Summer monsoon onset in the subtropical western North Pacific

Click Here for Full Article GEOPHYSICAL RESEARCH LETTERS, VOL. 36, L18810, doi:10.1029/2009gl040168, 2009 Summer monsoon onset in the subtropical western North Pacific Chi-Hua Wu, 1 Wen-Shung Kau, 1 and Ming-Dah Chou 1 Received 23 July 2009; revised 24 August 2009; accepted 31 August 2009; published 25 September 2009. [1] Monsoon onset in July over the subtropical western North Pacific (SWNP, 15 25 N; 130 150 E) is investigated by focusing on the westward movement of cloudy region east of the SWNP, which is associated with high vorticity in the upper troposphere. At least in seven of the 22 years between 1985 and 2006, the arrival of clouds at the SWNP from the east coincides with a significant change in the upper tropospheric circulation and a rapid northeastward extension of strong convections from the tropical western North Pacific, which essentially is monsoon onset. Before the monsoon onset, the sea surface temperature (SST) increases, but winds remain divergent over the SWNP. Right after the monsoon onset, winds turn convergent and convections enhance, leading to a rapid decrease of surface heating and SST. It is suggested that westward-moving upper-level disturbances might trigger onset of monsoon in July when low-level atmospheric conditions favor development of deep convections. Citation: Wu, C.-H., W.-S. Kau, and M.-D. Chou (2009), Summer monsoon onset in the subtropical western North Pacific, Geophys. Res. Lett., 36, L18810, doi:10.1029/2009gl040168. 1. Introduction [2] Climatologically, the summer monsoon onset in the subtropical western North Pacific (SWNP, 15 25 N; 130 150 E) occurs in mid-summer, which is much later than the onset in India and the South China Sea [LinHo and Wang, 2002]. The strong land-sea thermal contrast is a dominant force that initiates and maintains the South Asian monsoon. In contrast, the SWNP monsoon occurs over the vast oceanic warm pool, where the meridional sea surface temperature (SST) gradient is weak. The forcing due to the strong temperature contrast occurring in South Asia is absent in the SWNP. [3] The summer monsoon onset in the SWNP is generally characterized by a rapid deepening of the monsoon trough (MT) and a quick extension of strong convective region from the tropical western North Pacific to the SWNP [Ueda and Yasunari, 1996]. Wu and Wang [2001] demonstrated that the precipitation of summer monsoon over the western North Pacific began in the South China Sea in mid- May and extended abruptly to the southwestern Philippine Sea in early to mid-june, and finally penetrated to the SWNP around mid-july. On the other hand, there are 30- to 60-day intra-seasonal oscillations (ISO) [Hsu and Weng, 2001] propagating northward and northwestward in the 1 Department of Atmospheric Sciences, National Taiwan University, Taipei, Taiwan. Copyright 2009 by the American Geophysical Union. 0094-8276/09/2009GL040168$05.00 South China Sea and the Philippine Sea. It is likely that the rapid monsoonal transition is a result of the interaction between the smooth seasonal evolution of solar heating and the phase-locked ISO [Wu and Wang, 2001]. [4] Ueda and Yasunari [1996] suggested that warming of the Philippine Sea in early July might be related to the abrupt northward shift of convections. The processes involving local atmosphere-ocean interactions were further investigated by Wu [2002]. Recently, Ueda et al. [2009] used general circulation model (GCM) simulations to evaluate the impact of SST on rapid monsoon transitions. The GCM simulations showed that the impact of SST on the SWNP monsoon onset is weak when compared to solar radiation, land memory, and atmospheric transient effects. [5] Sato et al. [2005] studied the Marcus Convergence Zone (MCZ) formation, which followed the formation of the upper cold-low (UCL) in the subtropical North Pacific (20 25 N and 150 170 E). High vorticity air mass associated with the UCL migrated westward, reaching Marcus Island (24 18.0 0 N, 153 58.0 0 E) in mid-july, which might have implications on the monsoon onset in the SWNP. Sakamoto and Takahashi [2005] studied the weakening and cut-off processes of the UCL. More recently, Lu et al. [2007] investigated the relationship between the intraseasonal circulation in the mid-latitude upper tropospheric and convections in the SWNP. They suggested that the latter might be affected by the former. [6] In this study, we further studied the relationship between the mid-ocean upper tropospheric circulation and the rapid shift of convection in the SWNP using satelliteinferred SST, outgoing longwave radiation (OLR), highlevel thick clouds (HTC), and surface heat fluxes, as well as the model analysis of winds. The emphasis of this study is the thermal and dynamical conditions prior to and during the monsoon onset in the SWNP. 2. Data and Monsoon Onset Date [7] Data used in this study cover a period of 22 years from 1985 to 2006. Those data include: (a) OLR at the top of the atmosphere from National Oceanic and Atmospheric Administration (NOAA) [Liebmann and Smith, 1996]; (b) Upper- and lower-tropospheric winds from Version 2 of the National Centers for Environmental Prediction reanalysis (NCEP-R2) [Kanamitsu et al., 2002]; (c) Cloud amount from the International Satellite Cloud Climatology Project (ISCCP) D1 data set [Rossow and Schiffer, 1999]; (d) SST from the NOAA Optimum Interpolation (OI) analysis [Reynolds et al., 2007]; (e) The 10-m wind speed and surface latent and sensible heat fluxes from the Objectively Analyzed Air-Sea Heat Fluxes (OAFlux) [Yu et al., 2007]; (f) Surface net shortwave and longwave fluxes from the NASA World Climate Research Programme/Global Energy L18810 1of5

Figure 1. (a) The leading mode of the EOF (EOF1) derived from the outgoing longwave radiation (OLR, unit: Wm 2 )of 22-summers (June August). The box indicates the SWNP region (15 25 N; 130 150 E). (b) The principal component of the leading EOF mode (PC1) of seven years that have a clear shift in July. The onset dates when PC1 = 0 are given in Figure 1b. and Water-Cycle Experiment (WCRP/GEWEX) Surface Radiation Budget (SRB) project. We calculated the surface net heat flux using the OAFlux and SRB data. [8] All the data have a temporal resolution of 1 day. The spatial resolution is: (a) 2.5 latitude-longitude for the NOAA OLR, the NCEP-R2 winds, and the ISCCP D1 clouds; (b) 0.25 latitude-longitude for the NOAA OI SST; (c) 1 latitude-longitude for the OAFlux and SRB data. [9] Due to a large variation of tropical convections, the date identified as monsoon onset will vary with methods used. Most previous studies identified the monsoon onset using information either from the satellite-measured brightness temperature [Ueda and Yasunari, 1996] or from OLR [Lu et al., 2007]. In this study, we used the empirical orthogonal function (EOF) analysis of OLR to define the SWNP monsoon onset. To remove synoptic signals, a 5-day running averaging was first applied to the OLR data of the 22 summers (June-August) in 1985 2006. Seasonal means were then subtracted from the smoothed time series to remove interannual variations. Finally, the EOFs and their principal components (PCs) were then calculated for the domain (0 40 N; 120 180 E), which is much larger than the SWNP. We found that the leading EOF mode (EOF1) had the largest amplitude in the SWNP (Figure 1a), and the PC corresponding to the leading EOF mode (PC1) had large fluctuations in July. The date in July when the PC1 has a significant jump across the line of PC1=0 can be interpreted as the date of monsoon onset in the SWNP. A similar method was also used to define the first transition of Asian summer monsoon [Hung and Hsu, 2008]. There are seven years (1985, 1986, 1989, 1993, 1997, 1999 and 2005) exhibiting a clear monsoon onset. Figure 1b shows the PC1 for the seven years, with PC1=0 corresponding to the onset date. The onset dates are also shown in Figure 1b. It ranges by 2 weeks, from 9 to 25 of July. [10] We also used another method to define the onset date. In this method, the 15-day mean OLR over the SWNP after the onset date is less than that before the onset date by 30 Wm 2. Fourteen years of the 22 years were found to have a monsoon onset. The seven years determined by using the EOF method are among those 14 years determined by using the domain-mean OLR method with the largest reduction in OLR after the onset. Analyses of the monsoon onset for the 14 years yielded results similar to those for the 7 years determined by using the EOF method. 3. Thermal and Dynamical Conditions Surrounding Monsoon Onset [11] We defined the rapid change of OLR in July over the SWNP region as monsoon onset. For this definition to be meaningful, the thermal and dynamical conditions over East Asia and the western North Pacific should undergo significant change after the onset. Figure 2 shows the weekly composites of the 850-hPa streamlines and the high-level clouds with a large optical thickness before and after the onset. Data from the seven years that were identified as Figure 2. Spatial distributions of weekly mean high-thick clouds (HTC, shaded, unit: %) and 850-hPa streamlines. (a) Two weeks prior to the onset, (b) the week prior to the onset, and (c) the week after the onset. Clouds were classified as HTC for cloud top higher than the 440-hPa level and the optical thickness greater than 3.6. The box indicates the SWNP region. 2of5

Figure 3. The 200-hPa streamlines and vorticity (shaded areas). The thick lines show the ridge of the vorticity. The box indicates the SWNP region. The unit of vorticity is 10 6 sec 1. having a clear monsoon onset were used for constructing the composites. Two weeks prior to the onset (Figure 2a), the subtropical high-pressure system dominates the climate of East Asia and western North Pacific. The ridge of the high-pressure system (thick line) extends southwestward to 20 N near Taiwan. At the northern flank of the subtropical high is a band of HTC stretching from southeastern China to Japan, and further to the south of Bering Sea. To the southwest of the subtropical high, there is also a band of HTC associated with the MT. The convection over the SWNP is weak with a small amount of HTC. In the midocean, there are nearly no HTC, except in the region around 25 N and the dateline. This patch of high clouds is associated with the mid-ocean trough and the UCL as elaborated, respectively, by Chen et al. [2001] and Sato et al. [2005]. In the week prior to the onset (Figure 2b), the HTC at the northern and southern flanks of the subtropical high, as well as the SWNP, do not have a significant change from a week before. However, the small patch of high clouds in the mid-ocean at dateline now expands westward to 160 E. Right after the monsoon onset (Figure 2c), circulation and convection change drastically. The midocean cloudy region continues to move westward, and the MT located at 10 N deepens and penetrates eastward to 150 E. Convection strengthens not only along the Intertropical Convergence Zone (ITCZ) south of 10 N but also expands northward to the entire SWNP. Corresponding to the deepening of the MT, the west section of the anticyclonic ridge shifts northward by 10 latitude to the south of Japan. The convection zone along the northern flank of the subtropical high weakens significantly, which signifies the end of the late-spring and early-summer rainy season in China, Korea, and Japan. [12] To illustrate the association between the convection and the upper tropospheric circulation in the mid-ocean, we show in Figure 3 the 200-hPa streamline and vorticity. The thick lines show the ridge of the vorticity, which coincide with the trough of the geopotential height (not shown). Corresponding to the westward moving high clouds (Figure 2), the high-level vorticity changes significantly around the moving cloudy region (denoted by arrows in Figures 3a and 3b). Patches of high clouds are located on the northwest side of the mid-ocean trough, which is consistent with the results shown by Sato et al. [2005]. After the onset, the stretch of the high-level trough west of 170 E bends significantly southward (Figure 3c) and the mid-latitude jet weakens (not shown), indicating a large change in circulation not only in the lower troposphere but also in the upper troposphere. [13] The results shown in Figure 2 seem to suggest that the rapid enhancement of convection in the SWNP is related to the westward expansion of high-level clouds from the mid-ocean. To support this view, we investigated the evolutions of circulation and convection prior to the monsoon onset. Figure 4 shows differences in 850-hPa circulation (streamlines) and HTC between consecutive 3-day intervals prior to the onset. Along the 20 N zone, the HTC increases from the 3-day period of ( 7 9) to the period ( 4 6) at 170 E (Figure 4a). The region of enhanced HTC moves westward to 155 E in the following 3-day period, which is associated with an increase in lowlevel cyclonic circulation (Figure 4b). After the onset, the region of enhanced HTC further moves westward to the SWNP with a strong enhancement in cyclonic circulation (Figure 4c). At this stage, the cloud amount in the convec- Figure 4. Changes of the atmospheric circulation (represented by streamlines) and the HTC (shaded, unit: %) between consecutive 3-day intervals (a and b) prior to and (c) during the monsoon onset. The letter C denotes the center of cyclonic circulation. 3of5

Figure 5. The SWNP (15 25 N; 130 150 E) domainmean 925-hPa air convergence (bars), HTC amount (dots), and SST (pluses) before and after monsoon onset. tive bands previously located at the northern flank (25 50 N) and southern flank (south of 10 N) of the highpressure ridge decreases greatly, signifying the onset of a new phase of the summer monsoon circulation. [14] Figure 5 summarizes the thermal and dynamical conditions surrounding the SWNP onset. Within two weeks prior to the onset, the atmosphere in the SWNP is clear and stable. The surface wind is only 5 m sec 1 (not shown), and the surface air is diverging (bars). The stable condition leads to a small amount of HTC of 5% (dots). Corresponding to the weak winds and clear skies is a weak evaporative cooling and a strong solar heating of the ocean. The consequence is a large total heating of the ocean of 110 Wm 2 during the two-week period prior to the onset (not shown). The SST increases steadily toward the onset date by 0.45 C in the two-week period (pluses). (With a nearly constant surface heating of 110 Wm 2, an increase of SST by 0.45 C in two weeks is equivalent to an ocean mixed layer of 55 m.) Regardless of the increasing SST, the surface wind remains divergent. It indicates that, prior to the onset, the SWNP is under the strong influence of the subtropical high-pressure system and that the increase in SST does not destabilize the atmosphere. [15] A few days prior to the onset, the surface wind and HTC begin to increase, and the surface heating begins to decrease. However, the SST continues to increase and reaches a maximum at the onset date. So, the change of SST lags the change of the surface heating by a few days. After the onset, the HTC, surface wind, and convergence continue to increase. Simultaneously, the surface heating and SST turn to decrease. Cloud systems move westward to the SWNP and trigger a new phase of summer monsoon. 4. Concluding Remarks [16] We investigated the thermal and dynamical conditions surrounding the monsoon onset in the SWNP using data from satellite observations and model reanalysis. By applying EOF analyses to the OLR with a length of 22 years from 1985 to 2006, there are seven years identified as having a strong monsoon onset in the SWNP in July. Prior to the onset, convection is generally weak and sky is clear in the subtropical mid-ocean, except in the region around the dateline. This cloudy region around the dateline propagates westward and is associated with the mid-ocean trough (high vorticity) in the upper troposphere. When this mid-ocean cloud system propagates to the SWNP, the region of highthick cloud with low OLR expands from the tropical western North Pacific (south of 10 N) northward and northeastward to the SWNP. The ridge of surface pressure shifts northward by 10 to the south of Japan, which coincides with the end of the late-spring and early-summer rainy season in the region extending from the southeastern China, to Korea, and further to Japan. It essentially enters a new phase of the summer monsoon in East Asia and the western North Pacific. [17] Prior to the monsoon onset, the atmosphere in the SWNP is stable. The westerly surface wind is weak and divergent, and the cloud amount is small. As a result, the heating of the ocean is strong, and the SST is high and increasing. The diverging wind and the high SST is an indication that the latter is due to a strong surface heating associated with a calm, stable atmosphere under the influence of the subtropical high-pressure system. After the monsoon onset, the surface wind increases and becomes convergent. Convections strengthen and clouds increase. It leads to rapid decreases of surface heating and SST. All these thermal and dynamical conditions surrounding the monsoon onset in the SWNP suggest that the monsoon onset is triggered by the westward propagation of the midocean disturbances in July when both the upper and lower tropospheric conditions favor the strengthening and penetration of the MT from the tropical western North Pacific to the SWNP. [18] Acknowledgments. This work was supported by National Research Council, Taiwan, under grant NSC 97-2111-M-002-009 and NSC 96-2119-M-002-004-MY3. The authors thank the anonymous reviewer for his/her constructive comments. The authors are also grateful for the following data used in this study: OLR from the National Oceanic and Atmospheric Administration (NOAA); winds from NOAA NCEP-R2 reanalysis; cloud data from the International Satellite Cloud Climatology Project (ISCCP); SST from the NOAA OI analysis; surface wind and turbulent heat fluxes from the Woods Hole Oceanographic Institution OAFlux project; surface radiation from NASA WCRP/GEWEX Surface Radiation Budget project. References Chen, T.-C., M.-C. Yen, G.-R. Liu, and S.-Y. Wang (2001), Summer upperlevel vortex over the North Pacific, Bull. Am. Meteorol. Soc., 82, 1991 2006, doi:10.1175/1520-0477(2001)082<1991:sulvot>2.3.co;2. 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. Clim., 14, 3834 3850, doi:10.1175/1520-0442(2001)014<3834:npotio>2.0.co;2. Hung, C.-W., and H.-H. Hsu (2008), The first transition of the Asian summer monsoon, interseasonal oscillation, and Taiwan Mei-yu, J. Clim., 21, 1552 1568, doi:10.1175/2007jcli1457.1. 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. Am. Meteorol. Soc., 83, 1631 1643, doi:10.1175/bams-83-11-1631(2002)083<1631:nar>2.3.co;2. Liebmann, B., and C. A. Smith (1996), Description of a complete (interpolated) outgoing longwave radiation dataset, Bull. Am. Meteorol. Soc., 77, 1275 1277. LinHo, and B. Wang (2002), The time-space structure of Asian summer monsoon A fast annual cycle view, J. Clim., 15, 2001 2019, doi:10.1175/1520-0442(2002)015<2001:ttssot>2.0.co;2. Lu, R., H. Ding, C.-S. Ryu, Z. Lin, and H. Dong (2007), Midlatitude westward propagating disturbances preceding intraseasonal oscillations of convection over the subtropical western North Pacific during summer, Geophys. Res. Lett., 34, L21702, doi:10.1029/2007gl031277. Reynolds, R. W., T. M. Smith, C. Liu, D. B. Chelton, K. S. Casey, and M. G. Schlax (2007), Daily high-resolution-blended analyses for sea surface temperature, J. Clim., 20, 5473 5496, doi:10.1175/2007jcli1824.1. 4of5

Rossow, W. B., and R. A. Schiffer (1999), Advances in understanding clouds from ISCCP, Bull. Am. Meteorol. Soc., 80, 2261 2288, doi:10.1175/1520-0477(1999)080<2261:aiucfi>2.0.co;2. Sakamoto, K., and M. Takahashi (2005), Cut off and weakening processes of an upper cold low, J. Meteorol. Soc. Jpn., 83, 817 834, doi:10.2151/ jmsj.83.817. Sato, N., K. Sakamoto, and M. Takahashi (2005), An air mass with high potential vorticity preceding the formation of the Marcus Convergence Zone, Geophys. Res. Lett., 32, L17801, doi:10.1029/2005gl023572. Ueda, H., and T. Yasunari (1996), Maturing process of the summer monsoon over the western North Pacific A coupled ocean/atmosphere system, J. Meteorol. Soc. Jpn., 74, 493 508. Ueda, H., M. Ohba, and S. P. Xie (2009), Important factors for the development of the Asian Northwest Pacific summer monsoon, J. Clim., 22, 649 668, doi:10.1175/2008jcli2341.1. Wu, R. (2002), Processes for the northeastward advance of the summer monsoon over the western North Pacific, J. Meteorol. Soc. Jpn., 80, 67 83, doi:10.2151/jmsj.80.67. Wu, R., and B. Wang (2001), Multi-stage onset of the summer monsoon over the western North Pacific, Clim. Dyn., 17, 277 289, doi:10.1007/ s003820000118. Yu, L., X. Jin, and R. A. Weller (2007), Annual, seasonal, and interannual variability of air-sea heat fluxes in the Indian Ocean, J. Clim., 20, 3190 3209, doi:10.1175/jcli4163.1. M.-D. Chou, W.-S. Kau, and C.-H. Wu, Department of Atmospheric Sciences, National Taiwan University, No 1. Sec. 4, Roosevelt Road, Taipei 106, Taiwan. (kwenshung@gmail.com) 5of5