surges associated with cases of tropical cyclogenesis in the western North Pacific had a westerly component. He hypothesized that many tropical disturbances may experience an increase in westerly winds when they first encounter the far eastern end of the monsoon trough in their westward movement and thus undergo genesis.

Upper-level circulations have also been examined for possible atmospheric processes that might trigger tropical cyclone formation. McBride and Keenan (1982) documented that 45% of their cases of tropical cyclogenesis in the Australian region occurred when upper-tropospheric troughs were located within 5° and 80% within 15° latitude poleward of a disturbance. Sadler (1976, 1978) also found that tropical cyclogenesis in the western North Pacific occurred more readily when an upper-level trough was poleward and westward of a disturbance.

Westward-moving intense cyclonic cells in the tropical upper-tropospheric trough (TUTT) are believed to trigger cyclogenesis by providing the intense outflow channel necessary for the initiation as well as the intensification of a tropical cyclone (Sadler 1976). Challa and Pfeffer (1990) hypothesized that an upper-tropospheric wavelike disturbance could force tropical cyclogenesis through an eddy flux convergence of angular momentum. Montgomery and Farrell (1993) proposed a mechanism for the formation of tropical cyclones involving an interaction between migratory potential vorticity anomalies in the upper levels and low-level tropical disturbances. These upper-level troughs may force uplift through a deep layer, and thus enhance convection, by unbalancing the flow aloft.

Westward-propagating wave disturbances, such as those examined by Wallace (1971) and Reed and Recker (1971), have also been presented as a possible forcing mechanism for tropical cyclogenesis through a local increase in low-level convergence and/or relative vorticity (Shapiro 1977; Zehnder 1991). Landsea (1993) documented that 57% of tropical storms, 58% of minor hurricanes, and 83% of intense hurricanes in the Atlantic basin from 1967 to 1991 formed from easterly waves. In the western North Pacific, Ritchie (1995) showed that only approximately 10%–15% of tropical cyclogenesis cases in 1990–92 could be attributed to easterly waves.

In this study, emphasis is placed on identifying the possible roles of upper-level troughs, low-level wind surges, propagating wave disturbances, and preexisting tropical cyclones in triggering the formation of tropical cyclones. In order to ensure similarity in the cases of cyclogenesis studied, only storms that formed in the monsoon trough of the western North Pacific Ocean are included. The monsoon trough is the region of low-level convergence between the equatorial westerlies and the trade wind easterlies. Gray (1968) and Ramage (1974) have shown the presence of the monsoon trough, with its relatively low vertical wind shear and high relative vorticity, to be extremely favorable for tropical cyclogenesis. In a climatology of tropical cyclogenesis in the western Pacific Ocean, Ritchie (1995) found that slightly more than 75% of the classifiable genesis cases from 1990 to 1992 occurred in the monsoon trough. While the monsoon trough can be considered an extended region of forced uplifting, this research will focus on mechanisms within the monsoon trough that increase and focus this uplifting. The underlying hypothesis of this paper is that tropical cyclogenesis occurs when a synoptic-scale circulation system provides uplift through a deep layer, enhancing convection, in a region with environmental conditions favorable for genesis.

3. Data and methodology

Initialized wind fields from the ECMWF were examined on two different tropospheric pressure levels, 850 and 200 hPa. These levels include more data than other levels, thus providing the best description of the tropical tropospheric dynamics (Arpe 1989). The data are available twice daily on a 2.5° × 2.5° resolution latitude–longitude grid.

In the past decade, the inclusion of satellite data and other changes in the objective analysis have greatly improved the center’s forecasts. Of particular importance to this study are changes to the center’s forecasts that affect the magnitude and/or direction of the analyzed tropical divergent winds. The latest change having a significant beneficial effect on the analyzed divergence was in January 1988 when the constraint forcing the wind to be nondivergent within the analysis box was removed (Unden 1989). Therefore, the years 1988 and 1989 were selected for the present study. In order to identify any synoptic-scale circulation patterns leading to genesis, the data were examined with respect to the tropical cyclone’s center of circulation as obtained from the Joint Typhoon Warning Center (JTWC) best track data.

The JTWC Annual Tropical Cyclone Reports summarize the formation and progress of the various tropical cyclones in the Pacific Ocean. From these storm summaries (JTWC 1988, 1989) and the ECMWF streamline analyses, it was determined that 16 tropical cyclones out of a total of 26 in 1988 and 29 out of a total of 35 in 1989 formed in the monsoon trough. The locations of each storm at the time of genesis with respect to an idealized monsoon trough as determined from an examination of the streamline analyses are plotted in Fig. 1. Eleven of the 16 tropical cyclones that formed in the monsoon trough in 1988 and 17 of the 29 in 1989 did so in the eastern end where the trade winds on the equatorial side curl around to the westerly direction. This result is in agreement with the findings of Zehr (1992).

Systems that were declared tropical depressions by the JTWC, weakened, and then redeveloped were eliminated from this study. Additionally, in order to ensure that the storms in this study developed in similar environments, cases of tropical cyclogenesis in which the monsoon trough was located relatively far north of its normal climatological position (north of 30°N) were also eliminated. In total, 14 tropical cyclones for 1988 and 27 tropical cyclones for 1989 were retained in the dataset.

In order to reduce the random and sampling errors inherent to an objective analysis as well as resolve common features among the individual cases of tropical cyclogenesis, the data were composited. Each year’s set of storms was composited separately to avoid having the composites include differences in the mean flow between the two years. Compositing smooths the data, resulting in a diminishing of any signal present. Therefore, if a composite reveals evidence of a relationship between genesis and the environmental flow, the circulation feature observed is likely to be even stronger in individual analyses. Thus, for any feature observed in the composites, further study of the individual cases was performed.

Both the 200- and 850-hPa wind fields were composited using a 6000-km square grid placed over the nearest grid point to the center of each storm at the time it was declared a tropical depression. This location was defined as the genesis location. Composites were computed for every 24-h time period back to 5 days before tropical cyclogenesis. Each composite grid was subdivided into 300-km square boxes. Any of the objective analysis data points falling within one of these boxes was averaged with the other values in that box.

Since a change in the speed and/or direction of the wind prior to tropical cyclogenesis was expected, difference fields were obtained from the composites by subtracting the composited winds at one observation time from the composited winds in that same location for another time. Difference fields between a reference time, considered to be representative of the background flow, and each observation period were calculated.

In order to study the possible role played by propagating disturbances in triggering tropical cyclogenesis, Hovmöller diagrams of the 850-hPa zonal and meridional wind were also produced. The winds along the latitude and longitude of genesis were plotted in time–longitude and time–latitude cross sections, respectively. These time–distance cross sections can be used to track the movement of a maximum or minimum feature relative to a fixed location. Hovmöller diagrams of 5° bands on either side of the genesis latitude and on either side of the genesis longitude were also constructed, but since they produced basically the same results, only the Hovmöller diagrams of the winds along the latitude and longitude of genesis are presented.

4. Results

a. Composite analyses

1) Upper-level circulation

This section examines the composite wind field surrounding the location where tropical cyclogenesis occurred for time periods up to 5 days prior to storm formation. Velocity vector composites of the 200-hPa wind fields for the 1988 storms are shown in Fig. 2. The average genesis location of the storms in this year is 15.4°N, 135.2°E. Within approximately 1200 km of the average latitude of genesis, the winds are generally weak for each 24-h time period. All of the 200-hPa composites reveal strong westerly flow northward of an east–west line about 1500 km north of the genesis latitude. The 200-hPa composite for 72 h before genesis (Fig. 2a) shows a weak trough (labeled with a “T” in Fig. 2) appearing in this westerly flow approximately 3400 km to the northwest of the genesis location. The upper-level trough deepens and moves directly eastward so that by 24 h prior to genesis it is approximately 2200 km to the north-northwest of the genesis longitude (Fig. 2c). By the time of tropical cyclogenesis, the trough is positioned approximately 1500 km directly northward of the genesis location (Fig. 2d). On much closer examination of the small wind vectors, a very weak cyclonic circulation resembling the TUTT appears approximately 1200–1500 km to the east-northeast of the genesis location. This circulation remains basically unchanged and in the same location for all five time periods.

The velocity vector composites of the 200-hPa wind fields for the 27 storms in 1989 are shown in Fig. 3. The average genesis location of the storms in this year is 15.6°N, 137.6°E. The composites for all of the time periods, in agreement with those from 1988, show no major wind feature within approximately 1200 km of the genesis location, but with strong westerly flow farther to the north. Starting in the −96-h composite (not shown), a weak ridge lies about 2200 km to the northeast of the genesis location. This upper-level ridge (marked with an “R” in Fig. 3) remains stationary, although it strengthens continuously through each time period. The composite for 72 h before genesis (Fig. 3a) shows the presence of an upper-level trough (labeled with a “T” in Fig. 3) located approximately 1500 km to the northwest of the genesis location. This trough then weakens over the next 24–36-h period. Another trough lies in the composite for 24 h before genesis (Fig. 3c) approximately 2800 km northwest of the genesis location. At the time of tropical cyclogenesis, this trough is located about 1600 km directly north of the genesis location (Fig. 3d). The very weak TUTT circulation revealed in the 1988 composites also appears in all time periods for 1989. These results agree with those from the 1988 200-hPa composites.

2) Low-level circulation

Composites of the 850-hPa wind fields for 1988 storms are shown in Fig. 4. The composite for 120 h before tropical cyclogenesis (Fig. 4a) shows a weak monsoon trough flow in which the genesis location is embedded. There is southwesterly flow within the monsoon trough westward of a point about 500 km to the east of the genesis location at this time. The trade-wind easterlies are generally weak to the north of the genesis location. A weak cyclonic circulation, which we will term a “prior circulation,” first appears approximately 500 km to the west of the genesis location in the composite for 48 h before genesis (Fig. 4b). Accompanying this prior circulation (labeled “P” in Fig. 4) is strong southwesterly flow (marked with an arrow in Fig. 4) from the southwest quadrant into the genesis location. By 24 h before genesis (Fig. 4c), the prior circulation has greatly weakened leaving behind strong easterly flow to the north of the genesis location and westerly flow to the south (marked with arrows in Fig. 4). By the time of tropical cyclogenesis, a strong cyclonic circulation around the genesis location has fully developed (Fig. 4d).

The velocity vector composites for 1989 are shown in Fig. 5. The composite circulation for 120 h prior to genesis (Fig. 5a) shows the far eastern portion of the monsoon trough, with easterly flow to the east of the genesis location and southwesterly flow to the southwest of the genesis location. The region of transition between easterly and westerly monsoonal flow is farther west during 1989 than during 1988. The monsoon trough seen in the 1989 composites, with its higher wind speeds, appears much more clearly than that appearing in the 1988 composites. As noted by JTWC, the monsoon trough in 1989 was very active and extremely broad (JTWC 1989).

The low-level wind field remains basically unchanged until 48 h before tropical cyclogenesis. At this time (Fig. 5b), the winds begin to flow cyclonically around the genesis location with a clear increase in the southwesterly flow (marked with an arrow in Fig. 5) from the southwest quadrant into the genesis location. Also appearing in this composite is a second center (labeled with a “P”) of weak, large-scale rotation (the prior circulation) about 2000 km to the west along the average latitude of genesis.

By 24 h before tropical cyclogenesis (Fig. 5c), the primary, developing cyclonic circulation around the genesis location, the prior circulation, and the southwesterly inflow into the genesis location all have strengthened. By the time of tropical cyclogenesis, the large-scale cyclonic circulation around the genesis location has fully developed, and the prior circulation has moved so that it is approximately 3000 km to the west (Fig. 5d) of the genesis location. In general, the synoptic-scale features resolved in the 1989 composites are stronger than their counterparts in the 1988 composites.

In order to emphasize changes in the mean flow associated with tropical cyclogenesis, difference fields of the 850-hPa composite wind fields have been computed for both years. Particular emphasis is placed on how tropical cyclogenesis might be affected by the prior circulation (labeled “P” in Figs. 4–7) located to the west of the genesis location or by any wind surges (marked with arrows in Figs. 4–7) propagating toward the genesis location. Since the difference fields between 120 and 96 h before genesis for both 1988 and 1989 show very little change in the winds, the 120-h composites are assumed to be representative of the large-scale pregenesis flow. Each wind anomaly field is computed using the time of 120 h before genesis as a reference time period.

The wind anomaly fields for the 1988 850-hPa composites are shown in Figs. 6a–c. The prior circulation to the west of the genesis location first appears in the anomaly field for 48 h before genesis (Fig. 6a). This circulation feature seems to be associated with an increase in the southwesterly inflow from the southwest quadrant into the genesis location. An increase in the trade wind easterly flow also appears approximately 1500 km east-northeast of the genesis location in this anomaly field. In the 24-h anomaly field (Fig. 6b), the prior circulation has substantially weakened, but the easterly flow to the north of the genesis location has greatly increased. The flow to the southwest of the genesis location has also strengthened, becoming more westerly. The anomaly field for the time of tropical cyclogenesis (Fig. 6c) reveals a fully developed cyclonic circulation around the genesis location, with the easterly winds to the north of the storm being about twice as strong as the westerly winds to the south.

The anomaly fields for the 1989 850-hPa composites are presented in Fig. 7. Once again, a prior circulation appears in the anomaly field 48 h before genesis (Fig. 7a), located approximately 1500 km to the west of the genesis location. Westerly winds from this prior circulation flow from the southwest into the genesis location. In the anomaly field for 24 h before genesis (Fig. 7b), the prior circulation has moved westward to approximately 2000 km west of the genesis location. The westerly inflow has moved closer to the place of genesis and strengthened, feeding into the cyclonic flow around the genesis location. The anomaly field for the time of tropical cyclogenesis (Fig. 7c) shows that the cyclonic circulation around the genesis location has become well developed, while the prior circulation has moved westward to approximately 2700 km west of the genesis location.

The increase in the monsoonal southwesterly winds equatorward of the genesis location observed in the 1988 and 1989 850-hPa composites provides low-level convergence on the synoptic scale that might have triggered tropical cyclogenesis as described in Zehr’s conceptual model. The divergence fields calculated from each year’s composite winds were examined for evidence of low-level convergence near the genesis location that might be due to these southwesterly winds. In 1988, the convergence at the genesis location increased by 6 × 10−5 to 1 × 10−5 s−1; a very narrow area around the genesis location went from being divergent to convergent. In 1989, it increased by 2 × 10−5 to 3 × 10−5 s−1, with the maximum increase of 5 × 10−5 to 7 × 10−5 s−1 occurring 500 km to the east of the genesis point. Although the increase in convergence was larger in the 1988 composites, it occurred over a much more concentrated area. In 1988, the increase in convergence was confined to an area approximately 400 km wide centered over the genesis location. On the other hand, the 1989 increase occurred over an area approximately 2400 km wide centered over the genesis location. The larger area of convergence increase in 1989 corresponds with the stronger monsoon trough, prior circulation, and southwesterly wind surges observed in the 1989 composite wind fields.

The change in the ITCZ structure also causes an increase in the large-scale vorticity values in the trough. Calculations showed that in 1988, the relative vorticity values within 5° of the genesis location increased by a factor of 8 between 120 h before genesis and the time of genesis, while it quadrupled in 1989.

Since Gray (1968, 1975) demonstrated that tropical cyclogenesis occurs only with low vertical wind shear, the 200–850-hPa vertical shear of the zonal wind was calculated for each composite time period. The results for both 1988 and 1989 were basically the same. Five days prior to genesis (not shown), the zero zonal shear line was located approximately 800 km to the north of the genesis location. It moved closer and closer to the genesis location until it was located over the genesis location at the time of genesis. The vertical shear of the zonal wind for the 1988 composite at the time of genesis is plotted in Fig. 8. The zero shear line lies in a west-northwest to east-southeast direction across the genesis location. This result is consistent with those of McBride and Zehr (1981) for their developing cloud clusters. Calculations of the magnitude of the total vertical wind shear were also made for each composite. In both years, the magnitude remained below a threshold of 10 m s−1 within about 500 km of the genesis location for all time periods. The line of minimum magnitude of the total shear was located in a west-northwest to east-southeast line across the genesis point as did the line of zero zonal shear.

In any research in which a compositing technique is used, it is important to demonstrate the significance of the patterns observed in the composites. Specifically, it is important to show that the circulation patterns appearing in the composites are directly related to the individual genesis cases and that they would not appear in a random grouping of 200- or 850-hPa flow fields. In order to do so, composites of the monsoon trough in which genesis did not occur were constructed for both years and for both vertical levels. Because most of the genesis cases in this study occurred in the confluence region of the monsoon trough, this point was chosen as the center of the composite grid (i.e., is analogous to the genesis location). For each 0000 UTC 850-hPa wind field in which a monsoon trough circulation was visible, this reference point was chosen. Composites of both the 200- and 850-hPa wind fields were constructed centered on this point. The results for the two years were very similar, so only the results from 1989 are presented here. The 850-hPa composite revealed a strong monsoon trough circulation with westerlies to the southwest of the genesis point and easterlies to the east and northwest (Fig. 9). Neither the strong primary cyclonic circulation around the center point nor the prior circulation located to the west seen in the genesis composites appear in these nongenesis composites. In the 200-hPa composite field, no strong winds can be seen within approximately 1200 km of the center point (Fig. 10). As in the genesis composites, strong westerly flow appears about 1500 km to the north of the composite center, but there is no evidence of any upper-level troughs or ridges in that flow. Interestingly, a TUTT circulation resembling that found in the genesis-case composites is also found in the nongenesis-case composites in basically the same location. Zehr (1992) also found that the TUTT tended to be associated equally with developing and nondeveloping tropical disturbances.

Composites of a random sampling of these nongenesis monsoon trough cases were also constructed with basically the same results of the composites of all of the nongenesis monsoon trough cases. This confirms the significance of the patterns observed in the genesis composites.

b. Individual case analyses

1) Upper-level troughs

The 200-hPa streamlines for the individual storms in both 1988 and 1989 are examined for the presence of upper-level troughs. In general, the troughs are much more distinct in the individual analyses than they are in the composites. The results are presented in Table 1. At the time of tropical cyclogenesis, the streamline analyses reveals the presence of an upper-level trough within 2500 km of the genesis location for 9 of the 14 storms in 1988 and for 16 of the 27 storms in 1989.

The axes of these troughs are plotted with respect to the genesis location in Fig. 11. Approximately 65% of the upper-level troughs associated with a storm in 1988 and 75% in 1989 are located in the northwest quadrant relative to the genesis location. This strong tendency for troughs to be located northwest of the genesis location is responsible for the relatively weak troughs described in Figs. 2 and 3. It is not possible to discern from these analyses whether or not the individual troughs are causing large-scale uplifting over the genesis area, but approximately one-half of the troughs shown are located less than 1000 km from the genesis location. In order to determine whether the location of these upper-level troughs to the northwest of the genesis location influenced genesis or is simply fortuitous, an examination was made of the upper levels over monsoon troughs that were not associated with the genesis cases used in this study. Because the majority of the genesis cases occurred in the confluence region of the monsoon trough, this point was used as the reference point around which the upper levels were then studied. For each 0000 UTC 850-hPa wind analysis in which a monsoon trough was discernible, this reference point was identified. The corresponding 200-hPa wind analysis was then examined for the presence of an upper-level trough within 20° of the reference point. The locations of these troughs were then plotted relative to the locations of the confluence region (Fig. 12). The upper-level troughs are basically evenly distributed zonally within the westerly flow north of the ITCZ. This supports the hypothesis that upper-level troughs located to the northwest tend to be favorable for genesis.

2) Preexisting tropical cyclones

One question introduced by the 850-hPa composites concerns the origin of the prior circulation appearing to the west of the genesis location. This cyclonic circulation may have triggered cyclogenesis by increasing the southwesterly inflow into the genesis location, thereby providing low-level convergence, deep uplift, and thus deep convection approximately 72–48 h before genesis. This prior circulation might reflect the presence of a preexisting tropical cyclone. To test this hypothesis, the individual cases have been examined to see whether the prior circulation to the west of the genesis location corresponds with the presence of a second tropical cyclone.

At the time of genesis, 5 of the 14 cases in 1988 had one preexisting storm present, while 3 of the 14 had two preexisting storms. In 1989, 10 of the 27 cases had one preexisting storm present, while 9 of the 27 had two preexisting tropical cyclones present at the time of genesis. Figure 13 shows the locations of these preexisting tropical cyclones relative to the positions of new storms at the time of tropical cyclogenesis. Within the outlined 20° longitude by 20° latitude box to the west of the genesis location, centered approximately on the location of the prior circulations in the 850-hPa composites, lie 36% of the preexisting storms during 1988 and 43% of the storms for 1989.

The above results suggest that the prior circulations seen in the composites are largely due to the fact that in roughly one-third to one-half of the genesis cases there was an existing tropical cyclone within the monsoon trough, located west of the genesis location. The fact that a lower percentage of the genesis cases in 1988 had preexisting storms located to the west than did the 1989 cases supports the observation that the prior circulation in the 1988 850-hPa composite is weaker than the one in the 1989 composites. Cases in which the developing cyclone was the only storm within the composite domain have also been composited to test whether or not the prior circulation was indeed due to the presence of another tropical circulation. As expected, the prior circulation did not appear in these composites for either year (not shown).

From an examination of the tracks and streamline analyses of the individual storm genesis cases, one scenario that arises is one in which a storm develops in the far eastern end of the monsoon trough (confluence point between easterly and westerly flow) and then propagates westward. A new storm does not develop until the preexisting storm is on average approximately 20° to the west. This is consistent with Frank’s (1982) finding that when western North Pacific typhoons first reach typhoon intensity, the most common location for a preexisting storm is 15°–20° to the west of the new typhoon. Frank (1982) attributed this spatial pattern to the fact that the wake region of tropical cyclones is a region of enhanced upward motion and vorticity, favorable for cyclogenesis. Ritchie (1995) argued that downstream energy dispersion is responsible for the eastern side of typhoons being favorable regions for cyclogenesis. These hypotheses are being examined in ongoing research. An alternative viewpoint, however, is that the existence of a cyclone in the monsoon trough causes the southwesterly surge in the monsoonal flow that enhances convergence into the easternmost region of the trough, triggering genesis there.

It is difficult to tell from the streamline analyses whether the preexisting storm provides the southwesterly surge beginning about 48–72 h before the genesis time, as seen in the composites and described by Zehr (1992). It is possible that the monsoonal southwesterlies strengthen from other external influences or from internal processes. In an effort to determine whether increased southwesterly flow into the genesis location is associated with preexisting cyclones, time–longitude cross sections (Hovmöller diagrams) of the zonal wind are examined for each case along the latitude of the genesis location. The westerly flow on the southern side of a preexisting cyclone, when located within or just to the north of the latitude of genesis of the developing storm, generally shows up as a westward-moving band of westerlies on these Hovmöller diagrams. By tracking these bands, one can judge whether increased southwesterly surges are associated with the prior storms or whether they occur due to some other causes.

An example of one of these Hovmöller diagrams is shown in Fig. 14 for the genesis of Typhoon Ruby 1988. A prior storm, Typhoon Pat, formed 6 days before and approximately 6° to the east-southeast of the genesis location of Ruby. Pat then intensified slowly as it tracked first in a westward and then a northwestward direction, passing to the west of the genesis location of Ruby. The track of Typhoon Pat relative to the genesis point of Typhoon Ruby is plotted in Fig. 15. The winds plotted in Fig. 14 are the winds along the closest latitude in the analysis grid to the south of the genesis latitude of Typhoon Ruby for several days prior to and after the genesis of Pat. The horizontal crossbar represents the time, while the vertical represents the longitude of the genesis of Ruby.

As Pat tracks northward across the genesis latitude between 0000 and 1200 UTC 17 October, the winds along the genesis latitude near the longitude of Ruby switch from an easterly to a westerly direction. This shifting of the zonal wind along the genesis latitude can be seen in the Hovmöller diagram for Typhoon Ruby (Fig. 14). The band of westerlies (labeled “W” in Fig. 14) appears at the genesis longitude approximately 48 h before genesis of Ruby and then propagates to the west. In the Hovmöller diagram for the meridional wind (not shown), Pat appears as a westward-propagating band of southerlies. The location of the band of southerlies corresponds to the location of the band of westerlies. Thus, Pat appears to produce southwesterly inflow along the latitude of genesis that provides the necessary low-level convergence needed for the genesis of Ruby.

The time–longitude cross sections for 1988 revealed a total of 6 (out of 14 genesis events) in which a prior storm appeared to provide such a southwesterly inflow. For 1989, a prior storm appeared to cause a southwesterly surge into the genesis location in 8 cases (out of 27 genesis events). Thus, while preexisting cyclones appear to trigger genesis at least in part by increasing the southwesterly flow into the genesis location for about 15%–30% of the storms that form in the ITCZ, other mechanisms must be responsible for the observed surges in other cases (Table 2).

3) Easterly waves

While approximately 50% of tropical cyclones that form in the Atlantic Ocean begin as easterly wave troughs, this mode of genesis is thought to be significantly less common in the Pacific. Ritchie (1995) analyzed 850-hPa vorticity analyses and estimated that approximately 10%–15% of the western North Pacific cyclones during 1990–92 formed from easterly waves. In order to identify possible zonal propagation of easterly waves into the monsoon trough, the time–longitude Hovmöller diagrams of the 850-hPa zonal and meridional winds have been examined. Wave troughs in the diagrams for the zonal and meridional wind components appear as alternating bands of easterlies and westerlies, and northerlies and southerlies, respectively. Since the wave disturbances are generally propagating westward in the trade-wind easterly flow, the results for the zonal component of the wind are expected to be weaker than those for the meridional component (Reed and Recker 1971).

Only one instance of a propagating wave apparently related to a genesis event can be discerned in all of the time–longitude cross sections. In this one case, a westward-propagating wave disturbance appears in the Hovmöller diagram (Fig. 16) for the meridional component of the wind approximately 4 days before the time of genesis of Typhoon Pat (1988) and 45° to the east of the genesis longitude. This wave, with a wavelength of approximately 3800 km, propagates westward along the genesis latitude at a speed of about 10° longitude per day. This result is in agreement with Reed and Recker (1971), who documented that easterly waves have average wavelengths of 2500–4000 km and travel westward at an average speed of 7° longitude per day. A band of southerlies (labeled “S” in Fig. 16) due to this wave propagates westward along the genesis latitude, crossing the genesis longitude at approximately the time of genesis of Typhoon Pat. This band of southerlies strengthens approximately 24 h before genesis, providing increased inflow into the genesis location. This southerly surge may have provided the necessary low-level convergence for tropical cyclogenesis to occur.

4) Other southwesterly surges

The time–longitude cross sections reveal cases of southwesterly surges into the genesis location that are not associated with either prior storms or easterly waves. During 1988 there are four such cases of zonally propagating westerly or southerly wind maxima that appeared to result in southwesterly surges into the genesis location 4–5 days prior to the genesis time. The origins of these events have not yet been determined. During 1989, six such events have been found.

In order to identify possible meridional propagation of southwesterly surges, time–latitude Hovmöller diagrams have been constructed by plotting each component of the wind along the genesis longitude versus time. Although no propagating waves or preexisting storms can be readily distinguished in these time–latitude cross sections, a band of westerlies, located south of the genesis latitude, is observed in 6 of the 14 tropical cyclogenesis cases during 1988 and in 12 of the 27 cases during 1989. In each of these bands of westerlies, the wind increases 4–6 m s−1 approximately 48–72 h before genesis. Figure 17 shows the Hovmöller diagram for the 850-hPa zonal wind for Typhoon Skip (1988). The band of westerlies (labeled “W” in Fig. 17) first appears approximately 1 week before tropical cyclogenesis and propagates northward so that at the time of genesis, westerlies are located directly south of the genesis location. Starting 48 h before genesis, this westerly wind increases approximately 4 m s−1 directly to the south of the genesis location.

A southwesterly surge first appears in the composites about 48–72 h before genesis. While it is difficult to observe a local maximum in the southerly wind at this time in the time–latitude cross sections, such an increase in the westerly wind is observed. Thus, although these surges do lack the southerly component of the surge observed in the composites, it is apparent that these bands provide an increase in the westerly monsoonal flow, thus increasing vorticity and convergence near the genesis location, leading to tropical cyclogenesis.

5. Conclusions

The focus of the paper was to explore the relationships between large-scale forcing and tropical cyclogenesis. Twice-daily objectively analyzed data from ECMWF were examined for tropical cyclones forming in the western Pacific monsoon trough at two different levels, 850 and 200 hPa, for the years 1988 and 1989. The basic hypothesis of this research was that tropical cyclogenesis occurs when the large-scale circulation provides uplift through a deep layer, convection, and finally cyclogenesis. Both the upper- and lower-level wind fields were inspected for common circulation features that could be associated with the formation of tropical cyclones.

Composites of the 200-hPa wind field suggested that upper-level troughs may play a role in tropical cyclogenesis. Of the 41 tropical cyclones examined in this study, 25 storms formed when an upper-level trough was present, 16 when the trough was located within 2500 km to the northwest of the genesis location prior to genesis. It has been hypothesized that these troughs interact with low-level disturbances by providing upper-level vorticity advection, forcing upper-level divergence and uplifting, thus leading to tropical cyclogenesis. Although they typically occur to the northwest of the developing disturbance, there is great variability in both their locations and strengths. These results are consistent with those presented by Sadler (1976, 1978) and McBride and Keenan (1982). In contrast, Zehr (1992) found little difference between the relative occurrence of the TUTT associated with pre–tropical storm disturbances and with nondeveloping disturbances. However, Zehr (1992) examined the upper levels specifically for the presence of the TUTT, while the genesis-related troughs found in this study tended to be eastward-propagating troughs in the westerly flow. In addition, upper-level troughs for nongenesis monsoon trough cases were found to occur equally to both the northwest and the northeast of the confluence region. While it is difficult to conclude that upper-level troughs trigger cyclogenesis, they most likely influence it when they are located to the northwest of a disturbance.

Composites of the 850-hPa wind field revealed that southwesterly surges in the flow to the southwest of the genesis location occurred approximately 48–72 h prior to genesis. These surges, with a horizontal scale of approximately 1000 km, may force the low-level convergence and deep uplifting necessary for tropical cyclogenesis. The individual genesis cases were then examined in order to determine the origins of these southwesterly surges. In only one case was the surge found to be due to an easterly wave. In contrast, Ritchie (1995) found that 10%–15% of western North Pacific cyclones during 1990–92 formed from easterly waves. The difference between the two results most likely occurred because Ritchie (1995) examined all tropical cyclones, while only those that formed in the monsoon trough were analyzed here.

In the composites, these wind surges appeared to be related to the presence of a prior circulation approximately 2000 km to the west of the genesis location. Approximately 66% of the individual genesis cases had at least one preexisting tropical cyclone present at the time of genesis. However, time–longitude cross sections of the 850-hPa zonal and meridional flow demonstrated that in only about 34% of all of the genesis cases did a preexisting storm appear to be the source of the southwesterly surge into the genesis location.

Southwesterly surges unrelated to either preexisting tropical cyclones or easterly waves were also observed in the time–longitude Hovmöller diagrams. Such surges appeared in approximately 24% of the genesis cases.

Approximately 44% of the genesis cases exhibited westerly wind surges along the latitude of genesis in the time–latitude cross sections of the 850-hPa zonal wind. These surges appear to be increases in the westerly winds of the monsoon trough. In approximately one-half of these cases, the surges propagated northward across the genesis location. An examination of the 850-hPa streamlines for the individual cases revealed that this propagation did not appear to result from the northward migration of the monsoon trough.

Thus, approximately 85% of all storms had either an upper-level trough or an identifiable low-level feature, and 49% had both an upper- and lower-level feature. Given the limitations of the data and analyses over the tropical oceans, the fact that such a large percentage of genesis cases in the monsoon trough appear to be associated with large-scale external circulations provides strong evidence for the role of external forcing as a determining factor in tropical cyclogenesis. Improvements in the density of the observational network are needed for better understanding of the links between the larger-scale, external circulations and tropical cyclones.

REFERENCES

  • Arpe, K., 1989: Impacts of changes in the ECMWF analysis-forecasting scheme on the systematic error of the model. Proc. Ten Years of Medium-Range Weather Forecasting, Vol. 1, Shinfield Park, Reading, United Kingdom, European Centre for Medium-Range Weather Forecasts, 466–474.

  • Challa, M., and R. L. Pfeffer, 1990: Formation of Atlantic hurricanes from cloud clusters and depressions. J. Atmos. Sci.,47, 909–927.

  • Chen, S. S., and W. M. Frank, 1993. A numerical study of the genesis of extratropical convective mesovortices. Part I: Evolution and dynamics. J. Atmos. Sci.,50, 2401–2426.

  • Emanuel, K. A., N. Renno, L. R. Schade, M. Bister, M. Morgan, D. J. Raymond, and R. Rotunno, 1993: Tropical cyclogenesis over the eastern North Pacific: Some results of TEXMEX. Preprints, 20th Conf. on Hurricanes and Tropical Meteorology, San Antonio, TX, Amer. Meteor. Soc., 110–113.

  • Frank, W. M., 1982: Large-scale characteristics of tropical cyclones. Mon. Wea. Rev.,110, 572–586.

  • ——, 1988: Tropical cyclone formation. A Global View of Tropical Cyclones, R. L. Elsberry, Ed., 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.

  • ——, 1975: Tropical cyclone genesis. Atmospheric Science Paper 234, 121 pp. [Available from Dept. of Atmos. Sci., Colorado State University, Fort Collins, CO 80523.].

  • ——, 1988: Environmental influences on tropical cyclones. Aust. Meteor. Mag.,36, 127–139.

  • Joint Typhoon Warning Center (JTWC), 1988: Annual tropical cyclone report. U.S. Naval Oceanography Command Center, 216 pp. [Available from U.S. Naval Oceanography Command Center, COMNAVMARIANAS Box 12, FPO, San Francisco, CA 96630.].

  • ——, 1989: Annual tropical cyclone report. U.S. Naval Oceanography Command Center, 254 pp. [Available from U.S. Naval Oceanography Command Center, COMNAVMARIANAS Box 12, FPO, San Francisco, CA 96630.].

  • Landsea, C. W., 1993: A climatology of intense (or major) Atlantic hurricanes. Mon. Wea. Rev.,121, 1703–1713.

  • Lee, C.-S., R. Edson, and W. M. Gray, 1989: Characteristics of north Indian Ocean tropical cyclone activity. Mon. Wea. Rev.,117, 407–426.

  • McBride, J. L., 1981: Observational analysis of tropical cyclone formation. Part I: Basic definition of data sets. J. Atmos. Sci.,38, 1117–1131.

  • ——, and R. M. Zehr, 1981: Observational analysis of tropical cyclone formation. Part II: Comparison of non-developing versus developing systems. J. Atmos. Sci.,38, 1132–1151.

  • ——, and T. D. Keenan, 1982: Climatology of tropical cyclone genesis in the Australian region. J. Climatol.,2, 13–33.

  • Montgomery, M. T., and B. F. Farrell, 1993: Tropical cyclone formation. J. Atmos. Sci.,50, 285–310.

  • Ramage, C. S., 1974: Monsoonal influences on the annual variation of tropical cyclone development over the Indian and Pacific Oceans. Mon. Wea. Rev.,102, 745–753.

  • Raymond, D. J., and H. Jiang, 1990: A theory for long-lived mesoscale convective systems. J. Atmos. Sci.,47, 3067–3077.

  • Reed, R. J., and E. E. Recker, 1971: Structure and properties of synoptic-scale wave disturbances in the equatorial western Pacific. J. Atmos. Sci.,28, 1117–1133.

  • Ritchie, E. A., 1995: Mesoscale aspects of tropical cyclone formation. Ph.D. dissertation, Monash University, 167 pp. [Available from Monash University, Wellington Rd., Clayton, Victoria 3168, Australia.].

  • Sadler, J. C., 1976: A role of the tropical upper troposphere in early season typhoon development. Mon. Wea. Rev.,104, 1266–1278.

  • ——, 1978: Mid-season typhoon development and intensity changes and the tropical upper tropospheric trough. Mon. Wea. Rev.,106, 1137–1152.

  • Shapiro, L. J., 1977: Tropical storm formation from easterly waves: A criterion for development. J. Atmos. Sci.,34, 1007–1021.

  • Unden, P., 1989: Tropical data assimilation and analysis of divergence. Mon. Wea. Rev.,117, 2495–2517.

  • Velasco, I., and J. M. Fritsch, 1987: Mesoscale convective complexes in the Americas. J. Geophys. Res.,92, 9591–9613.

  • Wallace, J. M., 1971: Spectral studies of tropospheric wave disturbances in the tropical western Pacific. Rev. Geophys. Space Phys.,9, 557–612.

  • Zehnder, J. A., 1991: The interaction of planetary-scale tropical easterly waves with topography: A mechanism for the initiation of tropical cyclones. J. Atmos. Sci.,48, 1217–1230.

  • Zehr, R. M., 1992: Tropical cyclogenesis in the western North Pacific. NOAA Tech. Rep. NESDIS 61, 181 pp. [Available from U.S. Department of Commerce, NOAA/NESDIS, 5200 Auth Rd., Washington, DC 20233.].

Fig. 1.
Fig. 1.

Locations relative to an idealized monsoon trough of tropical cyclones at the time of genesis for 1988–89.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 2.
Fig. 2.

Composites of the 200-hPa winds for the 1988 storm cases at (a) −72 h, (b) −48 h, (c) −24 h, and (d) time of tropical cyclogenesis. Maximum wind vector is 40 m s−1.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 3.
Fig. 3.

Same as in Figure 2 but for 1989 storm cases.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 4.
Fig. 4.

Composites of the 850-hPa winds for the 1988 storm cases at (a) −120 h, (b) −48 h, (c) −24 h, and (d) time of tropical cyclogenesis. Maximum wind vector is 10 m s−1.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 5.
Fig. 5.

Same as in Fig. 4 but for 1989 storm cases.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 6.
Fig. 6.

Wind anomaly fields for the 1988 850-hPa composites: (a) −48 h, (b) −24 h, and (c) time of genesis. Maximum wind vector is 10 m s−1.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 7.
Fig. 7.

Same as in Figure 6 but for 1989 850-hPa composites.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 8.
Fig. 8.

The 200–850-hPa vertical shear of the zonal wind at the time of genesis for the 1988 composite.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 9.
Fig. 9.

Composite of the 850-hPa wind for the 1989 nongenesis cases of the monsoon trough circulation. The center point is the confluence region of the monsoon trough. Maximum wind vector is 10 m s−1.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 10.
Fig. 10.

Same as in Fig. 9 but for the 200-hPa wind. Maximum wind vector is 40 m s−1.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 11.
Fig. 11.

Locations relative to the genesis point of axes of upper-level troughs present at the time of tropical cyclogenesis.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 12.
Fig. 12.

Locations relative to the confluence region of the monsoon trough of the closest point of upper-level troughs for the nongenesis cases.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 13.
Fig. 13.

Locations of preexisting tropical cyclones relative to the positions of new storms at the time of tropical cyclogenesis.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 14.
Fig. 14.

Time–longitude Hovmöller diagram of the zonal component of the 850-hPa wind along the genesis latitude for Typhoon Ruby (1988). Times are 1200 UTC.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 15.
Fig. 15.

Track of Typhoon Pat relative to the genesis location of Typhoon Ruby. Genesis occurred at 1200 UTC 20 October 1988.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 16.
Fig. 16.

Time–longitude Hovmöller diagram of the meridional component of the 850-hPa wind along the genesis latitude for Typhoon Pat (1988). Times are 1200 UTC.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Fig. 17.
Fig. 17.

Time–latitude Hovmöller diagram of the zonal component of the 850-hPa wind along the genesis longitude for Typhoon Skip (1988). Times are 0000 UTC.

Citation: Monthly Weather Review 125, 7; 10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2

Table 1.

Occurrence of upper-level troughs in individual cases. The total number of cases refers to the total number of genesis cases used in this study. The number of cases with upper-level troughs includes those cases with upper-level troughs to the northwest or west.

Table 1.
Table 2.

Number of genesis cases associated with preexisting storms, easterly waves, and other surges as observed in the 850-hPa time–longitude and time–latitude cross sections. Because some genesis cases were not associated with any of these features in the cross sections while others were associated with more than one, the sum of the number of cases associated with the various features does not add up to the total number of cases.

Table 2.
Save