Impacts of Steering Flows with Different Time Scales on the Track of Typhoon Sanba (2012)

+ Author Affiliations + Find other works by these authors
  • Corresponding author: Qiao LIU, liuq94@163.com
  • Funds:

    National Natural Science Foundation of China (41630423), US National Oceanic and Atmospheric Administration (NA18OAR4310298), US National Science Foundation (AGS-1643297), National Natural Science Foundation of China (41875069), and Priority Academic Program Development of Jiangsu Higher Education Institutions and China Scholarship Council (N201908320496)

  • doi: 10.1007/s13351-021-0125-z
  • Note: This paper has been peer-reviewed and is just accepted by J. Meteor. Res. Professional editing and proof reading are underway. Please use with caution.

PDF

  • Typhoon Sanba (2012), the strongest tropical cyclone (TC) of the year worldwide, moved northward almost along 130° longitude during its lifetime and passed through different background flows from low to high latitudes. The steering flows with different time scales for Sanba are retrieved by using the NCEP reanalysis data with the total wind field separated into: a mean state, an interannual component, an intraseasonal component, and a synoptic component. Our analysis indicates that the intraseasonal time-scale wave train (WT) with east–west oriented circulations made the largest contribution to the movement of Sanba. The effects of the environmental steering with different time scales on Sanba’s movement are investigated with numerical simulations using the Weather Research and Forecasting (WRF) model. In the control simulation, total fields from the NCEP reanalysis are used as initial and boundary conditions, and the northward motion of Sanba is well captured. In sensitivity experiments, each of the intraseasonal and interannual components is removed one at a time. The steering vectors associated with these time scales can explain their influences on the movement of Sanba in the experiments. Vorticity budget analyses indicate that the horizontal vorticity advection made the largest contribution to the movement of the storm.
  • 加载中
  • Fig. 1.  Composite of the vertically averaged (850–300 hPa) horizontal wind field (m s−1) from 0600 UTC 11 to 1800 UTC 17 September. Dots represent the locations of Sanba’s center. Purple, red, and yellow dots denote the super typhoon, typhoon, and tropical storm categories, respectively.

    Fig. 2.  Composites of the vertically averaged (850–300 hPa) (a, b) intraseasonal, (c) interannual, and (d) mean-state horizontal winds (m s−1) from 0600 UTC 11 to 1800 UTC 17 September. Shadings in (a) and (b) represent the intraseasonal vorticity fields (10−5 s−1). The composite of (a) is based on the center of Sanba, while the composites of (b), (c), and (d) are geographically fixed. Red dots represent the location of Sanba’s center. Green dots denote the averaged positions of Sanba’s center during its lifetime.

    Fig. 3.  Steering vectors (m s−1) with different time scales and the actual moving vector for Sanba from the best track data at 0600 UTC for each day from 12 to 17 September. “M”, “IA”, “IS”, “S”, “Sum”, and “TC” represent the steering vector of the mean-state flow (brown), interannual flow (blue), intraseasonal flow (red), synoptic flow (non-TC; green), sum of “M”, “IA”, and “IS” components (purple), and actual speed of Sanba (black), respectively.(请问各个小图片的横纵坐标名称分别是?).

    Fig. 4.  Evolution of the vertically averaged (850–300 hPa) intraseasonal horizontal wind (vector; m s−1) and vorticity (shaded; 10−5 s−1) fields at 1800 UTC for each day from 12 to 17 September. Green dots denote the locations of Sanba’s center.

    Fig. 5.  Joint Typhoon Warning Center (JTWC) best track (black dots) and simulated tracks of Sanba in the “CTL_1” (purple dots), “NO_IA_1” (green dots), and “NO_IS_1” (blue dots) experiments from 0600 UTC 12 to 1800 UTC 17 September. The red and yellow lines indicate the typhoon and tropical storm categories, respectively(请作者核查图题修改后是否ok).

    Fig. 6.  The simulated 10–90-day-scale horizontal wind (vector; s m−1) and vorticity (shaded; 10−5 s−1) fields vertically averaged from 850 to 300 hPa in the “CTL_1” experiment. Green dots indicate the locations of the simulated Sanba center. The red and blue curves represent the simulated 10–90-day cyclonic and anticyclonic circulations, respectively.

    Fig. 7.  As in Fig. 5 but for (a) the “CTL_2” (purple dots), “NO_IS_2” (blue dots), and “NO_IA_2” (green dots) experiments starting at 0600 UTC 15 September; and (b) the “CTL_3” (purple dots), “NO_IS_3” (blue dots), and “NO_IA_3” (green dots) experiments starting at 0600 UTC 16 September.

    Fig. 8.  The storm-centered composite wavenumber-one vorticity tendency (vertically averaged from 850 to 300 hPa; shaded; 10−9 s−2) fields for each term in the “CTL_1” experiment from 0600 UTC 12 to 0500 UTC 13 September: (a) term A; (b) sum of terms B1, B2, B3, C, and D; (c) term B1, the horizontal vorticity advection; (d) term B2, the vertical vorticity advection; (e) term B3, the planetary vorticity advection; (f) term C, the tilting; and (g) term D, the divergence. The black vector represents the magnitude of the maximum wavenumber-one vorticity tendency averaged over a 400-km radius and points to the region of the vorticity tendency maximum. (0, 0) is the center of the simulated TC.

    Fig. 9.  As in Fig. 8 but for the “NO_IS_1” experiment.

    Fig. 10.  As in Fig. 8 but for the “NO_IA_1” experiment.

    Table 1.  The description of numerical experiments

    ExperimentDescription
    CTL_1Control run with unfiltered initial and boundary fields from 0600 UTC 12 to 1800 UTC 17 September
    NO_IS_1As in CTL_1, but without the intraseasonal component
    NO_IA_1As in CTL_1, but without the interannual component
    CTL_2Control run with unfiltered initial and boundary fields from 0600 UTC 15 to 1800 UTC 17 September
    NO_IS_2As in CTL_2, but without the intraseasonal component
    NO_IA_2As in CTL_2, but without the interannual component
    CTL_3Control run with unfiltered initial and boundary fields from 0600 UTC 16 to 1800 UTC 17 September
    NO_IS_3As in CTL_3, but without the intraseasonal component
    NO_IA_3As in CTL_3, but without the interannual component
    Download: Download as CSV
  • [1]

    Anthes, R. A., 1982: Tropical Cyclones: Their Evolution, Structure and Effects. American Meteorological Society, Boston, 208 pp.
    [2]

    Aragon, L. G. B., and A. G. Pura, 2016: Analysis of the displacement error of the WRF–ARW model in predicting tropical cyclone tracks over the Philippines. Meteor. Appl., 23, 401–408. doi: 10.1002/met.1564.
    [3]

    Bi, M. Y., T. Li, M. Peng, et al., 2015: Interactions between Typhoon Megi (2010) and a low-frequency monsoon gyre. J. Atmos. Sci., 72, 2682–2702. doi: 10.1175/JAS-D-14-0269.1.
    [4]
    [5]

    Carr III, L. E., and R. L. Elsberry, 1995: Monsoonal interactions leading to sudden tropical cyclone track changes. Mon. Wea. Rev., 123, 265–290. doi: 10.1175/1520-0493(1995)123<0265:MILTST>2.0.CO;2.
    [6]

    Chan, J. C. L., 1985: Identification of the steering flow for tropi-cal cyclone motion from objectively analyzed wind fields. Mon. Wea. Rev., 113, 106–116. doi: 10.1175/1520-0493(1985)113<0106:IOTSFF>2.0.CO;2.
    [7]
    [8]

    Chan, J. C. L., F. M. F. Ko, and Y. M. Lei, 2002: Relationship between potential vorticity tendency and tropical cyclone motion. J. Atmos. Sci., 59, 1317–1336. doi: 10.1175/1520-0469(2002)059<1317:RBPVTA>2.0.CO;2.
    [9]

    Chen, T.-C., S.-Y. Wang, M.-C. Yen, et al., 2009: Impact of the intraseasonal variability of the western North Pacific large-scale circulation on tropical cyclone tracks. Wea. Forecasting, 24, 646–666. doi: 10.1175/2008WAF2222186.1.
    [10]

    Chen, X. M., Y. Q. Wang, K. Zhao, et al., 2017: A numerical study on rapid intensification of Typhoon Vicente (2012) in the South China Sea. Part I: Verification of simulation, storm-scale evolution, and environmental contribution. Mon. Wea. Rev., 145, 877–898. doi: 10.1175/MWR-D-16-0147.1.
    [11]

    Duchon, C. E., 1979: Lanczos filtering in one and two dimensions. J. Appl. Meteor., 18, 1016–1022. doi: 10.1175/1520-0450(1979)018<1016:LFIOAT>2.0.CO;2.
    [12]

    Ge, X. Y., Z. Y. Yan, M. Peng, et al., 2018: Sensitivity of tropical cyclone track to the vertical structure of a nearby monsoon gyre. J. Atmos. Sci., 75, 2017–2028. doi: 10.1175/JAS-D-17-0201.1.
    [13]

    George, J. E., and W. M. Gray, 1976: Tropical cyclone motion and surrounding parameter relationships. J. Appl. Meteor., 15, 1252–1264. doi: 10.1175/1520-0450(1976)015<1252:TCMASP>2.0.CO;2.
    [14]

    George, J. E., and W. M. Gray, 1977: Tropical cyclone recurvature and nonrecurvature as related to surrounding wind-height fields. J. Appl. Meteor., 16, 34–42. doi: 10.1175/1520-0450(1977)016<0034:TCRANA>2.0.CO;2.
    [15]

    Ho, C.-H., H.-S. Kim, J.-H. Jeong, et al., 2009: Influence of stratospheric quasi-biennial oscillation on tropical cyclone tracks in the western North Pacific. Geophys. Res. Lett., 36, L06702. doi: 10.1029/2009GL037163.
    [16]

    Holland, G. J., 1983: Tropical cyclone motion: Environmental interaction plus a beta effect. J. Atmos. Sci., 40, 328–342. doi: 10.1175/1520-0469(1983)040<0328:TCMEIP>2.0.CO;2.
    [17]

    Holland, G. J., 1984: Tropical cyclone motion. A comparison of theory and observation. J. Atmos. Sci., 41, 68–75. doi: 10.1175/1520-0469(1984)041<0068:TCMACO>2.0.CO;2.
    [18]

    Islam, T., P. K. Srivastava, M. A. Rico-Ramirez, et al., 2015: Tracking a tropical cyclone through WRF–ARW simulation and sensitivity of model physics. Nat. Hazards, 76, 1473–1495. doi: 10.1007/s11069-014-1494-8.
    [19]

    Jiang, X. N., D. E. Waliser, P. K. Xavier, et al., 2015: Vertical structure and physical processes of the Madden-Julian oscillation: Exploring key model physics in climate simulations. J. Geophys. Res. Atmos., 120, 4718–4748. doi: 10.1002/2014JD022375.
    [20]

    Kanase, R. D., and P. S. Salvekar, 2015: Effect of physical parameterization schemes on track and intensity of cyclone LAILA using WRF model. Asia-Pacific J. Atmos. Sci., 51, 205–227. doi: 10.1007/s13143-015-0071-8.
    [21]
    [22]

    Kurihara, Y., M. A. Bender, and R. J. Ross, 1993: An initialization scheme of hurricane models by vortex specification. Mon. Wea. Rev., 121, 2030–2045. doi: 10.1175/1520-0493(1993)121<2030:AISOHM>2.0.CO;2.
    [23]

    Lander, M., and G. J. Holland, 1993: On the interaction of tropical-cyclone-scale vortices. I: Observations. Quart. J. Roy. Meteor. Soc., 119, 1347–1361. doi: 10.1002/qj.49711951406.
    [24]

    Li, R. C. Y., and W. Zhou, 2013: Modulation of western North Pacific tropical cyclone activity by the ISO. Part II: Tracks and landfalls. J. Climate, 26, 2919–2930. doi: 10.1175/JCLI-D-12-00211.1.
    [25]

    Liu, Q., T. Li, and W. C. Zhou, 2018: Impact of 10–60-day low-frequency steering flows on straight northward-moving typhoon tracks over the western North Pacific. J. Meteor. Res., 32, 394–409. doi: 10.1007/s13351-018-7035-8.
    [26]

    Liu, Q., W. C. Zhou, M. Peng, et al., 2019: Factors controlling northward and north-eastward moving tropical cyclones near the coast of East Asia. Front. Earth Sci., 13, 778–790. doi: 10.1007/s11707-019-0797-1.
    [27]

    Rayhun, K. M. Z., D. A. Quadir, M. A. M. Chowdhury, et al., 2015: Simulation of structure, track and landfall of tropical cyclone Bijli using WRF-ARW model. J. Bangladesh Acad. Sci., 39, 157–167. doi: 10.3329/jbas.v39i2.25949.
    [28]

    Skamarock, W. C., J. B. Klemp, J. Dudhia, et al., 2008: A description of the Advanced Research WRF Version 3. (No. NCAR/TN-475+STR), University Corporation for Atmospheric Research, Boulder, Colorado, USA, 113 pp, doi: 10.5065/D68S4MVH.
    [29]

    Takahashi, C., and R. Shirooka, 2014: Storm track activity over the North Pacific associated with the Madden-Julian Oscillation under ENSO conditions during boreal winter. J. Geophys. Res. Atmos., 119, 10,663–10,683. doi: 10.1002/2014JD021973.
    [30]

    Wang, B., and J. C. L. Chan, 2002: How strong ENSO events affect tropical storm activity over the western North Pacific. J. Climate, 15, 1643–1658. doi: 10.1175/1520-0442(2002)015<1643:HSEEAT>2.0.CO;2.
    [31]
    [32]

    Wang, C. Z., C. X. Li, M. Mu, et al., 2013: Seasonal modulations of different impacts of two types of ENSO events on tropical cyclone activity in the western North Pacific. Climate Dyn., 40, 2887–2902. doi: 10.1007/s00382-012-1434-9.
    [33]

    Wu, C. C., and K. A. Emanuel, 1995: Potential vorticity diagnostics of hurricane movement. Part 1: A case study of Hurricane Bob (1991). Mon. Wea. Rev., 123, 69–92. doi: 10.1175/1520-0493(1995)123<0069:PVDOHM>2.0.CO;2.
    [34]

    Wu, C.-C., and Y. Kurihara, 1996: A numerical study of the feedback mechanisms of hurricane–environment interaction on hurricane movement from the potential vorticity perspective. J. Atmos. Sci., 53, 2264–2282. doi: 10.1175/1520-0469(1996)053<2264:ANSOTF>2.0.CO;2.
    [35]

    Wu, L. G., and B. Wang, 2000: A potential vorticity tendency diagnostic approach for tropical cyclone motion. Mon. Wea. Rev., 128, 1899–1911. doi: 10.1175/1520-0493(2000)128<1899:APVTDA>2.0.CO;2.
    [36]

    Wu, L. G., and X. Y. Chen, 2016: Revisiting the steering princi-pal of tropical cyclone motion in a numerical experiment. Atmos. Chem. Phys., 16, 14,925–14,936. doi: 10.5194/acp-16-14925-2016.
    [37]

    Yang, L., Y. Du, D. X. Wang, et al., 2015: Impact of intraseasonal oscillation on the tropical cyclone track in the South China Sea. Climate Dyn., 44, 1505–1519. doi: 10.1007/s00382-014-2180-y.
    [38]

    Yonekura, E., and T. M. Hall, 2014: ENSO effect on East Asian tropical cyclone landfall via changes in tracks and genesis in a statistical model. J. Appl. Meteor., 53, 406–420. doi: 10.1175/JAMC-D-12-0240.1.
    [39]

    Yoshida, R., Y. Kajikawa, and H. Ishikawa, 2014: Impact of boreal summer intraseasonal oscillation on environment of tropical cyclone genesis over the western North Pacific. Sci. Online Lett. Atmos., 10, 15–18. doi: 10.2151/sola.2014-004.
  • 加载中
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Impacts of Steering Flows with Different Time Scales on the Track of Typhoon Sanba (2012)

    Corresponding author: Qiao LIU, liuq94@163.com
  • 1. Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters (CIC-FEMD) / Key Laboratory of Meteorological Disaster, Ministry of Education (KLME) / Joint International Research Laboratory of Climate and Environmental Change (ILCEC), Nanjing University of Information Science and Technology, Nanjing 210044, China
  • 2. University of Colorado, Colorado Spring, Colorado 80918, USA
  • 3. International Pacific Research Center and Department of Atmospheric Sciences, School of Ocean and Earth Science and Technology, University of Hawaii at Manoa, Honolulu, Hawaii 96822, USA
Funds: National Natural Science Foundation of China (41630423), US National Oceanic and Atmospheric Administration (NA18OAR4310298), US National Science Foundation (AGS-1643297), National Natural Science Foundation of China (41875069), and Priority Academic Program Development of Jiangsu Higher Education Institutions and China Scholarship Council (N201908320496)

Abstract:  Typhoon Sanba (2012), the strongest tropical cyclone (TC) of the year worldwide, moved northward almost along 130° longitude during its lifetime and passed through different background flows from low to high latitudes. The steering flows with different time scales for Sanba are retrieved by using the NCEP reanalysis data with the total wind field separated into: a mean state, an interannual component, an intraseasonal component, and a synoptic component. Our analysis indicates that the intraseasonal time-scale wave train (WT) with east–west oriented circulations made the largest contribution to the movement of Sanba. The effects of the environmental steering with different time scales on Sanba’s movement are investigated with numerical simulations using the Weather Research and Forecasting (WRF) model. In the control simulation, total fields from the NCEP reanalysis are used as initial and boundary conditions, and the northward motion of Sanba is well captured. In sensitivity experiments, each of the intraseasonal and interannual components is removed one at a time. The steering vectors associated with these time scales can explain their influences on the movement of Sanba in the experiments. Vorticity budget analyses indicate that the horizontal vorticity advection made the largest contribution to the movement of the storm.

    • Tropical cyclones (TCs) that impact East Asia and seriously threaten coastal regions are mainly formed in the western North Pacific (WNP). To accurately predict their tracks, it is important to understand the mechanisms that control TC motion in the WNP. Kasahara (1957) indicated that TC motion is basically governed by the steering flow and discussed the effect of the Coriolis force on TC motion based on the barotropic model. Holland (1984) found that the advection of a TC by the environmental flow can be well represented by the steering vector, determined by the vertically averaged horizontal wind field between 850 and 300 hPa within a radius of about 500 km around the storm center. Considerable efforts have been made toward an understanding of the relationship between background flows and TC tracks (Lander and Holland, 1993; Carr III and Elsberry, 1995; Takahashi and Shirooka, 2014; Yang et al., 2015). It is generally agreed that environmental steering plays a dominant role in TC motion, especially under the barotropic framework. As strong vertical shear or latent heat release can cause a change in the atmospheric temperature structure, many studies explain TC motion based on the concept of potential vorticity tendency (Wu and Emanuel, 1995; Wu and Kurihara, 1996). A baroclinic TC is likely to move toward the region of the potential vorticity tendency maximum, which is mainly contributed by potential vorticity advection and diabatic heating (Wu and Wang, 2000; Chan et al., 2002; Wu and Chen, 2016).

      As part of the environment, the atmospheric intraseasonal flow has significant impacts on nearby TCs (Chen et al., 2009; Yoshida et al., 2014). Using a barotropic vorticity model, Carr III and Elsberry (1995) showed that a sudden northward movement occurs when a TC approaches and collocates with the intraseasonal monsoon gyre. Bi et al. (2015) found that the interaction between Typhoon Megi (2010) and the monsoon gyre led to Megi’s sudden turn northward. Yang et al. (2015) revealed that the majority of TCs in the South China Sea move westward under the influence of the environmental flow, while eastward-moving TCs are controlled largely by the intraseasonal flow. Li and Zhou (2013) suggested that westward TCs in the WNP are linked to the convective–active phases of the Madden–Julian oscillation and more northwestward TCs are found in the convective‒inactive phases. Liu et al. (2018) identified three types of intraseasonal flows that can influence northward-moving typhoons in the WNP: the monsoon gyre, the wave train (WT), and the mid-latitude trough. In the scenario of the WT pattern, a (an) cyclonic (anticyclonic) circulation is located to the west (east) of the northward typhoon center and the typhoon moves northwards under the steering of the intraseasonal southerly associated with the WT.

      The interannual variability of the atmospheric circulation also affects TC motion (Wang and Chan, 2002; Wang et al., 2013). Camargo et al. (2007) demonstrated the influence of the El Niño–Southern Oscillation (ENSO) on TC tracks over the WNP as detected in specific TC track clusters. Among these, recurving TCs form the largest cluster and occur more often during La Niña years. Yonekura and Hall (2014) examined the effect of ENSO on East Asian TC landfall based on a statistical TC track model and found that landfall rates increase in the coastal regions of Vietnam, China, the Korean Peninsula, and Japan during La Niña years. The quasi-biennial oscillation can also affect the motion of TCs. Ho et al. (2009) found that more TCs approach the East China Sea during the westerly phases of the quasi-biennial oscillation, while the number of TCs approaching the eastern offshore of Japan is larger during the easterly phases.

      In August most TCs near the east coast of China move northward, while more TCs move northeastward in September, which is attributed to the change in the climatological mean flow between 25°N and 30°N (Liu et al., 2019). While imbedded in the strong northeastward climatological flow near the east coast of China, Typhoon Sanba (2012) moved almost straight northward during its lifetime from 12 to 17 September. The objective of this study is to investigate in detail the reasons behind the nearly straight northward movement of Sanba. Our focus is on the relative contributions of the steering flows with different time scales to the northward movement of Sanba.

      The rest of the paper is organized as follows. The data and methodology are described in Section 2. The history of Sanba and the effects of different time-scale steering flows on the movement of Sanba are discussed in Section 3 using reanalysis data. In Section 4, numerical simulations for Sanba and sensitivity experiments to understand the impact of modes with different time scales are presented. The diagnosis of vorticity tendency is given in Section 5. A summary is given in Section 6.

    2.   Data and methodology
    • The primary dataset used in this study is the NCEP Final (FNL) Operational Global Analysis (available at https://rda.ucar.edu/datasets/ds083.2/), with 1° × 1° resolution at a 6-h interval. These data were used for analyzing Sanba’s movement and as initial and boundary conditions for numerical simulations. Data for the period 2000–2014 were used to construct the flows with different time scales. The TC best track data were obtained from the Joint Typhoon Warning Center (JTWC), including the genesis time, positions, and intensities at a 6-h temporal resolution.

    • To examine the relative influences of the background flows with different time scales on Sanba’s motion, the Lanczos filter (Duchon, 1979) with different frequencies is applied to the NCEP reanalysis data from 2000 to 2014. The two additional years of data beyond 2012 are included to obtain a more complete picture of the interannual variations. With this filtering, the background field is separated into a 10-day high-frequency-pass filtered component that includes TCs and other synoptic-scale flows, an intraseasonal (10–90-day band-pass filtered) component, and a low-frequency (90-day low-pass filtered) component. A 15-year monthly averaged low-frequency field is obtained to define the mean state in this study. Thus, the total field is decomposed into four components: a mean state, an interannual component obtained by subtracting the mean state from the 90-day low-pass filtered field, an intraseasonal component, and a synoptic-scale component. The data fields of the four components from 0600 UTC 11 to 1800 UTC 17 September 2012 are examined to analyze the movement of Sanba (2012).

      The model used for the numerical simulations is the Weather Research and Forecasting (WRF) model (Skamarock et al., 2008). The WRF model has been used extensively to study various weather events, such as heavy rainfall (Wang et al., 2017) and TC activities (Islam et al., 2015; Kanase and Salvekar, 2015; Rayhun et al., 2015; Aragon and Pura, 2016; Chen et al., 2017). The configuration of the system in our simulation has two meshes with horizontal resolutions of 30 km and 10 km, respectively, and the inner domain moves with the storm. The outer domain covers a region within (3°–58°N and 92°–162°E) in the first group of experiments and (7°–58°N and 101°–152°E) in the second and third groups to accommodate different initial positions of Sanba with different starting dates of the integration. The WRF-single moment 6-class (WSM6) microphysics scheme, Dudhia shortwave radiation parameterization, and the Rapid Radiative Transfer Model (RRTM) longwave radiation parameterization are used in both nested domains. The Kain–Fritsch convective scheme is used only in the outer domain.

    3.   Typhoon Sanba and environmental flows with different time scales
    • At 0600 UTC 11 September 2012, Sanba, located to the east of the Philippines (Fig. 1), was identified as a tropical storm by the JTWC. Sanba moved northwestward over the next day and was upgraded to the typhoon category (winds greater than 64 knots) at 1200 UTC 12 September. Sanba continued moving in a north by northwest direction, reaching its peak intensity as a super typhoon by 1800 UTC 13 September with a minimum central pressure of 907 hPa. Sanba then moved almost straight northward to reach 600 km from the coast of eastern China on 16 September with a central pressure of 930 hPa and weakened to a typhoon when it reached South Korea late on 17 September. After making landfall in South Korea, Sanba weakened quickly into a tropical depression and transitioned to an extratropical cyclone.

      Figure 1.  Composite of the vertically averaged (850–300 hPa) horizontal wind field (m s−1) from 0600 UTC 11 to 1800 UTC 17 September. Dots represent the locations of Sanba’s center. Purple, red, and yellow dots denote the super typhoon, typhoon, and tropical storm categories, respectively.

      In addition to the track of Sanba, Fig. 1 shows the composite background flow of Sanba during its lifetime from 0600 UTC 11 to 1800 UTC 17 September. The composite is geographically fixed. During this period, Sanba was in a low-latitude convergent zone of westerlies and easterlies to the east of the Philippines and a mid-latitude trough to the east of Taiwan with the WNP occupied by a subtropical high-pressure system. Sanba was steered by the northwestward flow initially and was then steered by the northward flow in low latitudes and the northeastward flow in midlatitudes. To investigate the relative roles of different time-scale flows in the movement of Sanba, the total wind field is separated into: a mean state, an intraseasonal component, an interannual component, and a synoptic-scale component which includes Sanba’s circulation. Figure 2 shows the composites of the wind fields associated with the intraseasonal scale, the interannual scale, and the mean state from 0600 UTC 11 to 1800 UTC 17 September 2012 during Sanba’s lifetime. All wind fields are vertically averaged from 850 to 300 hPa. As Sanba moved from low to high latitudes, it would be influenced differently by these components.

      Figure 2.  Composites of the vertically averaged (850–300 hPa) (a, b) intraseasonal, (c) interannual, and (d) mean-state horizontal winds (m s−1) from 0600 UTC 11 to 1800 UTC 17 September. Shadings in (a) and (b) represent the intraseasonal vorticity fields (10−5 s−1). The composite of (a) is based on the center of Sanba, while the composites of (b), (c), and (d) are geographically fixed. Red dots represent the location of Sanba’s center. Green dots denote the averaged positions of Sanba’s center during its lifetime.

      Starting from the bottom panels, the mean state for September (Fig. 2d) shows that a large anticyclonic circulation occupied the WNP and Sanba was located to the west of it. Strong westerlies occurred north of 30°N, which would contribute to the eastward movement of a TC near the east coast of China. This is in agreement with Liu et al. (2019). However, Sanba still moved northward. What causes the northward movement of Sanba and what are the relative contributions of other time-scale flows to the northward track? The composite of the interannual flow shows a weak trough along the track of the storm south of 30°N and a strong, large-scale anticyclone north of 30°N, covering Japan and Korea (Fig. 2c). Within the averaged intraseasonal flow (Fig. 2b), a large cyclonic circulation was located at East Asia. Sanba moved northward along the east periphery of the cyclonic circulation. An intraseasonal WT (Liu et al., 2018) can be identified with a cyclonic circulation to the west and an anticyclonic circulation to the east of Sanba. The WT pattern is clear in the storm-centered composite (Fig. 2a). In the geographically fixed composite, the anticyclonic circulation has two parts, one to the north near Japan and the other south of 20°N. This is partially due to the composite of different time frames, to be discussed in more detail later.

      Sanba formed at 10°N and moved northwestward to 14°N in the southeasterly mean flow (Fig. 2d). However, the southwesterlies associated with the intraseasonal cyclonic circulation over East Asia and the interannual cyclonic circulation east of the Philippines prefer a northeastward track (Figs. 2cd). Between 15° and 25°N, the easterly mean flow favored the westward movement for a TC (Fig. 2d). However, the intraseasonal westerly (Fig. 2b) offsets the easterly flow of the mean state, causing the northward movement of Sanba. North of 30°N, the flow of the mean state was mainly a strong westerly that would steer Sanba to the east or the east–northeast and affect Japan (Fig. 2d). However, both the intraseasonal (Fig. 2b) and interannual flows (Fig. 2c) north of 30°N were in the northwest direction associated with a large anticyclonic circulation. The southeasterly winds associated with the intraseasonal and interannual components contributed to the northward component of the environmental flows and reduced the eastward component coming from the mean state.

      As Sanba’s own circulation is included in the synoptic-scale wind field, we use the TC-removal method proposed by Kurihara et al. (1993) by separating the circulation of Sanba from the total wind field and further separate the synoptic-scale component into a TC component and a non-TC component. To understand the relative roles of different time-scale modes in Sanba’s northward movement more clearly, here we compute the background steering vectors associated with the synoptic (non-TC) component, intraseasonal component, interannual component, and the mean state at 0600 UTC for each day from 12 to 17 September within a 500-km radius. The vectors are vertically averaged from 850 to 300 hPa (Fig. 3). The concept of steering flows has long been used to understand the movement of TCs (George and Gray, 1976, 1977; Anthes,1982; Chan and Gray, 1982; Chan, 1985).

      Figure 3.  Steering vectors (m s−1) with different time scales and the actual moving vector for Sanba from the best track data at 0600 UTC for each day from 12 to 17 September. “M”, “IA”, “IS”, “S”, “Sum”, and “TC” represent the steering vector of the mean-state flow (brown), interannual flow (blue), intraseasonal flow (red), synoptic flow (non-TC; green), sum of “M”, “IA”, and “IS” components (purple), and actual speed of Sanba (black), respectively.(请问各个小图片的横纵坐标名称分别是?).

      On the first two days when Sanba was at low latitudes, the northward steering was largely contributed by the intraseasonal flow. The westward components of the mean flow and the synoptic flow were, however, countered by the eastward steering from the interannual and intraseasonal flows. As Sanba moved northward away from the trade wind region over the next two days (14–15 September), the northeastward steering from the intraseasonal flows increased somewhat and largely cancelled the westward steering from the mean state and the southward steering from the synoptic component, resulting in the mostly northward motion of Sanba (Figs. 3cd). As Sanba continually moved northward, it moved to the mid-latitude westerly region. The eastward steering of the mean state was very large and countered the intraseasonal and interannual components, leading to a mainly northward movement of Sanba (Figs. 3ef). During Sanba’s lifetime, the synoptic steering always pointed to the west or the south, indicating that the synoptic (non-TC) component made a relatively small contribution to the northward movement of Sanba. Sanba was displaced westward and northward relative to the sum of the steering vectors from the synoptic (non-TC), intraseasonal, and interannual components, and the mean state. Observational studies show that, in most cases, TCs in the Northern Hemisphere move to the left of the steering current (George and Gray, 1976; Chan and Gray, 1982; Holland, 1983). Overall, the intraseasonal component played the major role in controlling the movement of Sanba and this warrants more investigation.

      Evolutions of the intraseasonal flow and its vorticity field during the lifetime of Sanba are displayed in Fig. 4. The cyclonic and anticyclonic circulations associated with the intraseasonal WT are outlined by red curves and blue curves, respectively. A WT occurred to the east of the Philippines from 1800 UTC 12 to 1800 UTC 14 September with Sanba initially located to the south of the cyclonic center (Figs. 4ac). During this period, Sanba moved northward for approximately seven-degrees latitude, while the cyclonic circulation of the WT moved northward at a slower speed. The centers of Sanba and the WT cyclonic circulation got closer with time. Sanba was in the southern part of the WT cyclonic circulation, and was thus steered by the local westerly south of the cyclonic center. The westward flow of the mean state (Fig. 2d) and the eastward flow associated with the intraseasonal WT had opposite effects on Sanba’s motion during this period. At 1800 UTC 15 September, the two smaller cyclonic circulations west of Sanba merged into a large cyclonic circulation east of China. There were also two anticyclonic circulations and the one in the northern part around 40°N strengthened with time. The two merged into one by 1800 UTC 16 September (Fig. 4e). Overall, the northward flow associated with the west–east-oriented WT acted as ventilation and steered Sanba to move northward.

      Figure 4.  Evolution of the vertically averaged (850–300 hPa) intraseasonal horizontal wind (vector; m s−1) and vorticity (shaded; 10−5 s−1) fields at 1800 UTC for each day from 12 to 17 September. Green dots denote the locations of Sanba’s center.

      In summary, the northward steering of the intraseasonal flow made the largest contribution to the northward movement of Sanba. When Sanba was located south of 30°N, the flow of the mean state was easterly but was countered by the westerly component of the intraseasonal flow. Meanwhile, the interannual flow displayed a weak trough near Sanba with little net flow in either direction. When Sanba moved to north of 30°N, its surrounding flow of the mean state became most westerly as a part of the persistent anticyclonic system in the WNP. The southerly wind of the intraseasonal flow between the cyclonic circulation to the west of Sanba and the anticyclonic circulation to the east was enhanced and responsible for the northward movement of Sanba. At the same time, the interannual flow was northeasterly, canceling the westerly steering from the mean state, and Sanba moved northward.

    4.   Numerical simulations
    • We use the Advanced Research WRF (ARW-WRF) model and the NCEP FNL Operational Global Analysis data to conduct numerical experiments with different starting dates. In the control experiment, identified as the “CTL_1” experiment, the total fields from the NCEP reanalysis are used as the initial and boundary conditions, updated every 6 h, and the SST is updated every 24 h. The first set of simulations starts at 0600 UTC 12 September, 6 h before Sanba reached the typhoon category, and integrates for 132 h. To examine the impacts of the intraseasonal flow on Sanba’s motion, we conduct a sensitivity experiment in which we filter out the 10–90-day mode of all relevant variables including meridional winds, zonal winds, geopotential height, pressure, sea-level pressure, surface pressure, temperature, relative humidity, and sea surface temperature in the initial and boundary conditions, identified as the “NO_IS_1” experiment. To investigate the contribution of the interannual component, we remove the interannual mode from the total fields, identified as the “NO_IA_1” experiment. The “_1” part of the label denotes different sets of experiments with different starting times. All the experiments are listed in Table 1.

      ExperimentDescription
      CTL_1Control run with unfiltered initial and boundary fields from 0600 UTC 12 to 1800 UTC 17 September
      NO_IS_1As in CTL_1, but without the intraseasonal component
      NO_IA_1As in CTL_1, but without the interannual component
      CTL_2Control run with unfiltered initial and boundary fields from 0600 UTC 15 to 1800 UTC 17 September
      NO_IS_2As in CTL_2, but without the intraseasonal component
      NO_IA_2As in CTL_2, but without the interannual component
      CTL_3Control run with unfiltered initial and boundary fields from 0600 UTC 16 to 1800 UTC 17 September
      NO_IS_3As in CTL_3, but without the intraseasonal component
      NO_IA_3As in CTL_3, but without the interannual component

      Table 1.  The description of numerical experiments

      Figure 5 shows the best track and simulated tracks for Sanba in the first set of experiments starting at 0600 UTC 12 September. In the “CTL_1” experiment, the TC moves mainly northward, in fairly good agreement with the JTWC best track. The simulated TC track is right on the spot before 17 September and then deviates somewhat to the west from the best track afterwards. Considering the long forecast lead time of 138 h for a limited-area model, the “CTL_1” experiment is able to simulate reasonably well the main feature of the track. In the “NO_IS_1” and “NO_IA_1” simulations, the TC-vortex moves westward and northwestward, respectively, suggesting that both the intraseasonal and interannual components have large impacts on the movement of Sanba. Note that the mean-state background flow shown in Fig. 2f indicates an easterly south of 25°N. When either the northeastward steering of intraseasonal flow or the eastward steering of the interannual flow (Fig. 3a) is missing, Sanba moves westward following the flow of the mean state and does not move northward at the later stage. Compared with the “NO_IA_1” experiment, the simulated track in the “NO_IS_1” experiment deviates more from the northward track in the control experiment, suggesting that removing the intraseasonal mode has a greater impact on the northward motion, as discussed earlier.

      Figure 5.  Joint Typhoon Warning Center (JTWC) best track (black dots) and simulated tracks of Sanba in the “CTL_1” (purple dots), “NO_IA_1” (green dots), and “NO_IS_1” (blue dots) experiments from 0600 UTC 12 to 1800 UTC 17 September. The red and yellow lines indicate the typhoon and tropical storm categories, respectively(请作者核查图题修改后是否ok).

      To examine the simulated intraseasonal flows, we extract the 10–90-day component from the model output fields in the “CTL_1” simulation. The method of a 5-day running mean subtracted from a 45-day running mean is used. Since the model-simulated output fields cover only 132 h, the NCEP FNL daily analysis data 22 days before and 22 days after the integration period are merged with the model outputs. Figure 6 shows the evolution of the simulated intraseasonal flows and their vorticity fields. Compared with the observation (Fig. 4), major circulations associated with the 10–90-day-scale flow were captured in the control experiment. A cyclonic circulation resides to the northeast of the Philippines from 12 to 14 September. During that time, the simulated Sanba is located in the southeastern part of the intraseasonal cyclonic circulation, moving northward with it. The intraseasonal cyclonic circulation then merges with another large cyclonic circulation over China by 1800 UTC 15 September. The cyclonic and the anticyclonic circulations east of Japan constitute an intraseasonal WT and the ventilation flow associated with the WT steers Sanba to move northward over the following two days.

      Figure 6.  The simulated 10–90-day-scale horizontal wind (vector; s m−1) and vorticity (shaded; 10−5 s−1) fields vertically averaged from 850 to 300 hPa in the “CTL_1” experiment. Green dots indicate the locations of the simulated Sanba center. The red and blue curves represent the simulated 10–90-day cyclonic and anticyclonic circulations, respectively.

      As flows with different time scales change from lower latitudes to higher latitudes, their influence on Sanba may change as Sanba moves forward (Fig. 2). We conduct two additional sets of experiments with different initial times, designed as “CTL_1” and its associated sensitivity experiments (Table 1). The second set of simulations (_2) starts at 0600 UTC 15 September and integrates for only 60 h and the third set of simulations (_3) starts at 0600 UTC 16 September and integrates for only 36 h as Sanba is approaching land. Figures 7a and b show the best track and the simulated tracks in the second and third sets of simulations, respectively. In the “NO_IA_2” and “NO_IA_3” simulations, when the interannual flow is removed, the simulated Sanba still moves northward, indicating the small effect of the interannual flow on TC motion during this period. This is consistent with the steering vectors displayed in Figs. 3de, in which the steering vector of the interannual component at the time is very small. However, the TC moves northwestward without the intraseasonal flow in the “NO_IS_2” and “NO_IS_3” experiments, suggesting that the intraseasonal component is more important in controlling the movement of Sanba at this stage, as in the case when Sanba was located at lower latitudes.

      Figure 7.  As in Fig. 5 but for (a) the “CTL_2” (purple dots), “NO_IS_2” (blue dots), and “NO_IA_2” (green dots) experiments starting at 0600 UTC 15 September; and (b) the “CTL_3” (purple dots), “NO_IS_3” (blue dots), and “NO_IA_3” (green dots) experiments starting at 0600 UTC 16 September.

      To summarize, when the intraseasonal component is filtered out, the direction and moving speeds of Sanba change in all experiments starting at different times (latitudes). The interannual mode has a significant impact only in the first set of experiments when Sanba is located at low latitudes (Fig. 5). These experiments demonstrate the most important role of the intraseasonal mode in the northward movement of Sanba.

    5.   Diagnosis of vorticity tendency
    • To investigate how flows with different time scales affect Sanba’s movement, we conduct vorticity budget analysis for our numerical experiments, as a TC usually moves toward the region of the vorticity tendency maximum (Holland, 1983; Chan et al., 2002; Bi et al., 2015; Ge et al., 2018). The vorticity tendency equation can be written as follows:

      $$ \begin{split} & \underbrace {\frac{{\partial \zeta }}{{\partial t}}}_A = \underbrace { - u\frac{{\partial \zeta }}{{\partial x}} - v\frac{{\partial \zeta }}{{\partial y}}}_{{B_1}}\underbrace { - \omega \frac{{\partial \zeta }}{{\partial p}}}_{{B_2}}\underbrace { - v\frac{{\partial f}}{{\partial y}}}_{{B_3}} + \underbrace {\frac{{\partial \omega }}{{\partial y}}\frac{{\partial u}}{{\partial p}} - \frac{{\partial v}}{{\partial p}}\frac{{\partial \omega }}{{\partial x}}}_C\\ & \underbrace { - \left( {\frac{{\partial u}}{{\partial x}} + \frac{{\partial v}}{{\partial y}}} \right)\left( {\zeta + f} \right)}_D, \end{split} $$ (1)

      where t is time, u is the zonal wind, v is the meridional wind, ω is the vertical velocity in a pressure coordinate, f is the Coriolis parameter, and $\zeta $ is relative vorticity. The relative vorticity tendency (A) can be affected by the five terms on the right-hand-side, the vorticity advection terms (B1, B2, and B3), which are the horizontal vorticity advection term (B1), the vertical vorticity advection term (B2), and the planetary vorticity advection term (B3), the tilting term (C), and the divergence term (D), with the friction term neglected. The vertical average from 850 to 300 hPa is computed for each term, consistent with the construction of the steering flow. We focus mainly on the azimuthal wavenumber-one component of the vorticity tendency which makes the largest contribution to the direction of TC motion (Bi et al., 2015).

      The moving directions of the simulated TCs are different during the first 24 h in the control and the three sensitivity experiments in the first set of simulations starting at 0600 UTC 12 September (Fig. 5). Vorticity budget analyses are conducted to understand the influences of different dynamic processes. Figure 8 shows the averaged wavenumber-one component of each vorticity tendency term in Eq. (1) during the first 24 h in the control experiment (“CTL_1”) from 0600 UTC 12 to 0600 UTC 13 September. The vorticity tendency, the left-hand-side of Eq. (1), is the average of the 24 vorticity differences between each hourly interval. The orientation of the maximum vorticity tendency represented by the black arrow in Fig. 8a points to the north, in good agreement with the simulated direction of Sanba’s motion.

      Figure 8.  The storm-centered composite wavenumber-one vorticity tendency (vertically averaged from 850 to 300 hPa; shaded; 10−9 s−2) fields for each term in the “CTL_1” experiment from 0600 UTC 12 to 0500 UTC 13 September: (a) term A; (b) sum of terms B1, B2, B3, C, and D; (c) term B1, the horizontal vorticity advection; (d) term B2, the vertical vorticity advection; (e) term B3, the planetary vorticity advection; (f) term C, the tilting; and (g) term D, the divergence. The black vector represents the magnitude of the maximum wavenumber-one vorticity tendency averaged over a 400-km radius and points to the region of the vorticity tendency maximum. (0, 0) is the center of the simulated TC.

      The magnitude and the horizontal pattern of the sum of all terms on the right-hand side of Eq. (1), terms B1 to D (Fig. 8b), are in good agreement with the left-hand-side of Eq. (1), the vorticity tendency (Fig. 8a), indicating that the vorticity budget analysis is reasonable. The horizontal vorticity advection term makes the largest contribution to TC motion (Fig. 8c). Meanwhile, the vertical vorticity advection and planetary vorticity advection also have positive impacts on the northward movement of the TC, but they are relatively small (Figs. 8de). However, the tilting term and the divergence term make negative contributions to the northward track (Figs. 8fg). But the tilting term is relatively small.

      In the “NO_IS_1” experiment, the maximum vorticity tendency is toward the west (Fig. 9a), consistent with the TC movement (Fig. 5). The westward vorticity tendency is primarily contributed by the horizontal vorticity advection term (Fig. 9c), as in the control experiment. In the absence of the intraseasonal component, the horizontal advection term forces the vortex to move toward the west. Moreover, the divergence term also favors the southwestward movement (Fig. 9g). The sum of the five terms on the right-hand side (Fig. 9b) has similar pattern and amplitude to the left-hand side of the vorticity equation (Fig. 9a).

      Figure 9.  As in Fig. 8 but for the “NO_IS_1” experiment.

      In the “NO_IA_1” experiment, the maximum vorticity tendency points to the northwest (Fig. 10a), conforming to the direction of movement of the simulated TC in the experiment (Fig. 5). The analysis of the wavenumber-one vorticity tendency budget for this experiment shows a consistent dynamic as in other experiments where the horizontal vorticity advection term plays the major role in determining the region and magnitude of the maximum vorticity tendency. With the interannual mode filtered out, the horizontal vorticity advection term has changed and leads the TC-vortex to move northwestward. The divergence term tends to oppose the TC’s direction of movement. Moreover, the tilting term hinders the northward motion but is small in magnitude. The sum of five terms on the right-hand side (Fig. 10b) has a similar pattern and amplitude to the left-hand side, indicating that the vorticity budget analysis is reliable, as in other experiments.

      Figure 10.  As in Fig. 8 but for the “NO_IA_1” experiment.

      Diagnosis of the vorticity tendency has also been conducted for the second and third sets of experiments with different starting times. The results are similar to the diagnosis of the “_1” experiments, showing that the horizontal vorticity advection has the largest effect on the TC motion (figures omitted)

    6.   Summary
    • Typhoon Sanba, the strongest TC worldwide in 2012, moved mostly northward through different background flows at different latitudes during its lifetime from 0600 UTC 11 to 1800 UTC 17 September and made landfall in South Korea. In this study, influences of steering flows with different time scales on the movement of Sanba have been investigated using the NCEP reanalysis data and numerical experiments with the ARW-WRF model. For diagnostics, the total field was separated into four different time-scale components: a synoptic component, an intraseasonal mode, an interannual mode, and a mean state.

      The westward steering of the mean state and the synoptic (non-TC) component opposed the eastward steering of intraseasonal and interannual components over 12–13 September when Sanba was located at lower latitudes south of 15°N. The westward steering of the mean state increased somewhat and the synoptic steering turned to the south, weakening the northeastward intraseasonal steering over 14–15 September. When Sanba moved to the north of 25°N, the strong eastward steering of the flow of the mean state (associated with the subtropical high in the WNP) was balanced by the effect of the intraseasonal flow in the east–west direction and Sanba continued moving northward. The northward movement of Sanba north of 30°N was controlled mainly by the intraseasonal WT with a cyclonic circulation to the west and an anticyclonic circulation to the east. The ventilating southerly associated with the intraseasonal WT made a large contribution to the northward movement of Sanba. The northward movement of Sanba was contributed largely by the intraseasonal component during Sanba’s lifetime among the four time-scale components. As shown by previous studies (Jiang et al., 2015), only a quarter of 27 current state-of-the-art global circulation models have skills in simulating realistic eastward propagation of the Madden–Julian Oscillation. Mid-latitude intraseasonal oscillations are even more difficult to reproduce. As the intraseasonal steering flow largely determines the northward movement of Sanba even in the mid-latitude westerly region, it is difficult to predict the northward track accurately if a model is unable to predict the intraseasonal change in the circulation system in the Western Pacific. This points out the importance of accurately predicting the intraseasonal flow, in order to predict TC tracks correctly.

      Numerical experiments were designed to identify the effects of modes with different time scales on Sanba’s movement. The control experiments with the total fields from the NCEP reanalysis given in the initial and boundary conditions can simulate the track of Sanba well, starting at different initial times. In the “NO_IS” experiment and the “NO_IA” experiment, the intraseasonal component and the interannual component are removed from the initial and boundary conditions, respectively. Deviations of the TC movements in these sensitivity experiments from the control experiment follow what is expected from the steering vectors of these individual time-scale components. Overall, the steering vector analysis and the sensitivity experiments indicate that the intraseasonal component played the most important role in the northward movement of Sanba.

      Vorticity budget analyses were conducted to examine the dynamics affecting the movement of Sanba in the three sets of numerical experiments with different starting times. The simulated Sanba moves toward the region of the vorticity tendency maximum which is contributed mainly by the horizontal vorticity advection term. The changes in direction of the simulated TC tracks follow the changes in the horizontal vorticity advection in the sensitivity experiments.

      Acknowledgments. Discussions with Dr. Mingyu Bi are greatly appreciated. We thank the editor and anonymous reviewers for their constructive comments and suggestions.

Reference (39)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return