Tropical Cyclone Formations in the South China Sea Associated with the Mei-Yu Front

2670 MONTHLY WEATHER REVIEW VOLUME 134

CHENG-SHANG LEE AND YUNG-LAN LIN Department of Atmospheric Sciences, National Taiwan University, Taipei, Taiwan

KEVIN K. W. CHEUNG National Science and Technology Center for Disaster Reduction, Taipei, Taiwan

(Manuscript received 17 February 2005, in final form 19 January 2006)

ABSTRACT

This study examines the 119 tropical cyclone (TC) formations in the South China Sea (SCS) during 1972–2002, and in particular the 20 in May and June. Eleven of these storms are associated with the weak baroclinic environment of a mei-yu front, while the remaining nine are nonfrontal. Seven of the 11 initial disturbances originated over land and have a highly similar evolution. Comparison of the frontal and nonfrontal formation shows that a nonfrontal formation usually occurs at a lower latitude, is more barotropic, develops faster, and possibly intensifies into a stronger TC. Six nonformation cases in the SCS are also identified that have similar low-level disturbances near the western end of a mei-yu front but did not develop further. In the nonformation cases, both the northeasterlies north of the front and the monsoonal southwesterlies are intermittent and weaker in magnitude so that the vorticity in the northern SCS does not spin up to tropical depression intensity. Because of the influence of a strong subtropical high, convection is suppressed in the SCS. The nonformation cases also have an average of 2–3 m sϪ1 larger vertical wind shear than the formation cases. A conceptual model is proposed for the typical frontal-type TC formations in the SCS that consists of three essential steps. First, an incipient low-level disturbance that originates over land moves eastward along the stationary mei-yu front. Second, the low-level circulation center with a relative vorticity maximum moves to the open ocean with the stationary front. Last, with strengthened northeasterlies, cyclonic shear vorticity continues to increase in the SCS, and after detaching from the stationary front, the system becomes a tropical depression.

1. Introduction end of the front, and a high pressure system was located to the north. From 0000 UTC 3 June to 0600 UTC 6 The mei-yu season in southeast Asia is from about June, the low pressure system moved eastward along mid-May to mid-June when a stationary front in the the mei-yu front but was still attached to it when the northern South China Sea (SCS) often occurs. During high pressure system to the north moved from land to May and June 2002 two tropical cyclones (TCs) formed sea (Figs. 1c,d). The low pressure system deepened and in the SCS associated with the mei-yu front, namely was declared a tropical depression by the Joint Ty- Tropical Depression 05W and Typhoon Noguri 07W. phoon Warning Center (JTWC) at 0000 UTC 6 June. Several days before the formation of Typhoon Noguri, Curvature in the cloud band of the system as seen in the a mei-yu front extended from Hainan Island to south- satellite imagery also suggests that the vorticity had al- ern Taiwan at 0000 UTC 3 June 2002 (Figs. 1a,b). A ready developed into a column structure. By 0600 UTC large area of low pressure was located at the western 8 June, TC Noguri had further intensified to 998 hPa and detached from the frontal system (Figs. 1e,f). The purpose of this study is to examine 1) how often this Corresponding author address: Cheng-Shang Lee, Department of Atmospheric Sciences, National Taiwan University, 1, Section type of TC formation occurs, and 2) the associated 4, Roosevelt Rd., 106 Taipei, Taiwan. mechanisms of formation. E-mail: [email protected] The mei-yu frontal system is part of the East Asia

MWR3221 OCTOBER 2006 L E E E T A L . 2671

FIG. 1. Surface weather maps and GMS-5 visible satellite imageries associated with Typhoon Noguri (in 2002) at (a), (b) 0600 UTC 3 Jun, (c), (d) 0600 UTC 6 Jun, and (e), (f) 0600 UTC 8 Jun 2002. monsoon system. After the summer monsoon starts ev- migration. When the subtropical high in the western ery year, organized precipitation associated with con- North Pacific (WNP) retreats to the east, there is often vection along the monsoon trough first occurs in the a subsequent frontal system approaching southeast SCS, and is sometimes stationary or has a northward Asia (Tao and Chen 1987). The persistent southwest-

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC 2672 MONTHLY WEATHER REVIEW VOLUME 134 erlies associated with the summer monsoon and the baroclinic disturbance into Diana was strongly modi- frontal easterlies create a general cyclonic flow envi- fied by the effects of latent heating. In the first stage, ronment in the northern SCS. Moreover, the onset of the low-level circulation is strengthened through the the summer monsoon is closely related to the eastward axisymmetrization of remote PV anomalies that are extension of the westerlies from the Bay of Bengal (Liu generated by condensational heating and then advected et al. 2002). Occasionally, disturbances can originate toward the incipient storm. After a warm-core vortex from these westerlies to become surface lows in the and spiral bands of convection begin to form, the mois- northern SCS. Chen and Chang (1980) showed that the ture in the core of the storm increases. Davis and mei-yu front and midlatitude fronts have similar struc- Bosart (2002) explored the model sensitivities in simu- tures in that they possess a large horizontal tempera- lating Diana. When the upper-level trough was re- ture gradient and vertical baroclinic structures. The ma- moved from the initial condition, cyclogenesis did not jor difference is that the mei-yu front is only weakly occur, which indicates the importance of appropriate baroclinic and there can be strong low-tropospheric coupling between the upper and lower levels in trans- vertical wind shear, which is especially true for the fron- forming the baroclinic system into a tropical depres- tal systems west of Japan (Ninomiya and Akiyama sion. 1992). In a further study, Davis and Bosart (2003) collected Lander (1996) examined TC formations during 1978– 10 cases of transformation from a STC to a TC during 94 associated with the so-called reverse-oriented mon- September–November in 2000 and 2001. A rapid de- soon trough, which is similar to the frontal orientation crease in the 200–900-hPa vertical wind shear occurred in the East Asia mei-yu season. However, Lander before the STC became a TC. On the other hand, those subtropical cyclones that did not transform to a TC (1996) did not explicitly identify the cases that originat- experienced vertical wind shear of about 15–20 m sϪ1. ed from the mei-yu front. Such mei-yu formations may Davis and Bosart (2003) also simulated the formation be related to the subtropical cyclone (STC), which He- of Hurricane Michael (in 2000) and found that the ver- bert and Poteat (1975) defined as a TC-like circulation tical wind shear had to decrease rapidly from 30 m sϪ1 that forms in the baroclinic subtropical environment to realize the warm-core structure. Therefore, the re- associated with a front. The cloud coverage in a STC duction of the vertical wind shear of the initial baro- usually measures 15° latitude across, which is larger clinic system is an essential requirement for its trans- than a typical TC. Similar to the naming convention for formation to a TC. tropical storms, a STC is called a subtropical storm Whereas the mei-yu front is a weak baroclinic system Ϫ1. when its maximum wind first exceeds 17 m s and occurs, on average, at a lower latitude than the Different statistics have been given on tropical cyclo- extratropical cyclones in the Atlantic, this study is ded- genesis associated with baroclinic or frontal systems. icated to identifying the mechanisms of transformation The results of Frank and Clark (1977) showed that less from baroclinic systems to TCs in the SCS. The orga- than 5% of all TC formations in the Atlantic occurred nization of this study is as follows. The data sources in close vicinity to a front or the east side of a westerly used in the study are discussed in section 2, and the trough. However, Gray (1998) indicated that about overall TC activity in the SCS and their large-scale forc- 25% of the global TC formations were not in a pure ing are examined in section 3. In section 4, the frontal- tropical environment. Bosart and Bartlo (1991) studied type TC formation in the SCS is defined, and it is com- Hurricane Diana (in 1984) in the Atlantic, and showed pared with the nonfrontal-type formation. It is also im- that the pre-Diana disturbance was also a STC that portant to understand the nonformation cases in which originated from a frontal system east of Florida. Bosart an incipient storm does not intensify; this is discussed in and Bartlo (1991) demonstrated an interaction between section 5. Finally, the mechanisms of frontal-type for- upper- and lower-level potential vorticity (PV) anoma- mation are summarized in section 6. lies. The PV anomaly associated with the upper-level trough was advected toward the low-level disturbance such that the entire system was able to evolve to a more 2. Data barotropic warm-core system. Davis and Bosart (2001) simulated the formation of Hurricane Diana using the Various data are utilized in this study. First, the cli- fifth-generation Pennsylvania State University– matology of TC formations in the SCS during 1972– National Center for Atmospheric Research (PSU– 2002 is established based on the best-track data from NCAR) Mesoscale Model (MM5) model and found the JTWC. Whereas the best-track data from 1950 are that the essential stage of transformation from a weak available, many of the records prior to 1972 begin with

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC OCTOBER 2006 L E E E T A L . 2673 tropical storm intensity (35 kt or 18 m sϪ1), and are not suitable for TC formation studies. Thus, only the records after 1972 are used. Detailed descriptions of the TCs are also extracted from the corresponding JTWC Annual Tropical Cyclone Reports. Surface features are defined from the daily weather charts of the Japan Meteorological Agency (JMA). When JMA charts are missing, the Taiwan Central Weather Bureau charts are used. In addition, the infra- red and visible satellite imageries from the Geostation- ary Meteorological Satellites (GMS) in the same 31-yr period are examined to compare with the surface features. The extended reconstructed sea surface temperature (ERSST) analyses from the National Oceanic and Atmospheric Administration (NOAA) with 2° latitude–longitude resolution are used (Smith and Reyn- olds 2003). The ERSST data have been statistically corrected (bias correction and error estimation) based on the Comprehensive Ocean–Atmosphere Data Set. The monthly mean interpolated outgoing longwave radiation (OLR) with 2.5° latitude–longitude resolution is also from NOAA, and the 6-hourly National Centers for Environmental Prediction (NCEP) reanalyses with the same resolution are used for the upper- level data. The SCS is defined as the region 0°–22°N, 105°– 120°E in this study (the area east of 120°E is called the FIG. 2. Latitudinal distribution (average per 2° latitude) of TC WNP). Any TCs entering the SCS from the WNP were formations in the SCS in May and June (solid), July–September omitted, so only TC formations associated with a fron- (dotted), and October–December (dashed). The abscissa indi- tal system in the SCS are considered. As in other TC cates percentage. formation studies (e.g., Cheung 2004), formation time is defined as the first time a TC reaches an intensity of 25 kt (ϳ13 m sϪ1) in the JTWC best-track data. between the two basins is in December in which the percentage of formations in the SCS is relatively larger (9% versus 5%). 3. Climatology of tropical cyclone formations in The TC formations in the SCS have a latitudinal the South China Sea variation during the season (Fig. 2). Formations during the first half of the season (May–September) are usu- a. Spatial and temporal distribution ally north of 15°N, while those in late season (October– A total of 119 (793) TC formations in the SCS December) are usually south of 15°N. This latitudinal (WNP) occurred during 1972–2002 with an annual av- variation is closely related to the meridional shift of the erage of 3.8 (25.6). Almost no formations occur in the monsoon shear line (where smaller vertical wind shear SCS from January to March, but the number increases is found) in a year (Cheung 2004), and also the en- rapidly to 16.8% (of the total number of formations in hanced convection in the tropical convergence zone the SCS) during May and June. The corresponding per- during cold surges related to the winter monsoon. centage for May and June in the WNP is only 9%, The interannual variation of TC formations in the which indicates that the two months of the mei-yu sea- SCS is also large compared with those in the WNP (Fig. son are particularly favorable for TC formations in the 3). No formations occurred in the SCS in 1991, while 10 SCS. Both the SCS and WNP have the highest monthly formations occurred in 1998. The latter is highly related percentages of formations in August (21% for the SCS to the La Niña during that year that persisted until early and 20% for the WNP), and then the formation rates 1999. For instance, the formation in January 1999 was decrease gradually until the end of the year. A contrast the first in that month since 1972. The enhanced con-

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC 2674 MONTHLY WEATHER REVIEW VOLUME 134

1979–2001, which is similar in magnitude to that of SST (negative because low OLR anomaly represents strong convection). The monthly mean SST and OLR anomalies (in the same regions defined above) are also correlated with the TC activity in the SCS (Fig. 4a). During the first four months of a year, the SST anomaly is still low, and OLR anomaly still high (minimum deep convection) in the SCS. The TC activity in the basin begins in May when both the SST and OLR anomalies change dramatically to be favorable for TC formation. These favorable conditions in May persist until September before the number of formations decreases. The mean SST is above 29°C in the SCS during May and June. The cold OLR anomaly well below 200 W mϪ2 is between the equator and the Borneo Island (latitudinal

FIG. 3. Interannual variation of the number of TC formations range of the island is about 3°S to about 5°N) from from 1972 to 2002 in the SCS (solid) and the WNP (dashed). January to April (Fig. 4b), the cold OLR anomaly shifts sharply after May to about 13°N, and stays at this latitude until September. This OLR migration is vective activity in the western (eastern) portions of the consistent with the latitudes of TC formations in the entire WNP basin tends to shift the mean positions of SCS during the summer (Fig. 2). Although the con- TC formations westward (eastward) substantially dur- vective clouds in the TCs certainly also contribute ing the La Niña (El Niño) (Wang and Chan 2002). The to the low OLR values, this should not be the dominat- La Niña (or post–El Niño) effect can also be seen in ing factor in the climatological values because the TCs 1973 and 1983. Therefore, the TC activity in the SCS last only a few days and occur at different times every and WNP are out of phase during the phases of the El year. Niño–Southern Oscillation. Perhaps a trend exists for an increasing number of formations in the SCS since c. Synoptic analyses the average number of cases during 1972–94 is only 2.81, while the average of 6.33 after 1994 is far higher The monthly mean wind fields from the NCEP re- than the climatological average of 3.8. Certainly, this analyses are consistent with the relatively active TC increasing trend might be related to interdecadal vari- formations in the SCS during May and June. During ability in TC activity (e.g., Ho et al. 2004) and should be winter, the dominant low-level (850 hPa) circulation subjected to further statistical verification. over the SCS is the high pressure center associated with the winter monsoon (not shown). Very often this anticyclone extends eastward from mainland China to b. Correlations with SST and OLR about 120°–125°E, which brings persistent northeaster- The interannual variability of TC activity in the SCS lies with negative relative vorticity over the SCS. In the can be partially explained by the corresponding vari- winter, positive 850-hPa relative vorticity is only found ability in SST and OLR in the area. For instance, the near the equator (e.g., near the west coast of the long-term mean of SST (during 1972–2002) in the re- Borneo Island). In May, southerlies and southwester- gion 8°–21°N, 110°–120°E is 27.8°C. The correlation lies associated with the summer monsoon start to be the coefficient between the SST anomalies (differences dominant circulation in the SCS (Fig. 5a). Positive 850- from the long-term mean) in this region and the num- hPa relative vorticity is also found at higher latitudes ber of TCs in the SCS is 0.55 during 1972–2001. How- (5°–10°N) such that TC formation is possible. In June, ever, this correlation coefficient changes substantially southwesterly flow covers almost the entire SCS, and in different decades: 0.4 in 1972–79; Ϫ0.6 in 1980–88; positive cyclonic vorticity associated with the south- and 0.9 in 1989–2001. The high correlation in the latest westerlies is common in the region (Fig. 5b). period is associated with the much above-average TC The midlevel (500 hPa) circulation depends mainly activity in the SCS. For OLR data averaged in the re- on the location of the subtropical high. During the win- gion 0°–22°N, 105°–120°E, the correlation between ter, the subtropical anticyclone is over the SCS such OLR anomaly and SCS TC activity is Ϫ0.5 during that the entire basin is associated with negative relative

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC OCTOBER 2006 L E E E T A L . 2675

Ϫ FIG. 4. (a) Monthly variation of averaged OLR anomaly (left vertical axis; W m 2), SST anomaly (right vertical axis; °C), and anomaly number of TC formations in the SCS (TC number; also using the left vertical axis) in the SCS. (b) Spatial distribution of OLR within the domain 0°–26°N, 105°–123°E from January to December. vorticity (not shown). In May, the subtropical high erlies induce large vertical wind shear in the region, center begins to retreat eastward to about 140°E, which is not favorable for development of any tropical which gives way to the southwesterlies from the west disturbances that might exist. However, an anticyclone (Fig. 5c). In June, the subtropical high shifts slightly builds over Indochina during May (Fig. 5e) and is lo- northward to about 20°N (Fig. 5d), which leads to a cated farther to the northwest in June (Fig. 5f). The confluent-type flow between the southwesterlies and divergent flow over the SCS associated with this anti- the subtropical high over the SCS, which is favorable cyclone during these months provides a favorable up- for TC formation. per-level environment for TC formation. Upper level (200 hPa) divergent flow is known to be favorable for TC formation. During the winter (not 4. Frontal-type formations shown), the Pacific anticyclone is centered at around 150°E such that southwesterlies are over the SCS. Among the 119 TC formations in the SCS during These strong southwesterlies over low-level northeast- 1972–2002, nine occurred in May and 11 in June. One of

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC 2676 MONTHLY WEATHER REVIEW VOLUME 134

FIG. 5. Monthly mean flow in (left) May and (right) June from 1972 to 2002 at (a), (b) 850; (c), (d) 500; and (e), (f) 200 hPa. Heavy contours in (a), (b) indicate wind speed greater than 6 m sϪ1 (interval: 2 m sϪ1). Shaded contours are (a)–(d) positive relative vorticity (interval: 0.5 ϫ 10Ϫ5 sϪ1) and (e), (f) divergence (interval: 0.2 ϫ 10Ϫ5 sϪ1). the most interesting features about these formations in June). To identify the mechanism for this type of for- May and June is that some of them are closely related mation, a TC formation is defined as a frontal-type to the quasi-stationary front at the northern part of the formation when the following conditions are satisfied SCS, which occurs in every mei-yu season (15 May–15 during the 4 days before the formation:

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC OCTOBER 2006 L E E E T A L . 2677

Ϫ 1) A closed surface isobar is located at the southwest TABLE 1. Maximum intensity (1 kt ϭ 0.5144 m s 1) and time end of the stationary front; taken to develop (when data are available) from a TCFA to a 2) The 925-hPa relative vorticity associated with the tropical storm (TS) for the (a) frontal-type and (b) nonfrontal- Ϫ5 Ϫ1 type formations during May and June 1972–2002. Cases A to G frontal zone is greater than 10 s and is oriented are the seven typical frontal-type formations used to compute the northeast–southwest; and composites in section 5 for comparing with the nonformation 3) A substantial north–south-oriented gradient of cases. equivalent potential temperature exists at 1000 hPa, (a) above the frontal zone. Max If all of these conditions are not met, the formation intensity Development time case is classified as a nonfrontal type formation. In to- Yr/month TC # Name (kt) TCFA → TS (h) tal, 11 frontal-type formations occurred in the 31-yr (A) 1974/6 TD 05w — 30 — period and the remaining nine cases were of the non- 1979/5 TD 05w — 30 — frontal type (Table 1), which indicates that the frontal 1981/6 TY 04w Ike 65 42 type is not uncommon during the mei-yu season. The (B) 1985/6 TD 04w — 30 — (C) 1986/5 TS 04w Mac 45 68 mean formation location of the 11 frontal-type cases is (D) 1987/6 TS 03w Ruth 35 36 at 18.5°N, 115.6°E (Fig. 6a). The mean motion of these (E) 1994/6 TS 05w Russ 55 17 TCs is toward the east-northeast, which is similar to (F) 1997/5 TS 05w Levi 45 54 those associated with a reverse-oriented trough forma- 2000/5 TD 03w — 30 — tion analyzed by Lander (1996). Among these systems, 2002/5 TD 06w — 25 — (G) 2002/6 TY 07w Noguri 85 66 only two (Ike in 1981 and Noguri in 2002) reached ty- Avg ϳ43 ϳ47 phoon intensity and five never reached tropical storm (b) intensity in their lifetimes. In addition, the average time interval for these TCs to develop from TC formation Max intensity Development time alert (TCFA) issued by the JTWC to tropical storm is Yr/month TC # Name (kt) TCFA → TS (h) 47 h, which indicates that they were not in an environment that was favorable for rapid intensification. Nev- 1972/6 TS 05w Mamie 50 24 1980/5 TS 06w Georgia 55 52.3 ertheless, the formation location of these frontal-type 1982/6 TS 06w Tess 35 60.5 TCs is close to cities in southeast Asia and could affect 1984/6 TS 01w Vernon 40 25.8 these cities in a short time after formation. Therefore, 1989/5 TY 04w Cecil 75 9.5 understanding their formation mechanisms is important 1990/6 TD 04w — 30 — for forecasters. 1996/5 TS 05w Cam 60 37.5 2000/5 TD 04w — 30 — The nine nonfrontal type cases in May and June had 2001/6 TY 05w Durian 75 15 a mean formation location of 14.0°N, 114.2°E, which is Avg 50 ϳ32 more than 4° latitude south of that of the frontal type. Moreover, the mean motion of these nine TCs was formation case. Therefore, only the composites with fixed northwestward because they were mainly steered by domain are presented here. The zero time of each case the easterlies of the subtropical high. On average, these Ϫ is the first time that V exceeded 25 kt (ϳ13 m s 1). nonfrontal TCs intensified more than the frontal ones max Three days before a frontal-type formation (Fig. 7a), [average maximum intensity reached was 50 kt (ϳ26 the 925-hPa flow in the northern SCS has an east–west- msϪ1)], and they also developed faster as the average oriented band of positive relative vorticity associated time interval from TCFA to tropical storm was 32 h with the monsoon trough and the surface front. South- compared with 47 h for the frontal type. Ϫ westerlies of up to 10 m s 1 are flowing into the SCS and meeting with northeasterlies associated with an an- a. Composites ticyclone over mainland China. From 48 to 36 h before To compare in detail the differences between frontal- formation (Fig. 7b), the synoptic situation remains simi- type and nonfrontal-type formations, composites of the lar. However, the vorticity maximum over the northern 11 frontal cases and the nine nonfrontal cases are com- SCS intensifies, the southwesterlies into the SCS are puted using the NCEP reanalyses. Because all of the 20 still strong, and the anticyclone to the north is still over cases formed near the same location (Fig. 6), the com- land. The surface front (not shown) is oriented north- posites with a fixed geographical domain have similar east–southwest, and is just south of Taiwan. characteristics to those composites with respect to the One day before formation (Fig. 7c), the southwest- system center (sea level pressure minimum) of each erlies increase in magnitude up to 12 m sϪ1, which fur-

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC 2678 MONTHLY WEATHER REVIEW VOLUME 134

FIG. 6. The best tracks and mean formation position (typhoon symbol) of the (a) frontal-type (18.5°N, 115.6°E) and (b) nonfrontal- type (14.0°N, 114.2°E) TCs in the SCS. The large arrows indicate their respective mean motions, and the insets show the formation positions of individual TCs. ther increases the vorticity maximum in the SCS. The trough, and the easterlies over the northern SCS are cold high to the north starts to move off the east coast associated with the subtropical high in the WNP. At of China (and so does the surface front), which brings 24 h before formation (Fig. 9b), both the southwester- some strong easterlies or northeasterlies to the north- lies associated with the monsoon trough and the eastern SCS. At the time of frontal-type formation (Fig. erlies to the north increase in magnitude, which gives 7d), the high has moved into the East China Sea, and rise to a strong vorticity center. The situation at 500 hPa the 925-hPa vorticity maximum has intensified as the (Figs. 9c,d) is similar to that at 925 hPa, which indicates tropical depression stage is achieved. that the development in this nonfrontal type of forma- At the midlevel (500 hPa), the dominant flows are tion is more barotropic. The subtropical high is shifted westerly and southwesterly over the SCS 3 days before to the western boundary of the WNP. The anticyclone formation (Fig. 8a). The subtropical high is east of at 200 hPa 2 days before formation (Fig. 9e) is similar to 140°E so that its influence over the SCS is weak. By the frontal-type composite; however, this anticyclone 36 h before formation (not shown), an area of stronger remains to the west (Fig. 9f) throughout the formation relative vorticity is located at the northern SCS because process without the development of a diffluent region of a developing short-wave trough. At the time of for- as in the frontal-type formation. mation (Fig. 8c), this short-wave trough further deep- Composites of the 1000-hPa equivalent potential ens in conjunction with the low-level system. Compari- temperature also show contrasts between the frontal son with the 925-hPa maps reveals the importance of and the nonfrontal formations. At 24 h before a frontal the low-level features in this frontal-type formation, and formation (Fig. 10a), the equivalent potential tempera- the baroclinicity of the system. At the 200-hPa level, an ture has a maximum in the northern SCS that is south anticyclone is located over northern Indochina three of a region with a large equivalent potential tempera- days before formation (Fig. 8b). By 36 h before forma- ture gradient. Although the situation for the nonfrontal tion (not shown), this anticyclone extends eastwards formations (Fig. 10b) is similar at this time, the maxi- into the northern SCS. At the formation time (Fig. 8d), mum temperature is further to the south where the the ridge further develops and leads to a diffluent flow development is occurring. At the formation time for the region over the SCS. Superposing divergence over the frontal type (Fig. 10c), the large equivalent potential low-level system is favorable for its intensification. temperature gradient in the northern SCS is just north The composites associated with the nonfrontal-type of the system center. In the nonfrontal formation case formation have distinct differences from the frontal (Fig. 10d), the gradient of equivalent potential tem- type. Two days before formation (Fig. 9a), the maxi- perature throughout the SCS is small. mum positive relative vorticity at 925 hPa is at a lower The magnitude of 200–850-hPa vertical wind shear latitude (about 15°N) associated with the monsoon (not shown) is about the same in the composites of

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC OCTOBER 2006 L E E E T A L . 2679

FIG. 7. Composite 925-hPa flow for frontal-type formations at (a) Ϫ72, (b) Ϫ48, (c) Ϫ24, and (d) 0 h. The heavy contours indicate wind speed greater than 6 m sϪ1 (interval: 2 m sϪ1) and the shaded contours are positive relative vorticity (interval: 0.5 ϫ 10Ϫ5 sϪ1). The typhoon symbols mark the mean formation positions. frontal-type and nonfrontal formations. As the mon- A–G in Table 1). Although the origin of the disturbance soon shear line shifts northward to about 15°N during for the remaining four cases is uncertain based on the May and June, low zonal wind shear exists over the available data, they were probably over the ocean. These northern SCS. The major contrast between frontal-type seven typical frontal-type TC formations during May and nonfrontal formations is in the meridional wind and June 1972–2002 have a unique formation mecha- shear, with northerly shear over the SCS for frontal- nism in close relationship to the development of the type formations and southerly shear for nonfrontal mei-yu front, which can be summarized in a conceptual type. That is, the frontal-type formations occur under model consisting of three essential steps (Fig. 11). the upper-level northerlies associated with the eastward 1) In the incipient step, a low-level disturbance over extension of the anticyclone over Indochina (Fig. 8d). land that originated to the west from cyclogenesis in The nonfrontal formations occur under weaker upper- a weakly baroclinic environment moves eastward level northerlies. along the stationary mei-yu front with a semiclosed surface isobar. b. Conceptual model 2) During the movement over sea step, the low-level Seven of the 11 frontal-type formation cases involved circulation center moves over the SCS along the sta- a disturbance cyclone that originated over land (labeled tionary front and with strengthened southwesterlies

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC 2680 MONTHLY WEATHER REVIEW VOLUME 134

FIG. 8. Same as in Fig. 7 but for (left) 500- and (right) 200-hPa flows at (a), (b) Ϫ72 and (c), (d) 0 h. Shaded contours are (a), (c) positive relative vorticity (interval: 0.5 ϫ 10Ϫ5 sϪ1) and (b), (d) divergence (interval: 0.2 ϫ 10Ϫ5 sϪ1). The typhoon symbols mark the mean formation positions.

now has a relative vorticity maximum and a closed and Holland indicates a region of convergence that ex- surface isobar. If the vertical wind shear over the tends eastward from the formation location and is con- disturbance decreases, the baroclinicity also weak- ducive to the development of sustained convection. For ens and vice versa. the frontal-type formations, this convergent region is 3) During the transition step, the northeasterlies along provided by the front, which then leads to the convec- the southeast China coast strengthen as the cold tive activity. The upper-level composite in Ritchie and high over China moves to the Yellow Sea, and thus Holland is also similar to the one for the frontal-type the cyclonic shear vorticity continues to increase formation (Fig. 8d) with a diffluent region aloft. Be- over the SCS. Deep convection associated with la- cause frontal-type formations in the SCS possess signa- tent heat release in the region reduces the Rossby ture characteristics that may be easily recognized, they radius of deformation of the system, and after de- may be treated as a special case of monsoon-shear line taching from the stationary front the system be- formation so that forecasters can estimate the potential comes a tropical depression. of such a TC formation in the region when similar conditions appear. As these frontal-type formations are embedded in the monsoon trough over the SCS, they may belong to 5. Formation versus nonformation the monsoon shear line formation pattern depicted in a. The nonformation cases Ritchie and Holland (1999) that accounts for 42% of all formation cases in the WNP (including the SCS). The Examination of weather maps near the time of for- composite of the monsoon shear line pattern in Ritchie mation of the seven typical cases of frontal-type forma-

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC OCTOBER 2006 L E E E T A L . 2681

FIG. 9. Same as in Fig. 7 but for (a), (b) 925-; (c), (d) 500-; and (e), (f) 200-hPa flows for nonfrontal-type formation at (a), (c), (e) Ϫ48 h and (b), (d), (f) Ϫ24 h. Shaded contours are (a)–(d) positive relative vorticity (interval: 0.5 ϫ 10Ϫ5 sϪ1) and (e), (f) divergence (interval: 0.2 ϫ 10Ϫ5 sϪ1). The typhoon symbols mark the mean formation positions. tions selected in the last section indicates that they all cases mentioned in this section refer exclusively to had the stationary mei-yu front accompanying the in- these seven cases. cipient surface low. Because the development in the The question is the following: Will a formation in the seven selected cases is highly similar, the formation SCS always occur whenever an incipient disturbance

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC 2682 MONTHLY WEATHER REVIEW VOLUME 134

FIG. 10. Composite 1000-hPa equivalent potential temperature (interval: 3 K) and mean formation position (typhoon symbol) for the (a), (c) frontal-type and (b), (d) nonfrontal-type formations at (a), (b) Ϫ24 and (c), (d) 0 h. moves from the west to near the mei-yu front? The 1000 km so that latent heat release from the mesoscale answer is negative because some surface low pressure system should have been very efficient in intensifying systems southwest of the mei-yu front did not eventu- the background disturbance. However, the system ally develop into tropical depressions. Yeh and Chen never developed into a tropical depression. To identify (2001) and Yeh et al. (2002) examined a mesoscale vor- these kind of nonformation systems, weather maps dur- tex in the northern SCS that formed along the mei-yu ing May and June 1972–2002 are reexamined. A non- front and lasted for about 4 days. The relative vorticity formation case is defined as a case where 4 days after a of the system intensified to 12.5 ϫ 10Ϫ5 sϪ1. The sys- surface low pressure disturbance is identified, the three tem’s Rossby radius of deformation decreased to about conditions in section 4 are satisfied (and that the closed

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC OCTOBER 2006 L E E E T A L . 2683

FIG. 11. Three steps in the conceptual model of TC formation associated with a mei-yu front. surface isobar lasts for at least 24 h but the system does the nonformation composites. For the seven typical for- not develop into a TD). Finally, the cases that persist mation cases, this reference time is, on average, about along the SCS are eliminated because they are easily 36 h before the formation time defined in section 4. affected by topography. Because the system identification conditions for the Six nonformation disturbances were identified in the formation and the nonformation systems are similar, 31-yr period according to the definition above (Table the composites several days before the reference time 2). The lifetime of these cases ranges from 2 days to also have little differences. Focus is thus put on times more than 4 days. Except for the system in 1975 that after the reference time. deepened to 998 hPa, the central pressures were slightly Because, by the zero reference time, the formation above 1000 hPa. The JMA classified four of the systems cases (not shown) have almost converted from a weakly as tropical depressions, but no TCFA was issued by the baroclinic to a more barotropic system, the composites JTWC. Most of the six nonformation systems moved of the seven typical formation cases have continuous slightly northward to south China, and the average time strong southwesterlies to the south of the incipient low period they stayed over the sea was about 2.5 days. The and easterlies to the north from 925 to 500 hPa (not surface features of these six cases are rather similar to shown). Divergent flow above the system starts to de- those of the formation cases. Some of them also have velop after the reference time, and the outflow struc- the high pressure system to the north that moved to the ture at 200 hPa is already clear by 48 h. Yellow Sea together with the mei-yu front. Therefore, The 925-hPa composite of the six nonformation cases the question is the following: What factors prohibited at the reference time has a vorticity maximum in the them from intensifying to the tropical depression stage? SCS and southwesterlies south of the system (Fig. 12a). However, the magnitude of the southwesterlies did not b. Composites increase in the next 24 h (Fig. 12b). Moreover, the cold Separate composites using the NCEP reanalyses for high to the north is not well defined and the accompa- the seven (typical frontal type) formation cases and the nying easterlies are relatively weak. Although a de- six nonformation cases are computed to identify the tached high pressure system is moving over the East prohibiting factors in the latter. For the nonformation China Sea, it is not as far east as in the frontal-type cases, the time when the closed surface isobar first formation. A 500-hPa short-wave trough is present in formed is used as the zero-time reference, and this ap- the SCS at the reference time (Fig. 12c), and a ridge is plies to the composites of the formation cases as well so located over Indochina such that at 24 h the short-wave that they do not have a higher intensity compared with trough becomes northwest–southeast-oriented (Fig.

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC 2684 MONTHLY WEATHER REVIEW VOLUME 134

TABLE 2. Lifetimes and minimum surface pressure attained by the nonformation cases during May and June 1972–2002. See section 5a for a definition of the nonformation cases.

No. Dates Lifetime (h) Min surface pressure (hPa) 1 0000 UTC 16 Jun–0000 UTC 19 Jun 1975 72 998 2 1200 UTC 5 Jun–0000 UTC 10 Jun 1983 108 1002 3 0000 UTC 1 Jun–0000 UTC 3 Jun 1990 48 1004 4 0600 UTC 26 May–1800 UTC 29 May 1998 84 1002 5 0000 UTC 14 Jun–1800 UTC 18 Jun 2000 114 1002 6 0600 UTC 27 May–0600 UTC 29 May 2001 48 1002

12d). At later times when the subtropical ridge contin- 6. Summary and conclusions ues to strengthen, southerlies persist in the SCS that are not favorable for further intensification. The 200-hPa The TC season in the SCS effectively starts in May as anticyclone (Fig. 12e) is at about the same location as in almost no TC formations occur in the first 4 months of the frontal-type formation at the reference time, and the year, and about 16.8% of all TCs form in May and this anticyclone extends farther eastward 24 h later June. The TC activity in the SCS correlates fairly well (Fig. 12f). However, the divergence over the SCS is not with the SST and OLR variations in the basin. In May particularly high. and June, areas with small OLR (less than 200 W mϪ2, The low-level equivalent potential temperature com- i.e., deep convection) and large SST (greater than posites of the nonformation cases are similar to those of 29°C) move to the northern SCS, which provides favor- the frontal formations (Figs. 10a,c), except the high able environmental conditions for TC formation in that equivalent potential temperature gradient regions are region. The synoptic flow also consists of persistent farther to the north (around 30°N) and the gradient in low-level monsoonal southwesterlies, the midlevel sub- the northern SCS is not particularly large. A maximum tropical high retreats eastward such that near souther- is not found as in Fig. 10 because the nonformation lies persist, and an upper-level anticyclone to the west cases do not intensify and go through a warming pro- over Indochina puts weak northerlies over the SCS. cess. This kind of synoptic flow provides an environment Aerial averages of several quantities can also distin- with positive low-level relative vorticity and weak ver- guish the environmental conditions in which typical tical wind shear, which are then suitable environmental frontal-type formation and nonformation are embed- conditions for TC formation. ded. The average 925-hPa relative vorticity within a Among the 119 TC formations in the SCS during 500-km radius centered on the surface pressure mini- 1972–2002, 20 were in May and June. In addition, 11 of mum for the formation cases maintains a magnitude these are associated with the weak baroclinic environ- between 2 ϫ 10Ϫ5 sϪ1 and 3 ϫ 10Ϫ5 sϪ1 after the ref- ment of a mei-yu front, while the remaining nine are erence time (Fig. 13a). Although the nonformation nonfrontal. Composites using NCEP reanalyses show cases have about the same magnitude of 925-hPa vor- that in the frontal-type formations, the low-level distur- ticity at the reference time, the vorticity then decreases bance typically forms over land and moves eastward as a result of the weakening southwesterlies to the along the mei-yu front to form a vorticity center over south and absence of strong easterlies to the north. The the northern SCS. At 200 hPa, a diffluent region over 200-hPa divergence associated with the formation case the northern SCS provides divergence. After forma- continues to be high several days after the reference tion, the mean motion of a frontal-type TC is still east- time, which is favorable for further intensification of ward because of the southwesterlies south of the mei-yu the tropical depression (Fig. 13b). In contrast, the di- front. In contrast, a nonfrontal TC formation is more vergence associated with the nonformation case is al- barotropic, develops faster, and possibly intensifies to a most zero throughout the period. The average 200–850- stronger TC. As the subtropical high is still dominant, hPa deep vertical wind shear for the nonformation the formation location is usually south of the frontal cases is on average 2–3msϪ1 larger than for the for- type. Moreover, the mean motion after formation is mation cases for several days after the reference time northwestward. (Fig. 13c). Although values of vertical wind shear of For contrast, six nonformation cases in the SCS dur- about 4–8msϪ1 are not large enough to cause dissipa- ing the period 1972–2002 are identified to distinguish tion, the probability of development is lower when the them from the typical frontal-type formation. These are vertical wind shear is higher. cases with similar low-level disturbances near the west-

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC OCTOBER 2006 L E E E T A L . 2685

FIG. 12. Same as in Fig. 7 but for (a), (b) 925-; (c), (d) 500-; and (e), (f) 200-hPa flows for the nonformation cases at (a), (c), (e) 0 and (b), (d), (f) 24 h. Shaded contours are (a)–(d) positive relative vorticity (interval: 0.5 ϫ 10Ϫ5 sϪ1) and (e), (f) divergence (interval: 0.2 ϫ 10Ϫ5 sϪ1). ern end of a mei-yu front, but without further develop- China coast because of the high pressure system north ment. Comparison of their respective composites re- of the mei-yu front moving to the East China Sea. In veals that the typical frontal-type formation has persis- the nonformation cases, both these northeasterlies and tent northeasterlies (for as long as 36 h) along the south the monsoonal southwesterlies are weaker and inter-

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC 2686 MONTHLY WEATHER REVIEW VOLUME 134

Frontogenesis calculations from this simulation suggest that the tilting term contributes to the weakening of the frontal structure, and indicates that an appropriate vertical shear environment may determine the formation or nonformation. Simulations of more of these frontal- type TC formations will be diagnosed in more detail in a later report.

Acknowledgments. Comments from Prof. Russell Elsberry of the Naval Postgraduate School and the two anonymous reviewers are much appreciated. The au- thors thank Ms. Erin Huang and Ms. Chia-Mei Huang for their help in manuscript preparation. This research is supported by the National Science Council of the Republic of China (Taiwan) under Grants NSC94-2111- M-002-019-AP1 and NSC93-2119-M-002-006-AP1.

REFERENCES Bosart, L. F., and J. A. Bartlo, 1991: Tropical storm formation in a baroclinic environment. Mon. Wea. Rev., 119, 1979–2013. Chen, G. T.-J., and C.-P. Chang, 1980: The structure and vorticity budget of an early summer monsoon trough (Mei-Yu) over southeastern China and Japan. Mon. Wea. Rev., 108, 942–953. Cheung, K. K. W., 2004: Large-scale environmental parameters associated with tropical cyclone formations in the western North Pacific. J. Climate, 17, 466–484. Davis, C. A., and L. F. Bosart, 2001: Numerical simulations of the Ϫ5 Ϫ1 genesis of hurricane Diana (1984). Part I: Control simulation. FIG. 13. (a) Averaged 925-hPa relative vorticity (10 s ), (b) 200-hPa divergence (10Ϫ5 sϪ1), and (c) 200–850-hPa vertical wind Mon. Wea. Rev., 129, 1859–1881. shear magnitude (m sϪ1) for the formation (dashed) and nonfor- ——, and ——, 2002: Numerical simulations of the genesis of mation (solid) cases. hurricane Diana (1984). Part II: Sensitivity of track and intensity prediction. Mon. Wea. Rev., 130, 1100–1124. ——, and ——, 2003: Baroclinically induced tropical cyclogenesis. mittent so that the vorticity in the northern SCS is Mon. Wea. Rev., 131, 2730–2747. smaller. Because of the influence of a strong subtropi- Frank, N. L., and G. Clark, 1977: Atlantic tropical systems of cal high, deep convection is suppressed in the SCS. Al- 1976. Mon. Wea. Rev., 105, 676–683. though a similar diffluent structure exists at the upper Gray, W. M., 1998: The formation of tropical cyclones. Meteor. level as in the typical frontal-type formation, it appears Atmos. Phys., 67, 37–69. Hebert, P. H., and K. O. Poteat, 1975: A satellite classification later in the development and divergence over the low- technique for subtropical cyclones. NOAA Tech. Memo. level disturbance is weaker. Whereas the typical fron- NWS SR-83, 25 pp. tal-type formations experience moderate vertical wind Ho, C.-H., J.-J. Baik, J.-H. Kim, and D.-Y. Gong, 2004: Interde- shear (about 5 m sϪ1), the nonformation cases experi- cadal changes in summertime typhoon tracks. J. Climate, 17, ence a large vertical wind shear (about 8 m sϪ1). 1767–1776. Lander, M., 1996: Specific tropical cyclone track types and un- Because of the low spatial and temporal data resolu- usual tropical cyclone motions associated with a reverse- tion, some important issues cannot be precisely deter- oriented monsoon trough in the western North Pacific. Wea. mined. For instance, the timing of the baroclinic-to- Forecasting, 11, 170–187. barotropic transition during the weakening of the fron- Lee, C.-S., and C.-H. Lee, 2002: A numerical simulation of the tal structure is not precisely known. The timing of when environmental momentum influences on typhoon formation. Proc. Fourth Conf. on East Asia and Western Pacific Meteo- the Rossby radius of deformation starts to decrease rology and Climate, Taipei, Taiwan, World Scientific, 261–271. such that latent heat can efficiently build up the warm Liu, Y., J. C. L. Chan, J.-G. Mao, and G.-X. Wu, 2002: The role of core also has to be examined further. Some successful Bay of Bengal convection in the onset of the 1998 South numerical simulations might provide suitable fields for China Sea summer monsoon. Mon. Wea. Rev., 130, 2731– further diagnosis (Lee and Lee 2002). Preliminary 2744. Ninomiya, K., and T. Akiyama, 1992: Multiscale feature of Baiu, simulations of Typhoon Noguri (in 2002) indicate the the summer monsoon over Japan and East Asia. J. Meteor. strengthening of northeasterlies associated with the Soc. Japan, 70B, 467–495. high north of the surface front play an important role. Ritchie, E. A., and G. J. Holland, 1999: Large-scale patterns as-

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC OCTOBER 2006 L E E E T A L . 2687

sociated with tropical cyclogenesis in the western Pacific. affect tropical storm activity over the western North Pacific. Mon. Wea. Rev., 127, 2027–2043. J. Climate, 15, 1643–1658. Smith, T. M., and R. W. Reynolds, 2003: Extended reconstruction Yeh, H.-C., and G. T.-J. Chen, 2001: Evolution of a meso-vortex of global sea surface temperatures based on COADS data along the Mei-yu front over South China Sea. Proc. Int. Conf. (1854–1997). J. Climate, 16, 1495–1510. on Mesoscale Meteorology and Typhoon in East Asia, Taipei, Tao, S.-Y., and L.-X. Chen, 1987: A review of recent research on Taiwan, National Science Council, Central Weather Bureau, the East Asia summer monsoon in China. Monsoon Meteo- and National Taiwan University, 359–363. rology, C. P. Chang and T. N. Krishnamurti, Eds., Oxford ——, ——, and W.-T. Liu, 2002: Kinematic characteristics of a University Press, 60–92. mei-yu front detected by the QuikSCAT oceanic winds. Mon. Wang, B., and J. C. L. Chan, 2002: How strong ENSO events Wea. Rev., 130, 700–711.

Unauthenticated | Downloaded 10/05/21 06:44 AM UTC