Seasonal Modulation of Tropical Intraseasonal ... - AMS Journals

Download PDF
0 downloads 0 Views 2MB Size Report
Comment
Dec 15, 2011 - Center for Monsoon System Research, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing ... (Manuscript received 12 November 2010, in final form 3 June 2011) ... the western North Pacific Ocean (WNP) is investigated in three periods of the ... 2009; Chand and Walsh 2010; Mao and.
15 DECEMBER 2011

HUANG ET AL.

6339

Seasonal Modulation of Tropical Intraseasonal Oscillations on Tropical Cyclone Geneses in the Western North Pacific PING HUANG Center for Monsoon System Research, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing

CHIA CHOU Research Center for Environmental Changes, Academia Sinica, and Department of Atmospheric Sciences, National Taiwan University, Taipei

RONGHUI HUANG Center for Monsoon System Research, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing (Manuscript received 12 November 2010, in final form 3 June 2011) ABSTRACT The seasonal modulation of tropical intraseasonal oscillation (TISO) on tropical cyclone (TC) geneses over the western North Pacific Ocean (WNP) is investigated in three periods of the WNP TC season: May–June (MJ), July–September (JAS), and October–December (OND). The modulation of the TISO–TC geneses over the WNP is strong in MJ, while it appears weaker in JAS and OND. In MJ, TISO propagates northward via two routes, the west route through the South China Sea and the east route through the WNP monsoon trough region, which are two clustering locations of TC geneses. TISO can synchronously influence most TC geneses over these two regions. In JAS, however, the modulation is out of phase between the monsoon trough region and the East Asian summer monsoon region, as well as the WNP subtropical high region, as a result of further northward propagation of TISO and scattered TC geneses. The TISO–TC genesis modulation in each individual region is comparable to that in MJ, although the modulation over the entire WNP in JAS appears weaker. In OND, TISO has a stronger influence on TC geneses west than east of 1508E because TISO decays and its convection center located at the equator is out of the TC genesis region when propagating eastward into east of 1508E. Midlevel relative humidity is the primary contribution to the modulations of TISO on the genesis environment, while vorticity could contribute to the modulation over the subtropics in JAS.

1. Introduction The western North Pacific Ocean (WNP) is the most active basin of tropical cyclone (TC) geneses because of the favorable environment, warm sea surface temperature (SST), and great relative humidity, for TC geneses in this region, (e.g., Gray 1968; McBride 1995). Over the WNP, TC geneses can be modulated by the variations of the large-scale circulation, such as low-level confluence (Holland 1995), monsoon trough (Harr and Elsberry 1995; Chen et al. 1998; Camargo et al. 2007b,c; Chen and Huang 2008), monsoon gyre (Chen et al. 2004), and tropical depression–type disturbance (Chen and Huang 2009).

Corresponding author address: Dr. Ping Huang, P.O. Box 2718, Bei-Er-Tiao 6#, Zhong-Guan-Cun, Beijing 100190, China. E-mail: [email protected] DOI: 10.1175/2011JCLI4200.1 Ó 2011 American Meteorological Society

At an intraseasonal time scale, Gray (1979) first found that TC geneses tend to alternately occur in clusters with 1–2 weeks, following a similar period of quiescence. Since then, the intraseasonal variation of TC formations is considered to be connected with large-scale tropical intraseasonal oscillations (TISO) over the WNP (Nakazawa 1986; Sui and Lau 1992). Around 70% of the WNP TCs form during the active period of the WNP TISO (Table 1). The impact of TISO on TC geneses over ocean basins has been successively revealed (e.g., Maloney and Hartmann 2000, 2001; Hall et al. 2001; Bessafi and Wheeler 2006; Camargo et al. 2009; Chand and Walsh 2010; Mao and Wu 2010). The large-scale conditions during the westerly phase of TISO are much more favorable for TC geneses than during the easterly phase (Liebmann et al. 1994; Maloney and Dickinson 2003; Kim et al. 2008; Ko and Hsu 2009).

6340

JOURNAL OF CLIMATE

VOLUME 24

TABLE 1. The statistics of TC geneses in four TISO categories during three periods. Categories where genesis numbers are statistically enhanced (suppressed) at 90%, 95%, 99%, and 99.9% significance levels are indicated by *, **, ***, and **** (^, ^^, ^^^, and ^^^^), respectively. The Avg, Total No., and Percentage rows show the average TGR during TISO, the total number of TC geneses during TISO, and the percentage of the total TC genesis number during TISO to the total number during each season, respectively. ESR is the difference of maximum TGR and minimum TGR relative to the average TGR. MJ

JAS

Phases

Total

SCS MT

WNP MT

Total

MT

EASM

WNPSH

OND

1&2 3&4 5&6 7&8 Avg Total No. Percentage ESR

1.85^^^ 14.18**** 11.83**** 3.17^^ 6.67 87 71.35 1.85

0.37^^ 4.00*** 4.66**** 0.95 1.86 28 82.35 2.31

1.48^^ 10.18**** 7.17* 2.22^^ 4.81 59 67.05 1.81

16.05^ 22.89* 27.93**** 15.84^ 19.86 361 66.73 0.62

6.98^^^^ 18.79*** 23.94**** 8.75^^^ 14.17 256 65.47 1.20

2.09** 0.22^ 1.17 0.95 1.05 19 65.52 1.79

6.98*** 3.89 2.82 6.15* 4.38 86 65.47 0.95

5.87^^^ 12.56* 12.57* 5.99^^^ 10.07 142 51.08 0.67

The locations of TC geneses are clustered near the cyclonic circulation during the TISO westerly phase, together with an enhanced monsoon trough and a moisture confluent zone, while TC geneses are infrequent and poorly organized in the TISO easterly phases because of unfavorable conditions (Ko and Hsu 2009). Besides the circulation, midlevel relative humidity anomalies associated with TISO make the largest contributions to the TISO–TC genesis modulation (Camargo et al. 2009). Although the impact mechanisms of TISO on TC genesis are independent on the season, Kim et al. (2008) argued that the TISO–WNP TC genesis modulation has a significant seasonal variation and that the modulation in June and July is much stronger than that in August and September. This seasonal variation of the modulation has not been explained yet. Actually, the seasonal variation of TISO has not been adequately considered when investigating the TISO–TC genesis modulation in previous studies. During the WNP TC genesis season, May–December, the propagation of TISO shows a significant seasonal variation. In boreal winter, TISO mainly propagates eastward and is commonly called the Madden–Julian oscillation (MJO; Madden and Julian 1971; Zhang 2005); In boreal summer, TISO propagates northward and northwestward in the South China Sea (SCS) and the Philippine Sea (Lau and Chan 1986; Wang and Rui 1990; Hsu and Weng 2001; Kemball-Cook and Wang 2001; Tsou et al. 2005; Chou and Hsueh 2010). The propagation of TISO in early summer is also different from that in late summer (Kemball-Cook and Wang 2001). Whether this seasonal variation of TISO during the TC season can be connected with the stronger modulation in early summer suggested by Kim et al. (2008) is unknown. The purpose of this paper is to reinvestigate the TISO– TC genesis modulation based on a more accurate description of the WNP TISO and to explain the reason of

the weaker TISO–TC genesis modulation during the boreal fall. The propagation of TISO in different periods during May–December is separately analyzed for a further investigation of the seasonal variation of the TISO– TC genesis modulation. The data and methodology are described in section 2. The seasonal variation of the TC genesis location and frequency over the WNP as well as the seasonal variation of the propagation of TISO are investigated in section 3. Section 4 presents the statistical results on the TISO–TC genesis modulation in different periods of the WNP TC season. Finally, conclusions are drawn in section 5.

2. Data and methods a. Data Convection associated with TISO is based on the interpolated outgoing longwave radiation (OLR) from the National Oceanic and Atmospheric Administration (NOAA), which are daily data on a 2.58 3 2.58 grid from January 1979 to December 2008 (Liebmann and Smith 1996). Other daily environmental data, such as winds, relative humidity, air temperature, and sea level pressure, are from the National Centers for Environmental Prediction (NCEP) reanalysis 2 from January 1979 to December 2008 (Kanamitsu et al. 2002). The 30–90-day Lanczos filtering is performed on the daily data (hereafter as ‘‘filtered’’ data versus the raw ‘‘unfiltered’’ data), which are used to composite the environmental fields during the active TISO period. The International Best Track Archive for Climate Stewardship (IBTrACS, which is available online at http://www. ncdc.noaa.gov/oa/ibtracs/) TC dataset over the WNP is used to study the number and location of TC genesis from 1979 to 2008 (Knapp et al. 2010). The genesis time and location are defined as the time and location for the first

15 DECEMBER 2011

HUANG ET AL.

record of a TC track. Fewer than 20 TCs forming over continents are ignored as error records.

b. Methods 1) DETERMINATION OF THE TISO PHASES In previous studies, the real-time multivariate MJO (RMM) index obtained from the Center for Australian Weather and Climate Research was used to determine the phase and amplitude of the MJO (Wheeler and Hendon 2004). The RMM index is based on the averages of 158S–158N and does not depend on the season. It describes the predominant intraseasonal oscillations propagating eastward from the Indian Ocean. Although some northward-propagating information can be observed in around 208N based on the RMM index (Wheeler and Hendon 2004), the RMM index cannot accurately describe the northward propagation of TISO over the WNP when TISO propagates northward north of 308N in the boreal summer. Therefore, for distinguishing different propagations during various seasons, the EOF analysis is performed on the filtered OLR over different domains for three periods: May–June (MJ), July– September (JAS), and October–December (OND). The classification of these periods is based on the seasonal characteristics of general circulation, the frequency and location of TC geneses, and the propagation of TISO over the WNP (with details in section 3). A zonal range from 1008E to the date line is chosen to exclude the signals of TISO in the Indian Ocean and other regions (Wang and Rui 1990; Kemball-Cook and Wang 2001). A meridional range of 108S–308N is chosen to emphasize the Northern Hemisphere convection in MJ and JAS because of the northward propagation of TISO, while a symmetric meridional range from 208S to 208N is chosen in OND. Before the EOF analysis, a long-term mean of OLR is removed, and the 30–90-day Lanczos filtering is performed, as in previous works (e.g., Matthews 2000; Kim et al. 2008). The principal components (PCs) of EOF are normalized by one standard deviation of each component, and the standard deviations are multiplied to the corresponding spatial modes. The first two EOF modes display a quadrature phase relationship explaining a similar variance and representing a propagating wave phenomenon, which combines these two modes (Zhang and Hendon 1997). The amplitude and phase of TISO can be expressed by [PC12(t) 1 PC22(t)]1/2 and tan21[PC2(t)/ PC1(t)], respectively. The day when its amplitude is greater than 1 is defined as an active TISO day. The entire cycle of TISO is divided into eight phases using the method described in Fig. 7 of Wheeler and Hendon (2004). Phases 1 and 8 are close to the first leading EOF

6341

pattern (Figs. 2a, 2c, and 2e), while phases 2 and 3 are close to the EOF2 pattern (Figs. 2b, 2d, and 2f). Considering the relationship with TC geneses, phase 1 is more similar to phase 2 than to phase 8 (not shown). Thus, the entire cycle of TISO is divided into four categories, phases 1–2, 3–4, 5–6, and 7—8.

2) QUANTIFICATION OF TC GENESES AND STATISTICAL SIGNIFICANT TESTS

The number of TC geneses per 100 days during a period or category, the TC genesis rate (TGR), is used to measure the frequency of TC geneses. Because the TC genesis number is normalized by time, it can be used to compare the TC genesis frequency among various TISO categories, which have different days. Larger (smaller) TGR in a category displays enhanced (suppressed) TC geneses during this category. Moreover, an index to represent the enhancement–suppression ratio (ESR) of the TISO–TC genesis modulation is defined as ESR 5 (TGRmax 2 TGRmin)/TGRavg, where TGRmax and TGRmin are the maximum and minimum TGR of four categories in one period, respectively, and TGRavg is the average of TGR in this period. ESR is also normalized and can be used to compare the TISO–TC genesis modulation among different periods. The Student’s t test is applied to calculate the statistical significance of the composites of environmental fields for different TISO categories. A statistical test for the significant anomalies of the TC genesis frequency in each TISO category is performed with a null hypothesis that the distribution of TC geneses is uniform, following Hall et al. (2001) and Kim et al. (2008).pThe relevant test ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi statistic is given by Z 5 (P 2 Pe )/ Pe (100 2 Pe )/N , where Pe and P are the expected and observed TC genesis number per 100 days in each TISO category, respectively. Here, N is the number of days for a particular TISO category. The test statistic Z follows a standard normal distribution and is tested using a two-tailed test with the critical values of Z 5 61.65, 61.96, 62.58, and 63.29 at the 90%, 95%, 99%, and 99.9% confidence levels, respectively.

3) GENESIS POTENTIAL INDEX To assess the total variations of environment associated with TISO, the genesis potential index (GPI) is used, which combines the most important variables related to the TC geneses, such as midlevel relative humidity, vertical wind shear, and vorticity (Gray 1979). GPI is developed by Emanuel and Nolan (2004) and some details are given in Camargo et al. (2007a). It successfully replicates the climatological annual cycle and spatial distribution of TC geneses in each ocean domains. GPI is defined

6342

JOURNAL OF CLIMATE

as GPI 5 j105 Vortj3/2 (RH/50)3 (PI/70)3 (1 1 0:1Vshear )22 , where Vort is absolute vorticity at 850 hPa in inverse seconds, RH is relative humidity at 600 hPa in percent, PI is potential intensity in meters per second, and Vshear is vertical wind shear between 200 and 850 hPa in meters per second. PI represents potential TC intensity at each grid (Bister and Emanuel 2002), and it is calculated based on sea surface temperature and pressure, as well as the vertical profiles of temperature and specific humidity. The weekly NOAA optimum interpolation SST V2 data from 1982 to 2008 (Reynolds et al. 2002) and monthly NOAA extended reconstructed SST V3b data from 1979 to 1981 (Smith et al. 2008) are temporally interpolated into daily SST data. A 7-day-running averaging is performed on the interpolated daily SST, together with the daily reanalysis as in Camargo et al. (2009). Since the running averaging but not a bandpass filtering is performed on these variables whose long-term mean is not removed, GPI includes both the intraseasonal variability and the climatological background of environmental variables. On the other hand, there is no effective method to obtain the location and frequency of anomalous TC geneses, so the collected TC geneses for various phases of TISO must include the climatological and abnormal geneses. Therefore, the GPI including the climatological background is more advantaged than the OLR anomalies to indicate the anomalous TC geneses during the evolution of TISO. A revised GPI with three climatological variables and one varied variable (3c1v_GPI) is used to obtain the contribution of each of four variables—absolute vorticity, vertical wind shear, potential intensity, and midlevel relative humidity—in the TISO–TC genesis modulation as in Camargo et al. (2007a, 2009). The 3c1v_GPI contributed by a variable is calculated by using the observed value of this variable and the long-term climatology of the other three variables. This method is useful to quantitatively estimate the relative importance of different factors in GPI (Camargo et al. 2009).

3. Seasonal variations of TC geneses, background environment, and propagations of TISO over the WNP a. TC geneses and background environment TC geneses have distinct seasonal variations accompanied by changes in background environment (e.g., Gray 1979). Early summer (MJ; Fig. 1a) is the transition season from inactive TC geneses to active TC geneses. In MJ, GPI is weak and TC geneses are relatively rare. The distribution of TC geneses cannot extend north of 108N over the WNP. Both GPI and the TC genesis location are clustering in two regions, the SCS and the southeastern

VOLUME 24

FIG. 1. The large-scale circulations (streamlines), GPI (shadings), and locations of TC geneses (points) in three periods, (a) MJ, (b) JAS, and (c) OND.

Philippine Sea. The former is the SCS monsoon trough (SCS MT) region, in which the monsoon westerlies from the Indian Ocean converge with the trade southeasterlies. The latter is the convergence region of the cross-equatorial southerlies from northern Australia and the trade southeasterlies, in which the WNP monsoon trough (WNP MT) is not apparent during this period (Chou et al. 2009). The two clustering TC genesis regions can be divided by 1208E (the dashed dark line in Fig. 1a); a total of 38 TCs form in the SCS, while 84 are in the southeastern Philippine Sea. In late summer and early fall, JAS, which is the major period of the WNP TC geneses, the SCS MT is mature and the WNP MT forms in the eastern Philippine Sea. The WNP MT is northwest–southeast oriented from Taiwan into the far southeastern Philippine Sea around 1558E and 58N (Chou et al. 2009). Major TC geneses also approximately distribute northwest–southeast along the monsoon trough (e.g., Gray 1968; Ding et al. 1977; Zehr

15 DECEMBER 2011

HUANG ET AL.

6343

FIG. 2. (left to right) The two leading EOF modes of filtered OLR anomalies in three periods (a),(b) MJ, (c),(d) JAS, and (e),(f) OND. Unit: W m22. The percentage of the explained variance of total filtered OLR in the corresponding period is shown at the top-right corner of each subplot.

1992). However, considerable TCs form in the WNP subtropical high (WNPSH) region, though the corresponding high pressure and anticyclonic circulation are unfavorable for TC geneses compared to the monsoon trough. There are some TCs forming in the East Asian summer monsoon (EASM) region due to abundant water vapor in the subtropics around 308N transported by the mature EASM. Overall, the JAS GPI is much higher and more widely expands from 58N to 358N, compared with the clustering GPI in MJ. The wide TC genesis region can be divided into the MT, EASM, and WNPSH regions, as shown by the dashed dark lines in Fig. 1b. During late fall and early winter (OND), the monsoon trough retreats and becomes some gyres around 108N. GPI is uniformly situated from the Maritime Continent to the date line around 58–158N. TCs are also uniformly distributed, similar to GPI.

b. Propagation of TISO Similar to the distribution of TC geneses and background environment, the propagation of TISO also has a significant seasonal variation during the WNP TC

season from summer to early winter. The dominant signals of TISO mainly propagate eastward during boreal winter (Madden and Julian 1971). In boreal summer, however, the WNP TISO propagates not only eastward but also northward into the subtropics (Chen and Murakami 1988; Wang and Rui 1990; Kemball-Cook and Wang 2001). Based on the seasonal variations of TC geneses and background environment over the WNP, together with the seasonal variations of the propagation of TISO, the WNP TC season from May to December can be divided into three periods, MJ, JAS, and OND. The regional EOF analysis is individually performed on the filtered OLR for each period. Two leading spatial modes and their variance contributions are shown in Fig. 2. In MJ, the two leading EOF modes (Figs. 2a and 2b) explain 16.24% and 13.25% of the variances of the filtered OLR, respectively. EOF1 exhibits positive convection anomalies over the Maritime Continent and negative anomalies over the SCS, the Philippine Sea, and northeast of Papua New Guinea; EOF2 exhibits positive convection anomalies south of the SCS and north of Papua New Guinea and negative anomalies north of

6344

JOURNAL OF CLIMATE

the SCS. During the second period (JAS), about 20.11% and 16.20% of the filtered OLR variances are explained by EOF1 and EOF2, respectively (Figs. 2c and 2d). The convection anomalies in EOF1 and EOF2 are similar to the two leading EOF modes in MJ but more horizontally oriented. It shows that intraseasonal convection anomalies in JAS mainly propagate northward. In OND, EOF1 is associated with negative convection anomalies over the WNP, while EOF2 displays positive anomalies over the Maritime Continent and negative anomalies west of the date line (Figs. 2e and 2f). EOF1 (EOF2) explains 20.35% (14.76%) of the filtered OLR variances in OND. The lead–lag correlation between PC1 and PC2 for each period is shown in Fig. 3. The maximum correlations of PC1 and PC2 occur when PC1 lead PC2 by around 10 days, while the strongest negative correlations occur when PC1 lag PC2 around 10 days. The correlation coefficients are up to around 0.5–0.6, which are close to the result in Wheeler and Hendon (2004). Thus, the oscillation described by EOF1 and EOF2 in all seasons should be a typical TISO with a period of around 40 days.

VOLUME 24

FIG. 3. The lead–lag correlation coefficient of PC1 and PC2 for three periods, MJ, JAS, and OND. Positive values of abscissa represent days by which PC1 leads PC2.

4. Modulation of TISO on TC geneses a. The early summer: MJ The statistics of TC geneses in three periods are shown in Table 1. Around 71% of TCs in MJ form during active TISO. The composites of the filtered OLR and 850-hPa winds during four categories are shown in Fig. 4. The composited GPI is also shown in Fig. 5 to assess the

entire modulation on background environment for TC geneses by TISO. In phases 1&2, positive convection anomalies occur in the Maritime Continent (Fig. 4a). From phases 1&2 to phases 3&4, positive convection anomalies in the Maritime Continent propagate along two directions, eastward into far southeast of the Philippine Sea and northward into the SCS (Fig. 4b). Thus,

FIG. 4. Composites of filtered OLR anomalies (shadings; W m22) and 850-hPa wind anomalies (vectors) for four TISO categories in MJ at the 95% significance level, based on the Student’s t test.

15 DECEMBER 2011

HUANG ET AL.

6345

FIG. 5. Composites of GPI (contours, interval 1) for four TISO categories in MJ.

two positive convection anomaly centers form south of the SCS and the southeastern Philippine Sea, which are two major TC genesis regions in MJ (Fig. 1). Cyclonic circulation anomalies, which coincide with positive convection anomalies, are also located on the SCS and the Philippine Sea, and GPI is enhanced in these phases (Fig. 5b). So, significantly more (at 99.9% confidence level) TC geneses with TGR up to 14.18 are induced. Although some TCs locate outside the significantly positive convection anomaly region, locations of TC geneses are entirely consistent with the distribution of high GPI while the TC geneses are infrequent over the region with low GPI (Fig. 5). This is attributed to the fact that GPI combines more factors related to TC geneses, not just the OLR, and includes the climatological background and the intraseasonal variations associated with TISO. From phases 3&4 to 5&6, two isolated positive convection anomaly centers over the southern SCS and the southeastern Philippine Sea propagate northeastward and northwestward, respectively, and converge around the Philippine Sea in phases 5&6 (Fig. 4c). Cyclonic circulation anomalies locate around areas with positive convection anomalies. Under the similar favorable environment as in phases 3&4, significantly more (at 99.9% confidence level) TCs form around the Philippine Sea with a high TGR of 11.83. Compared to two clustering TCs centers in phases 3&4, locations of TCs are clustered over the Philippine Sea due to the convergent convection anomalies. While positive convection anomalies are dominant over the Philippine Sea, negative convection anomalies appear over the Maritime Continent in phases 5&6 (Fig. 4c). Similar to the positive convection phases of TISO, negative convection anomalies also propagate southeast

of the Philippine Sea and the SCS in phase 7&8; two isolated negative convection anomalies converge around the Philippine Sea in the next phases, 1&2. TC geneses in phases 7&8 with TGR of 3.17 and in phases 1&2 with TGR of 1.85 are greatly suppressed at 95% and 99% confidence level, respectively. The ESR of the modulations in MJ is up to 1.85, which means that the difference of the TC genesis probability between positive convection phases 3&4 and negative phases 7&8 is around 1.85 times to the averaged TC genesis probability. ESR at 1.85 shows a dramatic modulation of TISO on TC geneses in MJ. It could be attributed to the climatological distributions of TC geneses and propagation of TISO in MJ. During this season, TC geneses cluster over the SCS and southeast of the Philippine Sea (Fig. 1a). The two clustering TC geneses coincide well with the two propagating routes of TISO (Fig. 2). The sensitive background environment for TC geneses over these two regions can be dramatically changed by TISO. Much high GPI in positive convection anomaly phases indicates a favorable environment for TC geneses (3&4 and 5&6; Figs. 5a and 5d), while low GPI in negative phases shows a suppressed environment for TC geneses (1&2 and 7&8; Figs. 5b and 5c). In consequence, the dramatically changed background environment induces a great TISO–TC genesis modulation with ESR up to 1.85. Additionally, the TISO–TC genesis modulation in the SCS and southeast of the Philippine Sea is not completely synchronous when two regions are separately considered as the SCS MT and WNP WP regions. The strongest impacts on TC geneses in the WNP MT occur during phases 3&4 and lead the impacts in the SCS MT, which peak in phases 5&6 (Table 1). It could be attributed to the fact that the latitude of the WNP MT region is much

6346

JOURNAL OF CLIMATE

VOLUME 24

FIG. 6. Differences of composited GPI in MJ between positive (phases 3&4 and 5&6) and negative (phases 1&2 and 7&8) convection anomaly categories of TISO. GPI is computed with seasonally varied (a) absolute vorticity, (b) vertical wind shear, (c) potential intensity, and (d) relative humidity, respectively, along with the climatology of three other variables.

lower than that of the SCS MT region, and thus TISO may arrive earlier in the WNP MT region than in the SCS MT region. ESR in the SCS MT is up to 2.31, which indicates an intense TISO–TC genesis modulation in this region, while the modulation in the WNP MT is relatively weak with ESR of 1.81. In Camargo et al. (2009), midlevel relative humidity is believed to be the most important factor impacting TC geneses during MJO. The 3c1v_GPIs for four variables are also calculated to investigate the importance of various factors in the TISO–TC genesis modulation. Changes in each 3c1v_GPI are represented by differences in each 3c1v_GPI between the enhancement categories (phases 3&4 and 5&6) and the suppression categories (phases 1&2 and 7&8). Figure 6 shows the difference between the enhancement and suppression categories for various 3c1v_GPIs. Midlevel relative humidity is a primary contributor to the modulation (Fig. 6d). The contribution of 3c1v_GPI from vorticity is secondly, and it is larger in the SCS than in the WNP MT. When convection anomalies propagate northward into the SCS, the vorticity anomalies associated with TISO would be larger than the response to convection anomalies related to MJO near the equator. Vertical wind shear tends to be out of phase with GPI, indicating that it negatively contributes to the modulation of TISO. The differences of GPI that are associated with potential intensity are very weak.

b. The late summer and early fall: JAS Around 67% of TCs form during active TISO in JAS (Table 1). The propagation of TISO during four

categories in JAS is presented by composites of the filtered OLR and wind anomalies (Fig. 7). The variation of GPI is shown in Fig. 8. Convection anomalies jump from the Indian Ocean into the low-latitude western Pacific, and then almost synchronously propagate northward through the SCS and the WNP routes as shown in Wang and Rui (1990). The favorable conditions for TC geneses also move northward, accompanied by the propagation of TISO. In phases 1&2, positive convection anomalies appear in the tropical western Pacific (Fig. 7a). They develop and move northward to around 108N (Fig. 7b) from phases 1&2 to 3&4 (Fig. 7b). The cyclonic circulation anomalies appear in the MT regions, accompanied by positive convection anomalies. Under these favorable conditions, GPI is high in the MT region and TC formations are enhanced over this region (Fig. 8b). When TISO propagates northward via the two routes and gathers at the Philippine Sea in phases 5&6, strong positive convection anomalies, as well as strong GPI, appear and enhance TC formations in this region (Figs. 7c and 8c). While positive convection anomalies dominate the Philippine Sea, negative convection anomalies appear over the Maritime Continent in phases 5&6. Negative convection anomalies propagate northward into the MT region in phases 7&8 and farther into the Philippine Sea in the next phases, 1&2. Since the low GPI, which combines the anomalies of various variables such as anticyclonic circulation anomalies and negative convection anomalies (Figs. 7a and 7d), dominates the MT region (Figs. 8a and 8d), TC formations over the MT region are suppressed in phases 1&2 and 7&8 (Figs. 7a

15 DECEMBER 2011

HUANG ET AL.

6347

FIG. 7. As in Fig. 4, but for JAS.

and 7d). Similar to the situation in MJ, TC genesis locations do not match well with positive convection anomalies, but with high GPI. Generally, there are significantly more (fewer) TC geneses during phases 3&4 and 5&6 (1&2 and 7&8) of TISO in the WNP (Table 1). Although the enhancement or suppression of TC geneses is significant during all categories, the ESR in JAS is only 0.62. This lower ESR relative to the ESR of 1.85 in MJ, implies that the TISO–TC genesis modulation in JAS is much weaker than that in MJ when the WNP is considered as a whole.

Positive convection anomalies in phases 5&6 can continuously move northward into Taiwan, and westerly anomalies also move northward to 208N in phases 7&8 and 1&2 (Figs. 7d and 7a), although they decay, relative to those in phases 5&6. The favorable conditions associated with TISO also propagate northward into high latitudes; thus there are two relatively high GPI centers around 258N, the EASM, and WNPSH regions (Figs. 8d and 8a). On the other hand, negative convection anomalies in phases 3&4 and 5&6 (Figs. 7b, 7c, 8b, and 8c) of TISO can also propagate northward into the EASM and

FIG. 8. As in Fig. 5, but for JAS.

6348

JOURNAL OF CLIMATE

VOLUME 24

FIG. 9. As in Fig. 6, but for JAS.

WNPSH regions, while positive convection anomalies dominate the MT region. There are synchronously opposite convection anomalies associated with TISO between low and middle latitudes due to the further northward propagation and short meridional wavelength of TISO. The statistics of the modulations in the MT, EASM, and WNPSH regions are shown in Table 1. As expected, TC geneses in the MT region are significantly more during phases 3&4 and 5&6 but fewer during phases 1&2 and 7&8. In contrast to the MT region, there are opposite modulations in the EASM and WNPSH regions, in which TC geneses are significantly enhanced during phases 1&2 and 7&8 but suppressed during phases 3&4 and 5&6. The ESR of 1.2 in the MT region is much higher than the ESR of 0.62 over the whole WNP since the opposite modulations in the subtropics are removed. The ESR in the EASM and WNPSH regions is up to 1.79 and 0.95, respectively. The ESR over each individual region in JAS is comparable with the ESR in MJ, which means that the TISO–TC genesis modulation in JAS is as strong as that in MJ. The apparent weaker modulation in JAS over the whole WNP is due to the wider impacts of TISO in JAS and the out-of-phase modulation between the MT and subtropical regions. Similar to MJ, midlevel relative humidity anomalies still have the most important contributions (Fig. 9d). The 3c1v_GPI contributed to by relative humidity in the MT region is out of phase to that in the EASM and WNPSH regions, due to the northward propagation of TISO (Fig. 9d). The vorticity also has contribution to the modulation over the subtropical EASM region. The 3c1v_GPI contributed to by vorticity is reversed in the

EASM and MT regions (Fig. 9a). It means that the tightly coupling pattern of the circulation and convection anomalies contributes more at midlatitudes than near the equator. In the WNPSH region, vertical wind shear also has a weak positive contribution during phases 1&2 and 7&8 (Fig. 9 b). Easterly shear anomalies in the WNPSH region during phases 1&2 and 7&8 can weaken the climatological westerly shear (Figs. 7b and 7c).

c. The late fall and early winter: OND In OND, around 51% of TCs form during active TISO (Table 1). TISO mainly propagates eastward at low latitudes and could be treated as MJO during boreal winter (Fig. 10). The eastward-propagating TISO also has a weak southward-propagating component. West of 1508E, convection anomalies mainly locate around 108N, while maximum convection anomalies locate at the equator when they propagate eastward to east of 1508E. Along the eastward migration path of convection anomalies, the westerly (easterly) anomalies appear, accompanied by negative (positive) OLR anomalies. The cyclonic circulation anomalies are not apparent in contrast to the coupling pattern of the circulation and convection anomalies at midlatitudes in MJ and JAS. Thus, it is reasonable to divide TISO into westerly and easterly phases in OND (Liebmann et al. 1994; Kim et al. 2008). In the westerly phase (3&4 and 5&6), large GPI, which indicates favorable conditions for TC geneses over the low-latitude WNP (Figs. 11b and 11c), and more TC geneses appear. In contrast, negative convection anomalies induce fewer TC geneses in the easterly phases (1&2 and 7&8) of TISO (Figs. 11a and 11d). ESR in OND is 0.67. As in MJ and JAS, TC geneses coincide more with

15 DECEMBER 2011

HUANG ET AL.

6349

FIG. 10. As in Fig. 4, but for OND.

GPI, related to the evolution of TISO, than with convection anomalies (Figs. 10 and 11). Since a half wavelength of TISO in OND (around 60 longitudinal degrees) is shorter than the width of the zonal uniform distribution of TC geneses, the modulation of TISO must be different over the WNP in one category. Figure 12 shows meridional mean (58N to158N) OLR anomalies during one TISO cycle and zonal locations of TC geneses. TC geneses almost synchronously migrate eastward, accompanied by convection anomalies. The ratio of the TC genesis number with positive to negative convection anomalies is up to 2.54. This ratio is

greater than the ratio of TC geneses during the westerly to easterly phase of TISO, which is 2.14. Moreover, TISO has stronger modulations on TC geneses west than east of the WNP. The ratio of TC geneses between positive and negative convection anomalies west of 1508E (45/12 5 3.75) is much larger than that east of 1508E (39/21 5 1.86). The different modulation could be attributed to the propagation of TISO. The TISO signal is stronger west of the WNP but decays east of the WNP. And maximum convection anomalies associated with TISO move southward, out of the TC genesis center east of 1508E (Figs. 2c and 10).

FIG. 11. As in Fig. 5, but for OND.

6350

JOURNAL OF CLIMATE

VOLUME 24

5. Conclusions

FIG. 12. The meridional averages (between 58 and 158N) of composited filtered OLR anomalies (contours, interval: 4 W m22) for 1 TISO cycle and the zonal location of TC geneses. At the top left, 45/12 is the number of TC geneses located at negative/positive OLR anomalies west of 1508E with the ratio of 3.75, and, at the top right, 39/21 is for the number east of 1508E with the ratio of 1.86.

For the four variables contributing to background environment, relative humidity anomalies have a predominant contribution (Fig. 13d) and vorticity anomalies are secondary (Fig. 13a). Vertical wind shear has a weak reversed contribution (Fig. 13b), while potential intensity is insignificant (Fig. 13c). This result is consistent with the findings of Camargo et al. (2009) on the modulations of MJO.

In this study, the modulation of TISO on the WNP TC geneses during a long TC season from May to December was reinvestigated based on a more accurate description of the WNP TISO. A regional and seasonal EOFs analysis was used to examine the propagation of TISO in three periods, early summer (MJ), late summer and early fall (JAS), and late fall and early winter (OND). The IBTrACS TC track data was used to study the locations and frequency of TC geneses. The variation of the TC genesis potential index (GPI), which combines the most important factors for TC geneses, was also examined during the evolution of TISO. This GPI includes both the intraseasonal variations and the climatological genesis environment. It is suitable for discussing the relationship between the variation of the environment and TC geneses, because the anomalous and climatological TC geneses are hardly separated. In early summer MJ, TISO propagates northward via two routes. The west route is from Sumatra into the SCS, and the east route is from northeast of Papua New Guinea into the midlatitude WNP. The circulation and convection anomalies associated with TISO also propagate via these two routes. This propagation is consistent with previous studies (Wang and Rui 1990; Kemball-Cook and Wang 2001). TC geneses also compactly locate in the SCS and southeast of the Philippine Sea in MJ. Therefore, TISO, together with the seasonal variation of large-scale environment can greatly modulate TC geneses. TC geneses are enhanced (suppressed) during positive (negative) convection anomaly phases of TISO. The enhancement– suppression ratio (ESR) of the modulation in MJ is up to 1.85. The TC geneses locations coincide well with high GPI

FIG. 13. As in Fig. 6, but for OND.

15 DECEMBER 2011

6351

HUANG ET AL.

associated with TISO, although the TC locations do not match well with positive convection anomalies, especially when maximum convection anomalies are out of the climatological TC genesis region. Midlevel relative humidity anomalies associated with TISO have a primary contribution to the modulation as discussed in previous study (Camargo et al. 2009). For vorticity anomalies, although they contribute little to the modulation near the equator as in Camargo et al. (2009), they have a much stronger contribution in the subtropics than in the low latitudes. In JAS, there also is significant enhancement (suppression) of TC geneses during positive (negative) convection anomaly phases of TISO. The TISO–TC genesis modulation in JAS is stronger than the result during August and September discussed in Kim et al. (2008). It could be attributed to the reasonable description of TISO in the present study based on the separation for three periods and the regional EOF analysis for each period. However, this modulation in the JAS over the whole WNP still appears weaker than that in MJ. The apparent weaker modulation over the whole WNP is due to the outof-phase modulation between low latitudes, the MT regions, and midlatitudes, the EASM and WNPSH regions. In JAS, TCs dispersedly locate over a wide range of the WNP, even over the midlatitude EASM and WNPSH regions; and TISO also can propagate northward into these higher-latitude regions. Because of the short meridional wavelength of TISO, there are out-of-phase convection anomalies associated with TISO in the lowlatitude MT region and the midlatitude EASM and WNPSH regions. This opposite-phase relationship between the MT and EASM regions is consistent with previous studies (e.g., Chen et al. 2000). Therefore, when the WNP TCs are considered as a whole, the modulation of the TISO–TC geneses over the MT region seems to be weaker. Individually, the TISO–TC genesis modulation in the MT, EASM, or WNPSH regions is comparable to that in MJ, so the modulation of TISO on TC geneses in JAS can be as strong as that in MJ. The TC geneses locations coincide well with high GPI but not negative OLR anomalies during various phases of TISO, similar to in MJ. Midlevel relative humidity has predominant contributions, while vorticity has secondary contributions in the EASM region. Vertical wind shear has positive contributions to TC geneses in the WNPSH region, but negative contributions at low latitudes (Camargo et al. 2009). In OND, TISO mainly propagates eastward at low latitudes and can be treated as MJO, which is divided into westerly and easterly phases. The westerly anomalies induce more TC geneses than the easterly anomalies. Moreover, since the TISO signal is stronger over west

of the WNP and decays over east of the WNP and maximum convection anomalies move southward out of the TC genesis center east of 1508E, the TISO–TC genesis modulation is greater west than east of 1508E. Relative humidity anomalies also have a predominant contribution and vorticity anomalies have much weaker contributions. Overall, TC geneses are closely related to the evolution of TISO in all three periods. However, due to the seasonal variations of the TC genesis distribution, background environment, and propagation of TISO, the modulations of TISO on TC geneses are varied in different periods. Acknowledgments. The authors thank Dr. Guanghua Chen for useful comments on the manuscript. We also thank two anonymous reviewers for their constructive advice. This research was supported financially by the National Natural Science Foundation of China (Grant 40921160379), the National Basic Research Program of China (Grant 2010CB950403), the Special Scientific Research Project for Public Interest (Grant 201006021), the National Natural Science Foundation of China (Grant 40905024), and ‘‘Strategic Priority Research Program’’ of the Chinese Academy of Sciences (Grant XDA05090401). Dr. Chou was supported by the National Science Council Grant NSC98-2745-M-001-005-MY3.

REFERENCES Bessafi, M., and M. Wheeler, 2006: Modulation of South Indian Ocean tropical cyclones by the Madden–Julian oscillation and convectively coupled equatorial waves. Mon. Wea. Rev., 134, 638–656. Bister, M., and K. Emanuel, 2002: Low frequency variability of tropical cyclone potential intensity. 1. Interannual to interdecadal variability. J. Geophys. Res., 107, 4801, doi:10.1029/ 2001JD000776. Camargo, S. J., K. A. Emanuel, and A. H. Sobel, 2007a: Use of a genesis potential index to diagnose ENSO effects on tropical cyclone genesis. J. Climate, 20, 4819–4834. ——, A. W. Robertson, S. J. Gaffney, P. Smyth, and M. Ghil, 2007b: Cluster analysis of typhoon tracks. Part I: General properties. J. Climate, 20, 3635–3653. ——, ——, ——, ——, and ——, 2007c: Cluster analysis of typhoon tracks. Part II: Large-scale circulation and ENSO. J. Climate, 20, 3654–3676. ——, M. C. Wheeler, and A. H. Sobel, 2009: Diagnosis of the MJO modulation of tropical cyclogenesis using an empirical index. J. Atmos. Sci., 66, 3061–3074. Chand, S., and K. Walsh, 2010: The Influence of the Madden– Julian oscillation on tropical cyclone activity in the Fiji region. J. Climate, 23, 868–886. Chen, G., and R. Huang, 2008: Influence of monsoon over the warm pool on interannual variation on tropical cyclone activity over the western North Pacific. Adv. Atmos. Sci., 25, 319–328.

6352

JOURNAL OF CLIMATE

——, and ——, 2009: Interannual variations in mixed Rossby– gravity waves and their impacts on tropical cyclogenesis over the western North Pacific. J. Climate, 22, 535–549. Chen, T.-C., and M. Murakami, 1988: The 30-50-day variation of convective activity over the western Pacific Ocean with emphasis on the northwestern region. Mon. Wea. Rev., 116, 892–906. ——, S. P. Weng, N. Yamazaki, and S. Kiehne, 1998: Interannual variation in the tropical cyclone formation over the western North Pacific. Mon. Wea. Rev., 126, 1080–1090. ——, M.-C. Yen, and S.-P. Weng, 2000: Interaction between the summer monsoons in East Asia and the South China Sea: Intraseasonal monsoon modes. J. Atmos. Sci., 57, 1373–1392. ——, S. Y. Wang, M. C. Yen, and W. A. Gallus, 2004: Role of the monsoon gyre in the interannual variation of tropical cyclone formation over the western North Pacific. Wea. Forecasting, 19, 776–785. Chou, C., and Y.-C. Hsueh, 2010: Mechanisms of northwardpropagating intraseasonal oscillation—A comparison between the Indian Ocean and the western North Pacific. J. Climate, 23, 6624–6640. ——, L.-F. Huang, L. Tseng, J.-Y. Tu, and P.-H. Tan, 2009: Annual cycle of rainfall in the western North Pacific and East Asian sector. J. Climate, 22, 2073–2094. Ding, Y., H. Fan, Q. Xue, and G. Chen, 1977: Preliminary study on coinstantaneous developments of multiple typhoons within the intertropical convergence zone (in Chinese). Chin. J. Atmos. Sci., 2, 89–98. Emanuel, K. A., and D. S. Nolan, 2004: Tropical cyclone activity and the global climate system. Preprints, 26th Conf. on Hurricanes and Tropical Meteorology, Miami, FL, Amer. Meteor. Soc., 10A.2. [Available online at http://ams.confex.com/ams/ 26HURR/techprogram/paper_75463.htm.] Gray, W., 1968: Global view of the origin of tropical disturbances and storms. Mon. Wea. Rev., 96, 669–700. ——, 1979: Hurricanes: Their formation, structure and likely role in the tropical circulation. Meteorology Over the Tropical Oceans, D. B. Shaw, Ed., Royal Meteorological Society, 155–218. Hall, J., A. Matthews, and D. Karoly, 2001: The modulation of tropical cyclone activity in the Australian region by the Madden–Julian oscillation. Mon. Wea. Rev., 129, 2970–2982. Harr, P., and R. Elsberry, 1995: Large-scale circulation variability over the tropical western North Pacific. Part I: Spatial patterns and tropical cyclone characteristics. Mon. Wea. Rev., 123, 1225–1246. Holland, G., 1995: Scale interaction in the western Pacific monsoon. Meteor. Atmos. Phys., 56, 57–79. Hsu, H.-H., and C.-H. Weng, 2001: Northwestward propagation of the intraseasonal oscillation in the western North Pacific during the boreal summer: Structure and mechanism. J. Climate, 14, 3834–3850. Kanamitsu, M., W. Ebisuzaki, J. Woollen, S.-K. Yang, J. J. Hnilo, M. Fiorino, and G. L. Potter, 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 1631–1643. Kemball-Cook, S., and B. Wang, 2001: Equatorial waves and air– sea interaction in the boreal summer intraseasonal oscillation. J. Climate, 14, 2923–2942. Kim, J., C. Ho, H. Kim, C. Sui, and K. Seon, 2008: Systematic variation of summertime tropical cyclone activity in the western North Pacific in relation to the Madden-Julian oscillation. J. Climate, 21, 1171–1191. Knapp, K. R., M. C. Kruk, D. H. Levinson, H. J. Diamond, and C. J. Neumann, 2010: The International Best Track Archive

VOLUME 24

for Climate Stewardship (IBTrACS). Bull. Amer. Meteor. Soc., 91, 363–376. Ko, K.-C., and H.-H. Hsu, 2009: ISO modulation on the submonthly wave pattern and recurving tropical cyclones in the tropical western North Pacific. J. Climate, 22, 582–599. Lau, K.-M., and P. H. Chan, 1986: Aspects of the 40–50-day oscillation during the northern summer as inferred from outgoing longwave radiation. Mon. Wea. Rev., 114, 1354–1367. Liebmann, B., and C. Smith, 1996: Description of a complete (interpolated) outgoing longwave radiation dataset. Bull. Amer. Meteor. Soc., 77, 1275–1277. ——, H. Hendon, and J. Glick, 1994: The relationship between tropical cyclones of the western Pacific and Indian Oceans and the Madden–Julian oscillation. J. Meteor. Soc. Japan, 72, 1–412. Madden, R. A., and P. R. Julian, 1971: Detection of a 40–50-day oscillation in the zonal wind in the tropical Pacific. J. Atmos. Sci., 28, 702–708. Maloney, E., and D. Hartmann, 2000: Modulation of eastern north Pacific hurricanes by the Madden–Julian oscillation. J. Climate, 13, 1451–1460. ——, and ——, 2001: The Madden–Julian oscillation, barotropic dynamics, and North Pacific tropical cyclone formation. Part I: Observations. J. Atmos. Sci., 58, 2545–2558. ——, and M. Dickinson, 2003: The intraseasonal oscillation and the energetics of summertime tropical western North Pacific sypnotic-scale disturbance. J. Atmos. Sci., 60, 2153–2168. Mao, J. Y., and G. X. Wu, 2010: Intraseasonal modulation of tropical cyclogenesis in the western North Pacific: A case study. Theor. Appl. Climatol., 100, 397–411. Matthews, A., 2000: Propagation mechanisms for the Madden– Julian oscillation. Quart. J. Roy. Meteor. Soc., 126, 2637–2651. McBride, J., 1995: Tropical cyclone formation: Global Perspectives on Tropical Cyclones. World Meteorological Organization Rep. TCP-38, WMO/TD-No. 693, 63–105. Nakazawa, T., 1986: Intraseasonal variations of OLR in the tropics during the FGGE year. J. Meteor. Soc. Japan, 64, 17–34. Reynolds, R. W., N. A. Rayner, T. M. Smith, D. C. Stokes, and W. Wang, 2002: An improved In situ and satellite SST analysis for climate. J. Climate, 15, 1609–1625. Smith, T. M., R. W. Reynolds, T. C. Peterson, and J. Lawrimore, 2008: Improvements to NOAA’s historical merged land– ocean surface temperature analysis (1880–2006). J. Climate, 21, 2283–2296. Sui, C., and K. Lau, 1992: Multiscale phenomena in the tropical atmosphere over the western Pacific. Mon. Wea. Rev., 120, 407–430. Tsou, C., P. Hsu, W. Kau, and H. Hsu, 2005: Northward and northwestward propagation of 30-60-day oscillation in the tropical and extratropical western North Pacific. J. Meteor. Soc. Japan, 83, 711–726. Wang, B., and H. Rui, 1990: Synoptic climatology of transient tropical intraseasonal convection anomalies: 1975–1985. Meteor. Atmos. Phys., 44, 43–61. Wheeler, M., and H. Hendon, 2004: An all-season real-time multivariate MJO index: Development of an index for monitoring and prediction. Mon. Wea. Rev., 132, 1917–1932. Zehr, R. M., 1992: Tropical cyclogenesis in the western North Pacific. NOAA Tech. Rep. NESDIS 61, 181 pp. Zhang, C., 2005: Madden-Julian Oscillation. Rev. Geophys., 43, RG2003, doi:10.1029/2004RG000158. ——, and H. Hendon, 1997: Propagating and standing components of the intraseasonal oscillation in tropical convection. J. Atmos. Sci., 54, 741–752.