• Barnes, G. M., , and P. Fuentes, 2010: Eye excess energy and the rapid intensification of Hurricane Lili (2002). Mon. Wea. Rev., 138, 14461458.

    • Search Google Scholar
    • Export Citation
  • Bosart, L. F., , C. S. Velden, , W. E. Bracken, , J. Molinari, , and P. G. Black, 2000: Environmental influences on the rapid intensification of Hurricane Opal (1995) over the Gulf of Mexico. Mon. Wea. Rev., 128, 322352.

    • Search Google Scholar
    • Export Citation
  • Braun, S. A., 2002: A cloud-resolving simulation of Hurricane Bob (1991): Storm structure and eyewall buoyancy. Mon. Wea. Rev., 130, 15731592.

    • Search Google Scholar
    • Export Citation
  • Bryan, G. H., , and R. Rotunno, 2009: The influence of near-surface, high-entropy air in hurricane eyes on maximum hurricane intensity. J. Atmos. Sci., 66, 148158.

    • Search Google Scholar
    • Export Citation
  • Chen, H., , and D.-L. Zhang, 2013: On the rapid intensification of Hurricane Wilma (2005). Part II: Convective bursts and the upper-level warm core. J. Atmos. Sci., 70, 146162.

    • Search Google Scholar
    • Export Citation
  • Chou, K.-H., , C.-C. Wu, , Y. Wang, , and C.-H. Chih, 2011: Eyewall evolution of typhoons crossing the Philippines and Taiwan: An observational study. Terr. Atmos. Oceanic Sci., 22, 535548.

    • Search Google Scholar
    • Export Citation
  • Cione, J. J., , and E. W. Uhlhorn, 2003: Sea surface temperature variability in hurricanes: Implications with respect to intensity change. Mon. Wea. Rev., 131, 17831796.

    • Search Google Scholar
    • Export Citation
  • Dudhia, J., 1989: Numerical study of convection observed during the Winter Monsoon Experiment using a mesoscale two-dimensional model. J. Atmos. Sci., 46, 30773107.

    • Search Google Scholar
    • Export Citation
  • Ek, M. B., , K. E. Mitchell, , Y. Lin, , E. Rogers, , P. Grunmann, , V. Koren, , G. Gayno, , and J. D. Tarpley, 2003: Implementation of Noah land surface model advances in the National Centers for Environmental Prediction operational mesoscale Eta Model. J. Geophys. Res., 108, 8851, doi:10.1029/2002JD003296.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., 1987: An air–sea interaction model of intraseasonal oscillations in the tropics. J. Atmos. Sci., 44, 23242340.

  • Emanuel, K. A., 1988: Observational evidence of slantwise convective adjustment. Mon. Wea. Rev., 116, 18051816.

  • Frisius, T., , and D. Schönemann, 2012: An extended model for the potential intensity of tropical cyclones. J. Atmos. Sci., 69, 641661.

    • Search Google Scholar
    • Export Citation
  • Fudeyasu, H., , Y. Wang, , M. Satoh, , T. Nasuno, , H. Miura, , and W. Yanase, 2010: Multiscale interactions in the life cycle of a tropical cyclone simulated in a global cloud-system-resolving model. Part I: Large-scale and storm-scale evolutions. Mon. Wea. Rev., 138, 42854304.

    • Search Google Scholar
    • Export Citation
  • Gray, S. L., , and A. J. Thorpe, 2001: Parcel theory in three dimensions and the calculation of SCAPE. Mon. Wea. Rev., 129, 16561672.

  • Gray, W. M., 1968: Global view of the origin of tropical disturbances and storms. Mon. Wea. Rev., 96, 669700.

  • Hack, J. J., , and W. H. Schubert, 1986: Nonlinear response of atmospheric vortices to heating by organized cumulus convection. J. Atmos. Sci., 43, 15591573.

    • Search Google Scholar
    • Export Citation
  • Holland, G. J., , and R. T. Merrill, 1984: On the dynamics of tropical cyclone structural changes. Quart. J. Roy. Meteor. Soc., 110, 723745.

    • Search Google Scholar
    • Export Citation
  • Holliday, C. R., , and A. H. Thompson, 1979: Climatological characteristics of rapidly intensifying typhoons. Mon. Wea. Rev., 107, 10221034.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., , and J.-O. Lim, 2006: The WRF single-moment 6-class microphysics scheme. J. Korean Meteor. Soc., 42, 129151.

  • Kain, J. S., 2004: The Kain–Fritsch convective parameterization: An update. J. Appl. Meteor., 43, 170181.

  • Kaplan, J., , and M. DeMaria, 2003: Large-scale characteristics of rapidly intensifying tropical cyclones in the North Atlantic basin. Wea. Forecasting, 18, 10931108.

    • Search Google Scholar
    • Export Citation
  • Kelley, O. A., , J. Stout, , and J. B. Halverson, 2005: Hurricane intensification detected by continuously monitoring tall precipitation in the eyewall. Geophys. Res. Lett., 32, L20819, doi:10.1029/2005GL023583.

    • Search Google Scholar
    • Export Citation
  • Kepert, J. D., , and Y. Wang, 2001: The dynamics of boundary layer jets within the tropical cyclone core. Part II: Nonlinear enhancement. J. Atmos. Sci., 58, 24852501.

    • Search Google Scholar
    • Export Citation
  • Kossin, J. P., , and M. Sitkowski, 2009: An objective model for identifying secondary eyewall formation in hurricanes. Mon. Wea. Rev., 137, 876892.

    • Search Google Scholar
    • Export Citation
  • Kurihara, Y., , M. A. Bender, , and R. J. Ross, 1993: An initialization scheme of hurricane models by vortex specification. Mon. Wea. Rev., 121, 20302045.

    • Search Google Scholar
    • Export Citation
  • Li, Q.-Q., , and Y. Wang, 2012: A comparison of inner and outer spiral rainbands in a numerically simulated tropical cyclone. Mon. Wea. Rev., 140, 27822805.

    • Search Google Scholar
    • Export Citation
  • Lin, I.-I., , C.-C. Wu, , I. F. Pun, , and D. S. Ko, 2008: Upper-ocean thermal structure and the western North Pacific category 5 typhoons. Part I: Ocean features and the category 5 typhoons’ intensification. Mon. Wea. Rev., 136, 32883306.

    • Search Google Scholar
    • Export Citation
  • McFarquhar, G. M., , B. F. Jewett, , M. S. Gilmore, , S. W. Nesbitt, , and T.-L. Hsieh, 2012: Vertical velocity and microphysical distributions related to rapid intensification in a simulation of Hurricane Dennis (2005). J. Atmos. Sci., 69, 35153534.

    • Search Google Scholar
    • Export Citation
  • Miyamoto, Y., , and T. Takemi, 2013: A transition mechanism for the spontaneous axisymmetric intensification of tropical cyclones. J. Atmos. Sci., 70, 112129.

    • Search Google Scholar
    • Export Citation
  • Mlawer, E. J., , S. J. Taubman, , P. D. Brown, , M. J. Iacono, , and S. A. Clough, 1997: Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102D, 16 66316 682.

    • Search Google Scholar
    • Export Citation
  • Molinari, J., , S. Skubis, , and D. Vollaro, 1995: External influences on hurricane intensity. Part III: Potential vorticity structure. J. Atmos. Sci., 52, 35933606.

    • Search Google Scholar
    • Export Citation
  • Montgomery, M. T., , M. E. Nicholls, , T. A. Cram, , and A. B. Saunders, 2006: A vortical hot tower route to tropical cyclogenesis. J. Atmos. Sci., 63, 355386.

    • Search Google Scholar
    • Export Citation
  • Moon, I.-J., , I. Ginis, , T. Hara, , and B. Thomas, 2007: A physics-based parameterization of air–sea momentum flux at high wind speeds and its impact on hurricane intensity predictions. Mon. Wea. Rev., 135, 28692878.

    • Search Google Scholar
    • Export Citation
  • Nakanishi, M., , and H. Niino, 2004: An improved Mellor–Yamada level-3 model with condensation physics: Its design and verification. Bound.-Layer Meteor., 112, 131.

    • Search Google Scholar
    • Export Citation
  • Nguyen, H. V., , and Y.-L. Chen, 2011: High-resolution initialization and simulations of Typhoon Morakot (2009). Mon. Wea. Rev., 139, 14631491.

    • Search Google Scholar
    • Export Citation
  • Ooyama, K. V., 1982: Conceptual evolution of the theory and modeling of the tropical cyclone. J. Meteor. Soc. Japan, 60, 369379.

  • Park, M.-S., , R. L. Elsberry, , and P. A. Harr, 2013: Vertical wind shear and ocean heat contents as environmental modulators of western North Pacific tropical cyclone intensification and decay. Trop. Cyclone Res. Rev., 1, 448457.

    • Search Google Scholar
    • Export Citation
  • Persing, J., , and M. T. Montgomery, 2003: Hurricane superintensity. J. Atmos. Sci., 60, 23492371.

  • Reasor, P. D., , M. D. Eastin, , and J. F. Gamache, 2009: Rapidly intensifying Hurricane Guillermo (1997). Part I: Low-wavenumber structure and evolution. Mon. Wea. Rev., 137, 603631.

    • Search Google Scholar
    • Export Citation
  • Rogers, R., 2010: Convective-scale structure and evolution during a high-resolution simulation of tropical cyclone rapid intensification. J. Atmos. Sci., 67, 4470.

    • Search Google Scholar
    • Export Citation
  • Rozoff, C. M., , W. H. Schubert, , D. Brian, , and J. Kossin, 2006: Rapid filamentation zones in intense tropical cyclones. J. Atmos. Sci., 63, 325340.

    • Search Google Scholar
    • Export Citation
  • Schubert, W. H., , and J. J. Hack, 1982: Inertial stability and tropical cyclone development. J. Atmos. Sci., 39, 16871697.

  • Shapiro, L. J., , and H. E. Willoughby, 1982: The response of balanced hurricanes to local sources of heat and momentum. J. Atmos. Sci., 39, 378394.

    • Search Google Scholar
    • Export Citation
  • Shutts, G. J., 1990: SCAPE charts from numerical weather prediction model fields. Mon. Wea. Rev., 118, 27462751.

  • Skamarock, W. C., , J. B. Klemp, , J. Dudhia, , D. O. Gill, , D. M. Barker, , W. Wang, , and J. G. Powers, 2008: A description of the Advanced Research WRF version 3. NCAR Tech Note-475+STR, 113 pp.

  • Smith, R. K., , M. T. Montgomery, , and V. S. Nguyen, 2009: Tropical cyclone spin-up revisited. Quart. J. Roy. Meteor. Soc., 135, 13211335.

    • Search Google Scholar
    • Export Citation
  • Steiner, M., , R. A. Houze Jr., , and S. E. Yuter, 1995: Climatological characterization of three-dimensional storm structure from operational radar and rain gauge data. J. Appl. Meteor., 34, 19782007.

    • Search Google Scholar
    • Export Citation
  • Vigh, J. L., , and W. H. Schubert, 2009: Rapid development of the tropical cyclone warm core. J. Atmos. Sci., 66, 33353350.

  • Wang, H., , Y. Wang, , and H.-M. Xu, 2013: Improving simulation of a tropical cyclone using dynamical initialization and large-scale spectral nudging: A case study of Typhoon Megi (2010). Acta Meteor. Sin., 27, 455–475.

    • Search Google Scholar
    • Export Citation
  • Wang, Y., 2008a: Rapid filamentation zone in a numerically simulated tropical cyclone. J. Atmos. Sci., 65, 11581181.

  • Wang, Y., 2008b: Structure and formation of an annular hurricane simulated in a fully-compressible, nonhydrostatic model—TCM4. J. Atmos. Sci., 65, 15051527.

    • Search Google Scholar
    • Export Citation
  • Wang, Y., 2009: How do outer spiral rainbands affect tropical cyclone structure and intensity? J. Atmos. Sci., 66, 12501273.

  • Wang, Y., , and J. Xu, 2010: Energy production, frictional dissipation, and maximum intensity of a numerically simulated tropical cyclone. J. Atmos. Sci., 67, 97116.

    • Search Google Scholar
    • Export Citation
  • Wang, Y., , and H. Wang, 2013: The inner-core size increase of Typhoon Megi (2010) during its rapid intensification phase. Trop. Cyclone Res. Rev., 2, 6580.

    • Search Google Scholar
    • Export Citation
  • Wu, C.-C., , K.-H. Chou, , H.-J. Cheng, , and Y. Wang, 2003: Eyewall contraction, breakdown and reformation in a landfalling typhoon. Geophys. Res. Lett., 30, 1887, doi:10.1029/2003GL017653.

    • Search Google Scholar
    • Export Citation
  • Wu, C.-C., , H.-J. Cheng, , Y. Wang, , and K.-H. Chou, 2009: A numerical investigation of the eyewall evolution of a landfalling typhoon. Mon. Wea. Rev., 137, 2140.

    • Search Google Scholar
    • Export Citation
  • Xu, J., , and Y. Wang, 2010a: Sensitivity of the simulated tropical cyclone inner-core size to the initial vortex size. Mon. Wea. Rev., 138, 41354157.

    • Search Google Scholar
    • Export Citation
  • Xu, J., , and Y. Wang, 2010b: Sensitivity of tropical cyclone inner-core size and intensity to the radial distribution of surface entropy flux. J. Atmos. Sci., 67, 18311852.

    • Search Google Scholar
    • Export Citation
  • Yuter, S. E., , and R. A. Houze Jr., 1995: Three-dimensional kinematic and microphysical evolution of Florida cumulonimbus. Part II: Frequency distributions of vertical velocity, reflectivity, and differential reflectivity. Mon. Wea. Rev., 123, 19411963.

    • Search Google Scholar
    • Export Citation
  • Zhang, D.-L., , and H. Chen, 2012: Importance of the upper-level warm core in the rapid intensification of a tropical cyclone. Geophys. Res. Lett., 39, L02806, doi:10.1029/2011GL050578.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    (a) Track of Typhoon Megi (2010) with every 6-h position indicated by solid circles with different colors for different categories in storm intensity, (b) the storm’s central sea level pressure (hPa), and (c) the maximum sustained 10-m wind speed (m s−1) from 0000 UTC 12 Oct to 0600 UTC 24 Oct 2010 based on the JTWC best track data. The modeled track, central sea level pressure, and maximum 10-m wind speed from 0000 UTC 15 Oct to 0000 UTC 22 Oct are also shown in red.

  • View in gallery

    (a) Satellite altimetry SST (°C) and (b) the upper OHC (kJ cm−2) on 17 Oct 2010 (shadings and white contours), overlapped with the track of Typhoon Megi (2010) from JTWC, with the storm track overlaid with the colored circles, indicating the intensity of the storm according to the Saffir–Simpson scale, which were produced by the Remote Sensing Laboratory at National Taiwan University and can be accessed online (http://data.eol.ucar.edu/codiac/dss/id=209.027).

  • View in gallery

    The satellite images at given times showing (top) the structural change before Typhoon Megi (2010) made landfall over Luzon Island and (bottom) the remarkable structural changes when the typhoon crossed Luzon Island and after it entered the SCS.

  • View in gallery

    Model domains with triply nested, movable meshes used for the Typhoon Megi (2010) simulation in this study.

  • View in gallery

    (a)–(c) (top) The model-simulated TBB (°C) and (bottom) the satellite infrared high-resolution (1 km) TBB at times given at the top of each panel for Typhoon Megi.

  • View in gallery

    Time evolution of vertical wind shear over the storm, shown are large-scale shear averaged between radii of 502 and 900 km (dotted) and shear between radii of 396 and 600 km (dashed) from domain 2, as well as shear averaged within 250-km radius in the nested domain 3 (solid).

  • View in gallery

    Wind vectors and geopotential height field (contours, gpm) from the NCEP GFS FNL at (a),(b) 200 and (c),(d) 850 hPa at 0000 UTC on (a),(c) 15 and (b),(d) 17 Oct 2010.

  • View in gallery

    The radius–time cross section of the azimuthal mean (a) tangential wind (m s−1) at 1-km height (contours) normalized by the maximum value at a given time (shading), (b) vertical velocity (cm s−1) at 3-km height, (c) normalized inertial stability (shading) and radial wind (m s−1, contours) at the lowest model level (about 28 m above the surface), and (d) surface rain rate (mm h−1).

  • View in gallery

    Radius–time cross section of the axisymmetricity parameter averaged in the layers between (a) 2 and 6, (b) 7 and 11, and (c) 12 and 16 km, respectively.

  • View in gallery

    CFAD (%) of simulated vertical velocity within 80-km radius in 24-h periods on (a) 15, (b) 16, and (c) 17 Oct.

  • View in gallery

    Cumulative contour frequency of distributions of vertical velocity within 80-km radius as a function of time for the simulated Typhoon Megi at (a) 2-, (b) 8-, and (c) 13-km heights.

  • View in gallery

    Time series of the convective (solid) and stratiform (dashed) percentages (a) within 80-km radius and (b) between 80- and 160-km radii.

  • View in gallery

    Number of grid points with convective bursts within 80-km radius: (a) convective bursts where the vertical velocity is greater than 7.5 m s−1 at 12-km height and (b) the maximum vertical velocity in the column between 2- and 12-km heights is greater than 7.5 m s−1.

  • View in gallery

    (a) Time–radius cross section of the azimuthal mean CAPE (J kg−1, shading), equivalent potential temperature (green contours with interval of 5 K), and RMW (dashed white). (b) The radial distribution of the number of grid points with vertical velocity greater than 7.5 m s−1 at 11-km altitude between 1200 UTC 15 Oct and 1200 UTC 16 Oct 2010.

  • View in gallery

    The radius–height cross section of the azimuthal mean vertical velocity (m s−1, shading) and radial wind (m s−1, contours) averaged on (a) 15 and (b) 16 Oct of the simulated Megi based on hourly model outputs.

  • View in gallery

    As in Fig. 14a, but for SCAPE calculated with the ascending air parcel arising from the lowest model level and following the azimuthal mean AAM surface and vertical velocities at 1-km altitude, with a contour interval of 0.2 m s−1 starting from 0.1 m s−1.

  • View in gallery

    Time evolution of the (a) temperature anomalies (K, shading) and equivalent potential temperature (K, contours) averaged within a 20-km radius and (b) inertial stability parameter in the boundary layer (s−2, red) and CAPE (J kg−1, blue) averaged within a 60-km radius of the simulated Megi.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 197 198 67
PDF Downloads 148 148 54

A Numerical Study of Typhoon Megi (2010). Part I: Rapid Intensification

View More View Less
  • 1 Key Laboratory of Meteorological Disaster of Ministry of Education, Nanjing University of Information Science and Technology, Nanjing, Jiangsu, China, and International Pacific Research Center and Department of Meteorology, School of Ocean and Earth Science and Technology, University of Hawai‘i at Mānoa, Honolulu, Hawaii
  • 2 International Pacific Research Center and Department of Meteorology, School of Ocean and Earth Science and Technology, University of Hawai‘i at Mānoa, Honolulu, Hawaii, and Key Laboratory of Meteorological Disaster of Ministry of Education, Nanjing University of Information Science and Technology, Nanjing, Jiangsu, China
© Get Permissions
Full access

Abstract

Typhoon Megi (15W) was the most powerful and longest-lived tropical cyclone (TC) over the western North Pacific during 2010. While it shared many common features of TCs that crossed Luzon Island in the northern Philippines, Megi experienced unique intensity and structural changes, which were reproduced reasonably well in a simulation using the Advanced Research Weather Research and Forecasting Model (ARW-WRF) with both dynamical initialization and large-scale spectral nudging. In this paper processes responsible for the rapid intensification (RI) of the modeled Megi before it made landfall over Luzon Island were analyzed. The results show that Megi experienced RI over the warm ocean with high ocean heat content and decreasing environmental vertical shear. The onset of RI was triggered by convective bursts (CBs), which penetrate into the upper troposphere, leading to the upper-tropospheric warming and the formation of the upper-level warm core. In turn, CBs with their roots inside of the eyewall in the boundary layer were buoyantly triggered/supported by slantwise convective available potential energy (SCAPE) accumulated in the eye region. During RI, convective area coverage in the inner-core region was increasing while the updraft velocity in the upper troposphere and the number of CBs were both decreasing. Different from the majority of TCs that experience RI with a significant eyewall contraction, the simulated Megi, as the observed, rapidly intensified without an eyewall contraction. This is attributed to diabatic heating in active spiral rainbands, a process previously proposed to explain the inner-core size increase, enhanced by the interaction of the typhoon vortex with a low-level synoptic depression in which Megi was embedded.

Corresponding author address: Dr. Yuqing Wang, IPRC/SOEST, Rm. 409G, POST Bldg., University of Hawai‘i at Mānoa, 1680 East–West Rd., Honolulu, HI 96822. E-mail: yuqing@hawaii.edu

Abstract

Typhoon Megi (15W) was the most powerful and longest-lived tropical cyclone (TC) over the western North Pacific during 2010. While it shared many common features of TCs that crossed Luzon Island in the northern Philippines, Megi experienced unique intensity and structural changes, which were reproduced reasonably well in a simulation using the Advanced Research Weather Research and Forecasting Model (ARW-WRF) with both dynamical initialization and large-scale spectral nudging. In this paper processes responsible for the rapid intensification (RI) of the modeled Megi before it made landfall over Luzon Island were analyzed. The results show that Megi experienced RI over the warm ocean with high ocean heat content and decreasing environmental vertical shear. The onset of RI was triggered by convective bursts (CBs), which penetrate into the upper troposphere, leading to the upper-tropospheric warming and the formation of the upper-level warm core. In turn, CBs with their roots inside of the eyewall in the boundary layer were buoyantly triggered/supported by slantwise convective available potential energy (SCAPE) accumulated in the eye region. During RI, convective area coverage in the inner-core region was increasing while the updraft velocity in the upper troposphere and the number of CBs were both decreasing. Different from the majority of TCs that experience RI with a significant eyewall contraction, the simulated Megi, as the observed, rapidly intensified without an eyewall contraction. This is attributed to diabatic heating in active spiral rainbands, a process previously proposed to explain the inner-core size increase, enhanced by the interaction of the typhoon vortex with a low-level synoptic depression in which Megi was embedded.

Corresponding author address: Dr. Yuqing Wang, IPRC/SOEST, Rm. 409G, POST Bldg., University of Hawai‘i at Mānoa, 1680 East–West Rd., Honolulu, HI 96822. E-mail: yuqing@hawaii.edu

1. Introduction

Megi (15W) was first identified as a tropical disturbance over the western North Pacific (WNP) by the Joint Typhoon Warning Center (JTWC) on 12 October 2010. The Japan Meteorological Agency (JMA) and JTWC began to monitor the low pressure circulation as a tropical depression (TD). The TD further intensified into a tropical storm (TS), named Megi by JMA at 1200 UTC on 12 October. Later on 14 October, the eye of the storm could be clearly seen from satellite image and JMA thus upgraded Megi to a severe tropical storm and JTWC upgraded it to a category-1 typhoon. On 15 October, JMA upgraded Megi to a typhoon.

As shown in Fig. 1, Megi initially moved northwestward and then turned west-southwestward. It experienced two periods of intensification before it made landfall at Luzon Island in the northern Philippines. The first intensification occurred from 1200 UTC 12 October to 0000 UTC 15 October during which the maximum 10-m wind speed increased by 35 m s−1 and the central sea level pressure (SLP) dropped by 45 hPa. The second intensification occurred from 0000 UTC 16 October to 1200 UTC 17 October. During this 36-h period the maximum 10-m wind speed also increased by 35 m s−1 while the central SLP dropped by 52 hPa. In the end, Megi attained its peak intensity with a central SLP of 905 hPa and a maximum 10-m wind speed of 80 m s−1, the most powerful supertyphoon over the WNP and South China Sea (SCS) in 2010. Based on the definition of rapid intensification (RI) by Holliday and Thompson (1979) for WNP tropical cyclones (TCs) and Kaplan and DeMaria (2003) for Atlantic TCs,1 the first intensification was not rapid. However, the second intensification can be classified as an RI case according to the definition proposed by Kaplan and DeMaria (2003). In this study, we will focus on the second intensification period, namely the RI phase of Typhoon Megi. During the RI period, Typhoon Megi moved west-southwestward east of the Philippines over the WNP with high sea surface temperature (SST) and high upper-ocean heat content (OHC) as shown in Fig. 2, both of which are favorable ocean conditions for RI of a TC (Lin et al. 2008).

Fig. 1.
Fig. 1.

(a) Track of Typhoon Megi (2010) with every 6-h position indicated by solid circles with different colors for different categories in storm intensity, (b) the storm’s central sea level pressure (hPa), and (c) the maximum sustained 10-m wind speed (m s−1) from 0000 UTC 12 Oct to 0600 UTC 24 Oct 2010 based on the JTWC best track data. The modeled track, central sea level pressure, and maximum 10-m wind speed from 0000 UTC 15 Oct to 0000 UTC 22 Oct are also shown in red.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

Fig. 2.
Fig. 2.

(a) Satellite altimetry SST (°C) and (b) the upper OHC (kJ cm−2) on 17 Oct 2010 (shadings and white contours), overlapped with the track of Typhoon Megi (2010) from JTWC, with the storm track overlaid with the colored circles, indicating the intensity of the storm according to the Saffir–Simpson scale, which were produced by the Remote Sensing Laboratory at National Taiwan University and can be accessed online (http://data.eol.ucar.edu/codiac/dss/id=209.027).

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

Megi made landfall over Luzon Island at around 0330 UTC on 18 October. It weakened to a category-2 typhoon immediately after its landfall. After crossing Luzon Island, Megi entered the SCS and turned northwestward and then suddenly north-northeastward on 20 October. During its northwest-to-north turning motion over the SCS on 19 October, Megi slowed down as it reintensified from category 2 to category 4 with a central SLP of 935 hPa and a maximum 10-m wind speed of 57 m s−1. Early on 20 October, Megi turned north-northeastward. It then weakened to a tropical storm, and made its second landfall at Zhangpu in Fujian Province, China, on 23 October and finally became a TD and dissipated gradually on the next day.

While Megi shared many common features of TCs that crossed Luzon Island (Chou et al. 2011), it also experienced some unique intensity, structural, and track changes. In addition to the subtle track and intensity changes (Fig. 1), Megi also experienced interesting structural changes (Fig. 3). For example, deep convection in the eyewall was widening without any signal of an eyewall contraction during RI. This is different from the majority of TCs experiencing RI. The RI ended as a concentric eyewall signal appeared before it made landfall over Luzon Island (Fig. 3), a not uncommon process at the end of an RI event (e.g., Kossin and Sitkowski 2009), but with the concentric eyewall cycle being interrupted by landfall for the Megi case. The storm experienced an eyewall breakdown when it crossed Luzon Island, and later on, a new outer eyewall formed at a larger radius as a result of the axisymmetrization of outer spiral rainbands after Megi entered the SCS (Fig. 3). Soon after, a small inner eyewall, which could have been the redevelopment of its original eyewall, appeared for several hours when it moved over the SCS. This could be the first double-eyewall structure observed to date as a result of the reappearance of the original eyewall within a newly formed outer eyewall. Compared with the well-studied Typhoon Zeb of 1998 (Wu et al. 2003, 2009), Megi experienced much richer structural changes, such as the lack of the eyewall contraction during RI before landfall and the development of the concentric eyewall structure, as well as a reintensification as it entered the SCS.

Fig. 3.
Fig. 3.

The satellite images at given times showing (top) the structural change before Typhoon Megi (2010) made landfall over Luzon Island and (bottom) the remarkable structural changes when the typhoon crossed Luzon Island and after it entered the SCS.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

In this study, based on a reasonable, week-long control simulation of Typhoon Megi in Wang et al. (2013), we focus on understanding some unique features of Megi, including its RI with no eyewall contraction, its structural changes during its landfall over Luzon Island, and its reintensification after it entered the SCS. In this paper, after a brief introduction of the high-resolution control simulation of Megi, we will present the analyses of RI before Megi made landfall over Luzon Island. The rest of the paper is organized as follows. Section 2 describes briefly the model setup, the dynamical initialization for Typhoon Megi, and the verification of the control simulation. The RI processes of the simulated Megi are analyzed in section 3. Our major results are summarized and discussed in the final section.

2. Model setup, dynamical TC initialization, and verification of simulation

a. Model setup

The numerical simulation for Typhoon Megi presented in this study was performed using version 3.3.1 of the Advanced Research Weather Research and Forecasting Model (ARW-WRF; Skamarock et al. 2008). Details of the model settings can be found in Wang et al. (2013). Table 1 summarizes the main parameters and selections of different physics parameterizations used in the simulation. The model domain is two-way interactive and triply nested (Fig. 4). The three meshes have sizes of 455 × 375, 436 × 436, and 328 × 328 grid points and horizontal grid spacings of 18, 6, and 2 km, respectively. The resolutions of the terrain height and land-use data for the three meshes are 5 min, 2 min, and 30 s (about 9, 4, and 1 km), respectively. There are 36 uneven σ levels in the vertical using terrain-following, hydrostatic pressure as the vertical coordinate extending from the surface to the model top at 50 hPa. While the outermost 18-km mesh is fixed, the two nested inner meshes automatically move following the TC during the model integration so that the model TC is always located near the mesh centers.

Table 1.

Configuration of the ARW-WRF used for the simulation of Megi in this study.

Table 1.
Fig. 4.
Fig. 4.

Model domains with triply nested, movable meshes used for the Typhoon Megi (2010) simulation in this study.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

The model physics include (i) the WRF single-moment 6-class cloud microphysics scheme (WSM6; Hong and Lim 2006) for grid-scale moist processes; (ii) the Mellor–Yamada–Nakanishi–Niino (MYNN) level-2.5 turbulence closure scheme (Nakanishi and Niino 2004) for subgrid-scale vertical mixing coupled with the Monin-Obukhov similarity theory for surface flux calculations over the ocean where the roughness length for momentum is modified for TC strength winds (Moon et al. 2007); (iii) the Rapid Radiative Transfer Model (RRTM; Mlawer et al. 1997) for longwave radiation calculation and the Dudhia scheme (Dudhia 1989) for shortwave radiation calculation; (iv) the Noah land surface scheme (Ek et al. 2003) for land surface processes; and (v) the Kain–Fritsch scheme (Kain 2004) for subgrid-scale deep and shallow convection parameterization in the outermost domain. Convection is assumed to be explicitly resolved in the two nested inner meshes. Dissipative heating is considered in all meshes.

The simulation discussed in this study was the control simulation documented in Wang et al. (2013). The model was initialized at 0000 UTC 15 October 2010 and integrated for 168 h up to 0000 UTC 22 October. The model initial and lateral boundary conditions for both the dynamical TC vortex initialization and the model simulation were interpolated from the National Centers for Environment Prediction (NCEP) Global Forecast System (GFS) Final Analysis (FNL), which has horizontal resolution of 1° × 1° on 27 uneven pressure levels. The daily SST dataset at 1° × 1° resolution is also from the GFS FNL. Since our interest is in achieving improved simulation, not prediction, we applied spectral nudging (SN) to preserve the large-scale flow with wavelengths longer than 1000 km both in the dynamical initialization and throughout the model simulation (Wang et al. 2013). This is acceptable since our focus is on both subsynoptic- and mesoscale processes. The large-scale SN is only applied to the outermost domain, which provides the lateral boundary conditions to the two nested inner meshes.

b. Dynamical initialization for Typhoon Megi

Since both the intensity and structure of Typhoon Megi in the FNL are unrealistic (not shown), we used the dynamical initialization (DI) scheme with the large-scale SN to spin up the axisymmetric TC vortex, as documented in Wang et al. (2013). The core of the DI scheme is to spin up the axisymmetric TC vortex through the 3-h cycle runs [instead of 1-h cycle runs used in Nguyen and Chen (2011)] and the combined use of the large-scale SN. This allowed the TC vortex to better adapt to the environment and to achieve the dynamical balance more sufficiently. The vortex separation algorithm developed by Kurihara et al. (1993) and later modified by Nguyen and Chen (2011) was utilized to subtract the axisymmetric vortex. In the first cycle run, the axisymmetric TC vortex in the FNL was relocated to the observed TC center. From the second cycle run, the axisymmetric TC vortex from the current cycle run was used to replace the axisymmetric TC vortex in the previous cycle run. The cycle run was repeated until the intensity of the TC vortex was comparable to that observed.

The DI scheme was used to spin up Typhoon Megi at 0000 UTC 15 October 2010. After five cycle runs, the intensity of the modeled Megi with the central SLP was 956.4 hPa and the maximum 10-m sustained wind speed was 45.2 m s−1, which are comparable with the observed results in the JTWC best track data (956 hPa and 45 m s−1, respectively). However, in the FNL field, the central SLP pressure of the TC vortex is 1002 hPa, which is too weak compared to the 956 hPa from the JTWC best track data, mainly because the storm core was not resolved well in the coarse-resolution global analysis.

c. Verification of simulation

The track and intensity of the simulated Megi are compared with those from the JTWC best track data in Fig. 1. The model simulated the west-northwestward movement in the first 2 days and the west-southwestward movement before the storm made landfall over Luzon Island. The location of landfall over Luzon Island was also captured in the model simulation except that the simulated storm made landfall about 4–5 h later than the observed. Nevertheless, the model captured nicely the timing of the observed northward-turning motion on 20 October (Fig. 1a). This can be attributed to the use of the large-scale SN, which kept the large-scale environmental flow in the model simulation close to that in the FNL, as demonstrated in Wang et al. (2013).

The model simulated reasonably well the intensity change as well in terms of the central SLP and the maximum sustained surface wind speed (Figs. 1b and 1c). The model captured the RI of Megi before it made landfall over Luzon Island, the rapid weakening over Luzon Island, and the reintensification of the storm after it entered the SCS. However, the model failed to simulate the slow weakening in the last 2 days of the simulation and thus overestimated the storm intensity after 0600 UTC 20 October when the storm moved north-northeastward over the SCS. This might be partially due to the exclusion of the negative ocean feedback because the daily mean SST was used in the simulation.

Verification of the simulated storm structure is difficult since few observations are available over the open ocean. Here, we compare the modeled and satellite-observed infrared cloud-top brightness temperature (TBB) during the RI period. The modeled TBB was calculated based on our outgoing longwave radiation (OLR) and the satellite TBB is the high-resolution (1 km) Multifunctional Transport Satellite (MTSAT) infrared image. Figure 5 shows the modeled and satellite-observed TBB at three given times. At 0300 UTC 16 October just after RI started, Megi displayed considerable convective asymmetric structure in the eyewall with an enhanced convective area in the northwest quadrant and outer spiral rainbands that extended southeastward east of the eyewall. By 2000 UTC 16 October, Megi was in the middle of RI and showed a clear eye and closed eyewall with a rainband extended from the north to the east. By 2300 UTC 17 October, just several hours before the storm made landfall, the storm increased its convective coverage in the inner-core region with a much-widened eyewall and somewhat of an increase in the eye diameter.

Fig. 5.
Fig. 5.

(a)–(c) (top) The model-simulated TBB (°C) and (bottom) the satellite infrared high-resolution (1 km) TBB at times given at the top of each panel for Typhoon Megi.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

The modeled TBB is not as low as the observed, indicating that convection in the eyewall in the simulation was not as deep as in the observation. This might be partly because the model resolution is not high enough to resolve the convective cores in the eyewall and partly because the model vertical grid spacing is not high enough near the cloud top to properly resolve the thick cirrus outflow cloud structure. In spite of the systematic bias in TBB, the modeled Megi shows a relatively larger eye size than the observed. Nevertheless, the modeled storm has a similar trend in the widening of the eyewall during RI to the observed Megi. Realistic simulation of the eyewall structure and its evolution are still quite challenging. Therefore, in our analysis below we keep in mind that the simulated Megi is not the real Typhoon Megi. Nevertheless, we consider that understanding of the model Megi could still have implications for the real Typhoon Megi. Note that we only focus on the RI phase of the modeled Megi in this part of this paper, the detailed structural changes after Megi made landfall and entered the SCS will be analyzed in a forthcoming Part II.

3. RI of Typhoon Megi

Megi experienced its RI from 0000 UTC 16 October to 1200 UTC 17 October and reached its peak intensity with a central SLP of 905 hPa and a maximum 10-m wind speed of 80 m s−1. In this section, the RI processes of the modeled Megi will be analyzed based on the high-resolution model simulation discussed in section 2c. Following the analysis of Rogers (2010), we will discuss the large-scale settings, vortex-scale evolution, and convective activities during RI of the simulated Megi.

a. Large-scale settings

The RI of a TC often occurs under some favorable environmental conditions, including high SST and large upper OHC, weak vertical wind shear (VWS), and high midtropospheric relative humidity (Molinari et al. 1995; Bosart et al. 2000; Kaplan and DeMaria 2003; Lin et al. 2008). In particular, both the OHC and VWS are key environmental parameters for RI (Park et al. 2013). A TC experiencing RI obtains considerable energy from the underlying ocean. High upper OHC serves as the energy source for RI and also limits the negative effect due to ocean upwelling resulting from the intensifying cyclonic wind stress curl and turbulent vertical mixing across the mixed layer base (Cione and Uhlhorn 2003; Lin et al. 2008). Megi traveled over the warm WNP with SSTs above 29°C along its track (Fig. 2a) and high upper OHC of over 100 kJ cm−2 (Fig. 2b) during its RI period. Therefore, the ocean conditions are favorable for the RI of Typhoon Megi.

VWS is a key atmospheric parameter for the RI of a TC. A common explanation of the effect of VWS concerns the ventilation of the upper-level warm core relative to the low-level cyclonic circulation, inhibiting the deepening of a TC (Gray 1968). Weak VWS can ensure that the warm core forms and is maintained over the low-level cyclonic circulation center and thus is favorable for RI. Megi intensified rapidly when the VWS magnitude decreased sharply from 10–12 to 1–2 m s−1, regardless of the fact that the averaged winds were calculated in the inner-core region or a large area (Fig. 6). This is consistent with the result of Chen and Zhang (2013) for Hurricane Wilma (2005). Note that the VWS calculated from the model simulation (Fig. 6) has the same trend as that calculated from the FNL but the magnitude is slightly larger, partially due to the coarse resolution of the FNL (not shown).

Fig. 6.
Fig. 6.

Time evolution of vertical wind shear over the storm, shown are large-scale shear averaged between radii of 502 and 900 km (dotted) and shear between radii of 396 and 600 km (dashed) from domain 2, as well as shear averaged within 250-km radius in the nested domain 3 (solid).

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

Figure 7 shows the 200- and 850-hPa geopotential and wind fields at 0000 UTC 15 and 17 October from the FNL. On 15 October prior to RI, Megi already developed an outflow layer with an outflow jet toward the northeast and then east-southeast at 200 hPa (Fig. 7a). The main midlatitude westerlies were located north of 30°N, far away from Megi. In the lower troposphere, Megi was located south-southwest of the WNP subtropical high and was embedded in a large-scale cyclonic circulation (depression) centered near 10°N, 130°E to the south-southwest of Megi (Fig. 7c). By 17 October during RI, the upper-level outflow intensified significantly (Fig. 7b). In addition to the outflow jet to the northeast, an outflow channel appeared southwest of the storm center. At this point, the midlatitude upper-tropospheric westerlies extended southward, reaching 25°N and potentially enhancing the outflow jet to the northeast of the storm (Fig. 7b). In the lower troposphere (Fig. 7d), Megi was still located in the easterlies south of the WNP subtropical high.

Fig. 7.
Fig. 7.

Wind vectors and geopotential height field (contours, gpm) from the NCEP GFS FNL at (a),(b) 200 and (c),(d) 850 hPa at 0000 UTC on (a),(c) 15 and (b),(d) 17 Oct 2010.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

b. Vortex- (storm) scale evolution

Figure 8 shows the time evolution of several azimuthal mean variables of the simulated Typhoon Megi. Megi exhibited a slow contraction in the first 18 h of simulation before it started its RI phase as inferred from the time evolution of the azimuthal mean tangential wind (Fig. 8a). The radius of maximum wind (RMW) of the simulated Megi was reduced by about 10 km from 50 to 40 km during this period. Although the RMW remained almost unchanged during the subsequent RI phase, the tangential wind field expanded outward considerably until the end of RI just before Megi made landfall over Luzon Island. For example, the radius of the azimuthal mean hurricane force wind (33 m s−1) was doubled, from 105 km at 0000 UTC 16 October to 210 km at 2100 UTC 17 October. This outward expansion of the azimuthal mean tangential wind is consistent with the widening of the eyewall convection in both the observed and simulated Typhoon Megi, as seen in Fig. 5. Although this outward expansion is partly a result of the storm intensification, showing little evidence of outward expansion of the normalized azimuthal mean tangential wind by its maximum value at the corresponding time (Fig. 8a), it also indicates an increase of inner-core size as inferred from the widening of eyewall convection as already discussed in Wang and Wang (2013). This is in sharp contrast to the majority of TCs during RI, where an eyewall contraction often occurs during RI with little increase in the outer-core circulation strength (Ooyama 1982; Schubert and Hack 1982; Holland and Merrill 1984; Rogers 2010).

Fig. 8.
Fig. 8.

The radius–time cross section of the azimuthal mean (a) tangential wind (m s−1) at 1-km height (contours) normalized by the maximum value at a given time (shading), (b) vertical velocity (cm s−1) at 3-km height, (c) normalized inertial stability (shading) and radial wind (m s−1, contours) at the lowest model level (about 28 m above the surface), and (d) surface rain rate (mm h−1).

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

Similar to the time evolution of the azimuthal mean tangential wind, the azimuthal mean vertical motion shows an eyewall contraction in the first 18 h of simulation as well but shows no eyewall contraction during the subsequent RI period (Fig. 8b). In addition to strong eyewall updrafts, relatively large azimuthal mean upward motion outside the eyewall reflects active spiral rainbands throughout the simulation. This suggests the possible contribution of diabatic heating in spiral rainbands in preventing the eyewall contraction and meanwhile increasing the outer-core strength (and the inner-core size) of the simulated Megi during RI (Wang and Wang 2013), a process previously proposed to explain the inner-core size increase in idealized simulations by Wang (2009) and Xu and Wang (2010a,b).

Figure 8c shows the time evolution of the normalized inertial stability (normalized by the square of the Coriolis parameter) at the lowest model level in the simulated Megi. Inertial stability2 is an important dynamical parameter that can measure the resistance to radial inflow and also determine the efficiency of the upper-tropospheric warming by eyewall heating (Shapiro and Willoughby 1982; Hack and Schubert 1986). Rogers (2010) suggested that the time evolution of inertial stability in the inner core could be an indicator of whether a TC is undergoing its RI. The inertial stability in the simulated Megi shows an off-center maximum inside the RMW prior to and during RI. This distribution is consistent with the off-centered maximum relative vorticity or potential vorticity (PV) distribution in the lower troposphere in a TC (e.g., Wang 2008a). After the onset of RI on 0000 UTC 16 October, inertial stability inside the RMW increased steadily as Megi intensified. Note that although inertial stability is the highest inside the RMW, relatively high azimuthal mean inertial stability is obvious outside the RMW and expanded radially outward during RI. This outward expansion of relatively high inertial stability is consistent with the outward expansion of the azimuthal mean tangential wind and the increase in the inner-core size.

Radial inflow in the boundary layer penetrated progressively inward outside the RMW and decelerated sharply near the RMW where the radial gradient in inertial stability is the largest (Fig. 8c). The maximum azimuthal mean radial inflow was only about 10 m s−1 prior to the onset of RI but increased to 20 m s−1 near the radius of about 50 km immediately outside the eyewall by the end of RI. The increase in both low-level inflow and inertial stability inside the RMW implies the strengthening of the radial mass and moisture convergence in the boundary layer, contributing to strong eyewall updrafts and intense convection in the eyewall. Latent heating released in eyewall convection would effectively enhance the upper-tropospheric warming of the high inertial stability core (Schubert and Hack 1982) and lower the central SLP of the storm. The latter in turn would drive stronger boundary layer inflow, further enhancing eyewall updrafts and convection. This forms a positive feedback (among inertial stability, eyewall updraft and convection, and upper-tropospheric warming) that contributes to RI of the simulated Megi, as collectively discussed in earlier studies (e.g., Ooyama 1982; Schubert and Hack 1982; Hack and Schubert 1986; Vigh and Schubert 2009; Smith et al. 2009).

Another important parameter is the axisymmetricity, which is a measure of the degree of axisymmetry for a TC vortex (e.g., Fudeyasu et al. 2010; Miyamoto and Takemi 2013). Fudeyasu et al. (2010) defined the axisymmetricity as the ratio of the azimuthal mean kinetic energy to the total kinetic energy within a radius of 301 km. Miyamoto and Takemi (2013) defined the axisymmetricity as the ratio of the squared azimuthal mean PV to the azimuthal mean squared total PV. Here, following Fudeyasu et al. (2010), we define the axisymmetricity using kinetic energy. Note that instead of the area-averaged axisymmetricity used in Fudeyasu et al. (2010), we calculated the axisymmetricity for each radial band of 2 km to examine the time evolution of axisymmetricity as a function of radius and height. Figure 9 shows the time evolution of the axisymmetricity in the simulated Typhoon Megi as a function of the radius averaged in three layers, representing the lower, middle, and upper troposphere, respectively. Consistent with earlier studies, the storm experienced axisymmetrization with the onset of RI, in particular in the middle troposphere in the simulated Megi case. The axisymmetrization started from the mid- to lower troposphere upward during RI. The maximum axisymmetricity parameter (close to 1.0) appeared just outside the eyewall where the filamentation time is relative short (Rozoff et al. 2006; Wang 2008a). The storm became more axisymmetric at larger radii during RI.

Fig. 9.
Fig. 9.

Radius–time cross section of the axisymmetricity parameter averaged in the layers between (a) 2 and 6, (b) 7 and 11, and (c) 12 and 16 km, respectively.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

The axisymmetrization during RI was also a subsequence of the merging of a low-level synoptic depression into the typhoon vortex. As already discussed in Wang and Wang (2013), there was a binary interaction between the typhoon vortex and a low-level synoptic depression in which Megi was embedded prior to RI (Figs. 7c and 7d). This binary interaction played a critical role in causing the inner-core size increase of Megi during its RI phase. On one hand, the shear deformation, filamentation, and merging of the low-level synoptic-scale depression by the typhoon vortex and the subsequent axisymmetrization enlarged both the inner- and outer-core sizes of Megi considerably. On the other hand, the binary interaction also played a role in triggering/enhancing active spiral rainbands in Megi. Diabatic heating in spiral rainbands drove inflow in the mid- to lower troposphere, accelerating tangential winds outside the RMW, leading to the significant outward expansion of the tangential wind and thus the inner-core size increase of the storm. This also prevented the eyewall’s contraction during its RI phase.

c. Convective-scale structure and evolution

Convective activity in the inner-core region is the key to the upward transport of energy gained from the ocean to the atmosphere above and, thus, is critical to the intensification and maintenance of a TC. Here, we will examine the convective-scale structure and evolution in the simulated Megi prior to and during its RI stage following the analysis done by Rogers (2010). We first examine the contoured frequency by altitude diagrams (CFADs; Yuter and Houze 1995) for vertical velocity; the CFADs illustrate the frequency distribution of the vertical velocity of the indicated values at different altitudes. Figure 10 shows 24-h-averaged CFADs of the simulated vertical velocity binned every 0.1 m s−1 at each altitude within a radius of 80 km. Three panels in Fig. 10 show the three stages of the simulated Megi: prior to and after the onset of RI, and during RI, respectively.

Fig. 10.
Fig. 10.

CFAD (%) of simulated vertical velocity within 80-km radius in 24-h periods on (a) 15, (b) 16, and (c) 17 Oct.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

A comparison of the vertical velocity CFADs 24 h prior to and after the onset of RI (Figs. 10a and 10b) indicates that the distributions are broader on 16 October after the onset of RI, with peak updrafts of 24 m s−1 and downdrafts of −8 m s−1 compared to peak updrafts of 20 m s−1 and downdrafts of −11 m s−1 on 15 October prior to the onset of RI. Nevertheless, the overall distributions of the CFADs 24 h prior to and after the onset of RI were very similar, with peak updrafts at 13–14-km altitude. Large differences occurred on the second day of RI (17 October; Fig. 10c). Compared to the first day of RI, the intense updrafts (top 0.01%) were largely reduced at the upper levels but increased at the middle levels, with the peak of updrafts at 8–9-km altitude. Another distinct feature were the greatly reduced downdrafts, in particular in the mid- to upper troposphere. The downdrafts showed a second peak in a layer between 4- and 5-km altitude. This was where the freezing level was located in the inner-core region, suggesting that downdrafts were driven by both the melting of snow and graupel and the evaporation of rain. The upper-level peak in downdrafts on 15 and 16 October could be partially due to sublimation of ice-phase hydrometeors partially forced by large vertical wind shear (Fig. 6).

The above features of vertical velocity in the inner-core region can be seen more clearly in the time evolution of the cumulative contour frequency for both updrafts and downdrafts within a radius of 80 km at three altitudes (2, 8, and 13 km), as plotted in Fig. 11. The three levels are chosen as representatives of the lower, middle, and upper troposphere, respectively. The distribution of the cumulative contour frequency for vertical motion shows upward motion confined mostly within 4 m s−1 throughout the depth of the troposphere during the analyzed time period. There are also differences in magnitude from the middle to the upper troposphere: the outliers (the envelope of the extreme values) at 8-km altitude increased during RI, while those at the upper levels showed a shrinking trend. The outliers of the upward motion at 13 km increased prior to RI and reached their maximum at the beginning of RI, and then decreased continuously during RI. A similar narrowing of the outliers in vertical velocity in the upper troposphere was also found during RI of the simulated Hurricane Dennis (2005) by McFarquhar et al. (2012). They suggested that the narrowing could be related to the increased precipitation loading. In the simulated Megi, the increased outward tilt of the eyewall (Wang 2008b) and the stabilization due to the rapid upper-level warming could be responsible for the narrowing of the cumulative contour frequency distribution for both updrafts and downdrafts at upper levels in the inner-core region.

Fig. 11.
Fig. 11.

Cumulative contour frequency of distributions of vertical velocity within 80-km radius as a function of time for the simulated Typhoon Megi at (a) 2-, (b) 8-, and (c) 13-km heights.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

The results above suggest that convection in the inner-core region become less strong and less penetrative as the simulated Megi rapidly intensified. It is interesting to examine the evolution of the area coverage by convection in the inner-core region during RI. Based on the convective-stratiform partitioning algorithm used in Steiner et al. (1995) and Rogers (2010), we obtained the time evolution of different precipitation types for both the inner-core region (within a radius of 80 km) and the outer-core region (between radii of 80 and 160 km) as shown in Fig. 12. The percentages of convective coverage decreased slightly in both regions before the onset of RI. After the onset of RI, the convective coverage in the inner-core region increased steadily from about 50% to 80% during RI. The convective coverage in the outer-core region showed little change and remained around 30% from 1500 UTC 15 October to 1200 UTC 17 October, and then slightly increased to about 40% later on 17 October, which reflects the enhanced convection in the outer-core region due to the interaction of the storm with orography when it approached Luzon Island. The percentage coverage by stratiform precipitation in the inner-core region increased from 15% to 35% in the first 12 h of RI, and then decreased to about 5% by the end of RI. In the outer-core region, the stratiform coverage increased from 20% prior to the onset of RI to about 60% by the end of RI, with significant fluctuations in between. This increase in stratiform coverage in the outer-core region was mainly related to the active outer spiral rainbands (Li and Wang 2012) and played an important role in preventing the eyewall contraction while increasing the inner-core size during RI of the simulated Meg (Wang and Wang 2013).

Fig. 12.
Fig. 12.

Time series of the convective (solid) and stratiform (dashed) percentages (a) within 80-km radius and (b) between 80- and 160-km radii.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

The above results suggest that the percentage of convective coverage increased while the cumulative contour frequency of vertical velocity in the inner-core region decreased at the upper levels during RI of the simulated Megi. Previous studies have found that convective bursts (CBs, deep and intense convective cells) are important to RI of TCs (Kelley et al. 2005; Montgomery et al. 2006; Reasor et al. 2009; Rogers 2010; Zhang and Chen 2012; Chen and Zhang 2013). This indeed is the case, as we can see from Fig. 13, which shows the time evolution of the total number of grid points with CBs defined as vertical velocity at 11-km altitude greater than 7.5 m s−1 (Fig. 13a) or as the maximum vertical velocity in the layer between 2 and 12 km greater than 7.5 m s−1 (Fig. 13b) within a radius of 80 km in the simulated Megi case.3 Consistent with the results of Chen and Zhang (2013), in the simulated Megi CBs in the inner-core region showed a significant increase associated with the onset of RI and then a decrease during RI. This indicates that CBs played an important role in triggering the initial warm-core formation while they would be suppressed as the storm rapidly intensified, although the convective coverage increased in the inner-core region (Fig. 12).

Fig. 13.
Fig. 13.

Number of grid points with convective bursts within 80-km radius: (a) convective bursts where the vertical velocity is greater than 7.5 m s−1 at 12-km height and (b) the maximum vertical velocity in the column between 2- and 12-km heights is greater than 7.5 m s−1.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

A question arises as to what triggered CBs during the onset of RI. The occurrence of maximum updrafts in the upper troposphere suggests the dominant control of CBs by convective available potential energy (CAPE), which is defined as
e1
where the subscript “env” denotes environmental variable while “parcel” denotes the variable of a parcel, Tυ is virtual temperature, zLFC is the level of free convection (LFC), and zLNB is the level of neutral buoyancy (LNB) following the motion of the parcel. CAPE is a measure of the maximum kinetic energy per unit mass that a buoyant parcel could obtain by ascending from a state of rest at the LFC to the LNB near the tropopause. The maximum vertical motion is thus expected to occur near the LNB and can be approximated by . Note that the CAPE defined in (1) is an overestimate of the actual updraft velocity because both the entrainment of the environmental air and the negative buoyancy due to water loading are ignored. Figure 14a shows the time evolution of the azimuthal mean CAPE together with the boundary layer equivalent potential temperature θe and the RMW. The azimuthal mean CAPE as well as boundary layer θe inside the RMW increased with time prior to the onset of RI of the simulated Megi. Large CAPE occurred in the eye region within a radius of 30 km prior to RI while decreased with time during RI. However, θe in the eye region increased continuously during RI.
Fig. 14.
Fig. 14.

(a) Time–radius cross section of the azimuthal mean CAPE (J kg−1, shading), equivalent potential temperature (green contours with interval of 5 K), and RMW (dashed white). (b) The radial distribution of the number of grid points with vertical velocity greater than 7.5 m s−1 at 11-km altitude between 1200 UTC 15 Oct and 1200 UTC 16 Oct 2010.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

In an observational study for Hurricane Lili (2002), Barnes and Fuentes (2010) defined an eye excess energy as the difference in θe between the eye and the eyewall and in the depth over which θe in the eye is larger than that in the eyewall. They found that the eye excess energy was large at the beginning of RI but diminished during RI. They hypothesized that the eye excess energy might serve as a boost for deep convection in the eyewall and thus the onset of RI. In a recent study, Miyamoto and Takemi (2013) found high CAPE in the eye region prior to RI and suggested that the high CAPE was accumulated due to the long residence time of boundary layer air inside the RMW where inertial stability was high. They proposed that high CAPE triggered deep convection inside the RMW and the onset of RI. However, they did not show whether or how the enhanced eyewall convection was really supported by high CAPE in the eye region.

In the simulated Megi case, CBs associated with the onset of RI and in the early stage of RI (from 1200 UTC 15 October to 1200 UTC 16 October) occurred mainly between 40- and 70-km radii (Fig. 14b), outside the region with high CAPE (Fig. 14a). Nevertheless, considering the outward slope of eyewall updrafts [slantwise nature of eyewall convection; Emanuel (1987)], high CAPE in the eye region could contribute to CBs in the eyewall and to the onset of RI in the simulated Megi although CAPE in the eye region might contribute little to the maximum storm intensity at the mature stage (Bryan and Rotunno 2009; Wang and Xu 2010). This can be demonstrated by the azimuthal mean vertical velocity and radial wind averaged on 15 and 16 October, respectively, as shown in Fig. 15. We can see that upward motion in the eyewall tilted outward with height with the maximum upward motion at 11 km occurring at a radius of 70 km on 15 October and of 55 km on 16 October, respectively. The inner edge of the upward motion at the top of the inflow boundary layer was located at about 25 km on 15 October and 20 km on 16 October, respectively, corresponding to low-level outflow there as a result of supergradient wind (Kepert and Wang 2001). This indicates that boundary layer high-θe air in the eye region was being transported into the eyewall, increasing the buoyancy, and thus enhancing updrafts, convection, and CBs in the eyewall.

Fig. 15.
Fig. 15.

The radius–height cross section of the azimuthal mean vertical velocity (m s−1, shading) and radial wind (m s−1, contours) averaged on (a) 15 and (b) 16 Oct of the simulated Megi based on hourly model outputs.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

In addition, the effect of high CAPE in the eye region on eyewall updrafts can also be understood in terms of lateral mixing between the eye and the eyewall. These processes have been well demonstrated in several previous studies (Braun 2002; Persing and Montgomery 2003; Barnes and Fuentes 2010; Wang and Xu 2010). The decrease in CAPE in the eye region during RI is consistent with both the decrease in the frequency of occurrence of CBs (Fig. 13) and the lowering of the height with maximum upward motion in the inner-core region (Fig. 10). The decrease in CAPE in the eye region during RI indicates the stabilization of the air column in the eye region due to the rapid development of the upper-level warm-core structure. Although deep convection in the eyewall might consume some CAPE in the eye region through lateral mixing, this could be secondary since θe remained high or even increased during RI (Fig. 14a).

The above process can be understood alternatively in the notion of slantwise CAPE (SCAPE; e.g., Emanuel 1988; Shutts 1990; Gray and Thorpe 2001). In (1), CAPE is defined as the undiluted air parcel ascending vertically. However, this is not the case in the eyewall of a strong TC since the eyewall ascent generally follows the absolute angular momentum (AAM) surface. Braun (2002) showed that SCAPE should be larger than CAPE in the eyewall because a part of the air parcels have their origin in the eye region in the boundary layer (see his Fig. 18). In a more recent study, Frisius and Schönemann (2012) showed that SCAPE arises in the boundary layer outside of the eyewall and could contribute to the maximum potential intensity of a TC. Here, we would show that SCAPE arising from the eye region does contribute to buoyancy and CBs in the eyewall. To simplify the calculation of SCAPE in the simulated Megi, we assume that the simulated storm was quasi-axisymmetric and estimate the SCAPE along the azimuthal mean AAM surface. In this case, SCAPE can be calculated using the following modified form of (1):
e2
where the buoyancy at each height on the right-hand side was calculated following the azimuthal mean AAM surface with the origin at the lowest model level (we tested different starting levels between the lowest model level and 1-km height; the results were quite similar). Figure 16 shows the time evolution of the azimuthal mean SCAPE, together with the vertical velocity at 1-km height. In general SCAPE shows the distribution and time evolution very similar to the CAPE shown in Fig. 14a, but with relatively higher values extending to the inner edge of the eyewall updrafts at the top of the inflow boundary layer. We can see clearly that part of the parcels arising in the eyewall stem from the eye region with high SCAPE, indicating that SCAPE in the eye region contributes considerably to the slantwise convection and upper-tropospheric CBs in the simulated Megi. Therefore, SCAPE provides buoyancy to convection and CBs in the mid- to upper troposphere, contributing to the onset of RI and the early RI of the simulated Megi.
Fig. 16.
Fig. 16.

As in Fig. 14a, but for SCAPE calculated with the ascending air parcel arising from the lowest model level and following the azimuthal mean AAM surface and vertical velocities at 1-km altitude, with a contour interval of 0.2 m s−1 starting from 0.1 m s−1.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

d. Development of the upper-level warm core and RI

Figure 17a shows the time evolution of the temperature anomalies and potential temperature averaged in the eye region (within 20-km radius) of the simulated Megi. The warm core first appeared in the midtroposphere between 3 and 8 km prior to RI. The midlevel warm core remained in both height and magnitude during RI, suggesting that it contributed little to the RI of the simulated Megi. A new upper-level warm core formed at about 16–17 km after the RI started, and then amplified and extended downward during RI. By the end of RI, the upper-level warm core was located between 14- and 16-km heights. This downward development of the upper-level warm core is consistent with the descent of the contour of 380-K potential temperature that originated from the tropopause. This suggests that the high potential temperature air detrained from the lower stratosphere by CBs could contribute to RI of the simulated storm. The results suggest that in the environment with decreasing vertical wind shear, CBs triggered the onset of RI and the formation of the upper-tropospheric warm core with a possible contribution from the high potential temperature air detrained from the lower stratosphere in the simulated Megi (Fig. 17a). This is similar to that documented for Hurricane Wilma (2005) by Chen and Zhang (2013), although Wilma was stronger than Megi. Chen and Zhang (2013) showed that the air detrained from the eyewall surrounding tall CBs into the eye region warmed the TC eye, initiating the formation of the upper-level warm core and the onset of RI. In the simulated Megi case, CBs were not as strong as those in Hurricane Wilma, but the overall structure and evolution were similar.

Fig. 17.
Fig. 17.

Time evolution of the (a) temperature anomalies (K, shading) and equivalent potential temperature (K, contours) averaged within a 20-km radius and (b) inertial stability parameter in the boundary layer (s−2, red) and CAPE (J kg−1, blue) averaged within a 60-km radius of the simulated Megi.

Citation: Monthly Weather Review 142, 1; 10.1175/MWR-D-13-00070.1

A question arises as to why CBs became less active as the storm rapidly intensified. We have already mentioned above that the weakening of CBs was consistent with the decrease in CAPE in the eye region (Figs. 14a and 17b). This is largely due to the increased static stability as a result of the upper-level warming (Fig. 17a). The reduction in CBs does not mean the slowing down of the upper-level warming. Since the efficiency of the upper-level warming is not only determined by CBs and updrafts in the eyewall but also depends on the inertial stability in the region of diabatic heating (Schubert and Hack 1982; Hack and Schubert 1986). As shown in Schubert and Hack (1982), the efficiency of the upper-level warming induced by eyewall convective heating increases as inertial stability in the inner-core region increases. In the simulated Megi case, the inertial stability in the inner-core region increased continuously throughout the troposphere (similar to that in Fig. 17b) as the storm rapidly intensified, indicating the increasing efficiency in the upper-tropospheric warming forced by eyewall convection during RI.

Vigh and Schubert (2009) examined the dependence of the warm-core development on the radial location of the diabatic heating based on the balanced vortex model. They found that the response of the upper-level warming to diabatic heating reaches its maximum when diabatic heating is centered in the region with high inertial stability inside the RMW. In the simulated Megi case, diabatic heating in the eyewall was indeed located inside the RMW and near the edge of the maximum radial gradient of low-level inertial stability during RI (Fig. 8). The large radial gradient of inertial stability also infers the large deceleration of the boundary layer inflow and, thus, the strengthening of the forced upward motion in the eyewall. Therefore, although CBs became less active as the storm intensified rapidly, the mean eyewall updrafts were still strong but less penetrative than those prior to RI (Fig. 10). This led to the lowering of the height of the upper-level warm core during RI (Fig. 17a).

In their numerical study for Hurricane Wilma (2005), Chen and Zhang (2013) showed that there was a shallow inflow layer immediately above the prevailing upper-level outflow layer, which originated in the lower stratosphere and emerged at the time of the eye formation. They considered the shallow inflow layer to be a result of mass continuity due to the mass sink in the eye region as the storm rapidly intensified. They also argued that the inflow can effectively carry environmental potential temperature air all the way into the eye, where air descends adiabatically to enhance the warm core due to the absence of inertial stability. We also observed similar processes in the simulated Megi case. Prior to the onset of RI, the upper-level inflow appeared but less organized (Fig. 15a). However, during RI a systematic shallow inflow layer developed immediately above the upper-level outflow layer (Fig. 15b), which arguably contributed positively to the formation and maintenance of the upper-level warm core in the simulated Megi.

4. Conclusions and discussion

Typhoon Megi was the most powerful and longest-lived TC over the WNP and SCS in 2010. While it shared many common features of TCs that crossed Luzon Island in the northern Philippines, Megi experienced unique intensity and structural changes, such as the lack of an eyewall contraction and an increase in inner-core size during its RI phase, drastic structural changes when it made landfall and crossed Luzon Island, and reintensification with the reformation of a large eyewall and the redevelopment of the original eyewall as an inner eyewall after it entered the SCS. Most of these features were reasonably well reproduced in a week-long control simulation using the ARW-WRF with both DI and large-scale SN, as documented in Wang et al. (2013). The DI scheme spun up the axisymmetric component of the TC vortex, which was rather weak in the coarse-resolution global analysis, and the large-scale SN kept the environmental flow with the wavelength longer than 1000 km as close to the driving field as possible in both DI and subsequent simulations.

This has paper presented an investigation into the processes responsible for the RI of the simulated Megi before it made landfall over Luzon Island based on the control simulation of Wang et al. (2013). The results show that, as is the case in most other TC cases studied earlier, Typhoon Megi experienced its RI over the WNP with high upper OHC and in an environment with decreasing vertical wind shear. On the storm scale, inertial stability in the inner-core region was already high prior to the onset of RI and the axisymmetrization throughout the troposphere was accompanied by RI. Frictionally induced boundary layer inflow penetrated progressively inward outside the RMW and decelerated sharply near the RMW where the radial gradient in inertial stability is the largest. This strengthened the radial mass and moisture convergence in the boundary layer, leading to strong eyewall updrafts and intense convection in the eyewall. Latent heating released in eyewall convection effectively enhanced the upper-tropospheric warming of the high inertial stability core and lowered the central sea level pressure of the storm. The latter in turn drove stronger boundary layer inflow, further enhancing eyewall updrafts and convection. This forms a positive feedback among inertial stability, eyewall updraft and convection, and upper-tropospheric warming, contributing to RI of the simulated Megi, as has been collectively documented in earlier studies (Ooyama 1982; Schubert and Hack 1982; Hack and Schubert 1986; Vigh and Schubert 2009; Smith et al. 2009).

Different from the majority of TCs that experience RI with an eyewall contraction, Megi rapidly intensified without any eyewall contraction. The lack of eyewall contraction is attributed to diabatic heating in active spiral rainbands, a process previously proposed to explain the inner-core size increase in idealized simulations by Wang (2009) and Xu and Wang (2010a,b). Actually, in the Megi case the inner-core size also increased considerably during its RI (Wang and Wang 2013). This is largely attributed to the binary interaction between the typhoon vortex and a low-level synoptic depression in which Megi was embedded and diabatic heating in active spiral rainbands enhanced by the binary interaction (Wang and Wang 2013).

On the convective scales, the onset of RI was found to be triggered by CBs in the eyewall, which penetrated into the upper troposphere and played a critical role in triggering the formation of the upper-tropospheric warm core. We have shown that slantwise convective available potential energy (SCAPE) accumulated in the eye region contributed to slantwise eyewall convection and CBs and thus the onset of RI. Once RI started, the convective area coverage in the inner-core region increased while the updraft velocity in the upper troposphere decreased. The number of CBs decreased as well during RI because of the reduction of SCAPE resulting from the stabilization of the inner core due to upper-level warming.

Our results are in agreement with those of Rogers (2010), who studied the role of convective processes in the RI of Hurricane Dennis (2005) via a high-resolution simulation, and those of Chen and Zhang (2013), who numerically studied the RI processes in Hurricane Wilma (2005). Our results however seem to differ from those of McFarquhar et al. (2012), who compared different definitions of CBs and found that in their simulated Hurricane Dennis (2005) the number of CBs did not show any increase prior to RI, but continuously increased during RI. The difference might reflect the case dependence of convective activities. Similar analyses should be undertaken for more cases in future studies to examine the majority of the convective evolution and robustness of the RI triggering mechanisms.

Acknowledgments

We thank two anonymous reviewers for their constructive review comments that resulted in substantial improvements to the manuscript. This study has been supported in part by the National Basic Research Program of China (2009CB421505) and the National Natural Science Foundation of China under Grant 41130964 and in part by NSF Grants ATM-0754039 and AGS-1326524. Additional support has been provided by the JAMSTEC through its sponsorship of the International Pacific Research Center (IPRC) in the School of Ocean and Earth Science and Technology (SOEST) at the University of Hawaii.

REFERENCES

  • Barnes, G. M., , and P. Fuentes, 2010: Eye excess energy and the rapid intensification of Hurricane Lili (2002). Mon. Wea. Rev., 138, 14461458.

    • Search Google Scholar
    • Export Citation
  • Bosart, L. F., , C. S. Velden, , W. E. Bracken, , J. Molinari, , and P. G. Black, 2000: Environmental influences on the rapid intensification of Hurricane Opal (1995) over the Gulf of Mexico. Mon. Wea. Rev., 128, 322352.

    • Search Google Scholar
    • Export Citation
  • Braun, S. A., 2002: A cloud-resolving simulation of Hurricane Bob (1991): Storm structure and eyewall buoyancy. Mon. Wea. Rev., 130, 15731592.

    • Search Google Scholar
    • Export Citation
  • Bryan, G. H., , and R. Rotunno, 2009: The influence of near-surface, high-entropy air in hurricane eyes on maximum hurricane intensity. J. Atmos. Sci., 66, 148158.

    • Search Google Scholar
    • Export Citation
  • Chen, H., , and D.-L. Zhang, 2013: On the rapid intensification of Hurricane Wilma (2005). Part II: Convective bursts and the upper-level warm core. J. Atmos. Sci., 70, 146162.

    • Search Google Scholar
    • Export Citation
  • Chou, K.-H., , C.-C. Wu, , Y. Wang, , and C.-H. Chih, 2011: Eyewall evolution of typhoons crossing the Philippines and Taiwan: An observational study. Terr. Atmos. Oceanic Sci., 22, 535548.

    • Search Google Scholar
    • Export Citation
  • Cione, J. J., , and E. W. Uhlhorn, 2003: Sea surface temperature variability in hurricanes: Implications with respect to intensity change. Mon. Wea. Rev., 131, 17831796.

    • Search Google Scholar
    • Export Citation
  • Dudhia, J., 1989: Numerical study of convection observed during the Winter Monsoon Experiment using a mesoscale two-dimensional model. J. Atmos. Sci., 46, 30773107.

    • Search Google Scholar
    • Export Citation
  • Ek, M. B., , K. E. Mitchell, , Y. Lin, , E. Rogers, , P. Grunmann, , V. Koren, , G. Gayno, , and J. D. Tarpley, 2003: Implementation of Noah land surface model advances in the National Centers for Environmental Prediction operational mesoscale Eta Model. J. Geophys. Res., 108, 8851, doi:10.1029/2002JD003296.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., 1987: An air–sea interaction model of intraseasonal oscillations in the tropics. J. Atmos. Sci., 44, 23242340.

  • Emanuel, K. A., 1988: Observational evidence of slantwise convective adjustment. Mon. Wea. Rev., 116, 18051816.

  • Frisius, T., , and D. Schönemann, 2012: An extended model for the potential intensity of tropical cyclones. J. Atmos. Sci., 69, 641661.

    • Search Google Scholar
    • Export Citation
  • Fudeyasu, H., , Y. Wang, , M. Satoh, , T. Nasuno, , H. Miura, , and W. Yanase, 2010: Multiscale interactions in the life cycle of a tropical cyclone simulated in a global cloud-system-resolving model. Part I: Large-scale and storm-scale evolutions. Mon. Wea. Rev., 138, 42854304.

    • Search Google Scholar
    • Export Citation
  • Gray, S. L., , and A. J. Thorpe, 2001: Parcel theory in three dimensions and the calculation of SCAPE. Mon. Wea. Rev., 129, 16561672.

  • Gray, W. M., 1968: Global view of the origin of tropical disturbances and storms. Mon. Wea. Rev., 96, 669700.

  • Hack, J. J., , and W. H. Schubert, 1986: Nonlinear response of atmospheric vortices to heating by organized cumulus convection. J. Atmos. Sci., 43, 15591573.

    • Search Google Scholar
    • Export Citation
  • Holland, G. J., , and R. T. Merrill, 1984: On the dynamics of tropical cyclone structural changes. Quart. J. Roy. Meteor. Soc., 110, 723745.

    • Search Google Scholar
    • Export Citation
  • Holliday, C. R., , and A. H. Thompson, 1979: Climatological characteristics of rapidly intensifying typhoons. Mon. Wea. Rev., 107, 10221034.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., , and J.-O. Lim, 2006: The WRF single-moment 6-class microphysics scheme. J. Korean Meteor. Soc., 42, 129151.

  • Kain, J. S., 2004: The Kain–Fritsch convective parameterization: An update. J. Appl. Meteor., 43, 170181.

  • Kaplan, J., , and M. DeMaria, 2003: Large-scale characteristics of rapidly intensifying tropical cyclones in the North Atlantic basin. Wea. Forecasting, 18, 10931108.

    • Search Google Scholar
    • Export Citation
  • Kelley, O. A., , J. Stout, , and J. B. Halverson, 2005: Hurricane intensification detected by continuously monitoring tall precipitation in the eyewall. Geophys. Res. Lett., 32, L20819, doi:10.1029/2005GL023583.

    • Search Google Scholar
    • Export Citation
  • Kepert, J. D., , and Y. Wang, 2001: The dynamics of boundary layer jets within the tropical cyclone core. Part II: Nonlinear enhancement. J. Atmos. Sci., 58, 24852501.

    • Search Google Scholar
    • Export Citation
  • Kossin, J. P., , and M. Sitkowski, 2009: An objective model for identifying secondary eyewall formation in hurricanes. Mon. Wea. Rev., 137, 876892.

    • Search Google Scholar
    • Export Citation
  • Kurihara, Y., , M. A. Bender, , and R. J. Ross, 1993: An initialization scheme of hurricane models by vortex specification. Mon. Wea. Rev., 121, 20302045.

    • Search Google Scholar
    • Export Citation
  • Li, Q.-Q., , and Y. Wang, 2012: A comparison of inner and outer spiral rainbands in a numerically simulated tropical cyclone. Mon. Wea. Rev., 140, 27822805.

    • Search Google Scholar
    • Export Citation
  • Lin, I.-I., , C.-C. Wu, , I. F. Pun, , and D. S. Ko, 2008: Upper-ocean thermal structure and the western North Pacific category 5 typhoons. Part I: Ocean features and the category 5 typhoons’ intensification. Mon. Wea. Rev., 136, 32883306.

    • Search Google Scholar
    • Export Citation
  • McFarquhar, G. M., , B. F. Jewett, , M. S. Gilmore, , S. W. Nesbitt, , and T.-L. Hsieh, 2012: Vertical velocity and microphysical distributions related to rapid intensification in a simulation of Hurricane Dennis (2005). J. Atmos. Sci., 69, 35153534.

    • Search Google Scholar
    • Export Citation
  • Miyamoto, Y., , and T. Takemi, 2013: A transition mechanism for the spontaneous axisymmetric intensification of tropical cyclones. J. Atmos. Sci., 70, 112129.

    • Search Google Scholar
    • Export Citation
  • Mlawer, E. J., , S. J. Taubman, , P. D. Brown, , M. J. Iacono, , and S. A. Clough, 1997: Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102D, 16 66316 682.

    • Search Google Scholar
    • Export Citation
  • Molinari, J., , S. Skubis, , and D. Vollaro, 1995: External influences on hurricane intensity. Part III: Potential vorticity structure. J. Atmos. Sci., 52, 35933606.

    • Search Google Scholar
    • Export Citation
  • Montgomery, M. T., , M. E. Nicholls, , T. A. Cram, , and A. B. Saunders, 2006: A vortical hot tower route to tropical cyclogenesis. J. Atmos. Sci., 63, 355386.

    • Search Google Scholar
    • Export Citation
  • Moon, I.-J., , I. Ginis, , T. Hara, , and B. Thomas, 2007: A physics-based parameterization of air–sea momentum flux at high wind speeds and its impact on hurricane intensity predictions. Mon. Wea. Rev., 135, 28692878.

    • Search Google Scholar
    • Export Citation
  • Nakanishi, M., , and H. Niino, 2004: An improved Mellor–Yamada level-3 model with condensation physics: Its design and verification. Bound.-Layer Meteor., 112, 131.

    • Search Google Scholar
    • Export Citation
  • Nguyen, H. V., , and Y.-L. Chen, 2011: High-resolution initialization and simulations of Typhoon Morakot (2009). Mon. Wea. Rev., 139, 14631491.

    • Search Google Scholar
    • Export Citation
  • Ooyama, K. V., 1982: Conceptual evolution of the theory and modeling of the tropical cyclone. J. Meteor. Soc. Japan, 60, 369379.

  • Park, M.-S., , R. L. Elsberry, , and P. A. Harr, 2013: Vertical wind shear and ocean heat contents as environmental modulators of western North Pacific tropical cyclone intensification and decay. Trop. Cyclone Res. Rev., 1, 448457.

    • Search Google Scholar
    • Export Citation
  • Persing, J., , and M. T. Montgomery, 2003: Hurricane superintensity. J. Atmos. Sci., 60, 23492371.

  • Reasor, P. D., , M. D. Eastin, , and J. F. Gamache, 2009: Rapidly intensifying Hurricane Guillermo (1997). Part I: Low-wavenumber structure and evolution. Mon. Wea. Rev., 137, 603631.

    • Search Google Scholar
    • Export Citation
  • Rogers, R., 2010: Convective-scale structure and evolution during a high-resolution simulation of tropical cyclone rapid intensification. J. Atmos. Sci., 67, 4470.

    • Search Google Scholar
    • Export Citation
  • Rozoff, C. M., , W. H. Schubert, , D. Brian, , and J. Kossin, 2006: Rapid filamentation zones in intense tropical cyclones. J. Atmos. Sci., 63, 325340.

    • Search Google Scholar
    • Export Citation
  • Schubert, W. H., , and J. J. Hack, 1982: Inertial stability and tropical cyclone development. J. Atmos. Sci., 39, 16871697.

  • Shapiro, L. J., , and H. E. Willoughby, 1982: The response of balanced hurricanes to local sources of heat and momentum. J. Atmos. Sci., 39, 378394.

    • Search Google Scholar
    • Export Citation
  • Shutts, G. J., 1990: SCAPE charts from numerical weather prediction model fields. Mon. Wea. Rev., 118, 27462751.

  • Skamarock, W. C., , J. B. Klemp, , J. Dudhia, , D. O. Gill, , D. M. Barker, , W. Wang, , and J. G. Powers, 2008: A description of the Advanced Research WRF version 3. NCAR Tech Note-475+STR, 113 pp.

  • Smith, R. K., , M. T. Montgomery, , and V. S. Nguyen, 2009: Tropical cyclone spin-up revisited. Quart. J. Roy. Meteor. Soc., 135, 13211335.

    • Search Google Scholar
    • Export Citation
  • Steiner, M., , R. A. Houze Jr., , and S. E. Yuter, 1995: Climatological characterization of three-dimensional storm structure from operational radar and rain gauge data. J. Appl. Meteor., 34, 19782007.

    • Search Google Scholar
    • Export Citation
  • Vigh, J. L., , and W. H. Schubert, 2009: Rapid development of the tropical cyclone warm core. J. Atmos. Sci., 66, 33353350.

  • Wang, H., , Y. Wang, , and H.-M. Xu, 2013: Improving simulation of a tropical cyclone using dynamical initialization and large-scale spectral nudging: A case study of Typhoon Megi (2010). Acta Meteor. Sin., 27, 455–475.

    • Search Google Scholar
    • Export Citation
  • Wang, Y., 2008a: Rapid filamentation zone in a numerically simulated tropical cyclone. J. Atmos. Sci., 65, 11581181.

  • Wang, Y., 2008b: Structure and formation of an annular hurricane simulated in a fully-compressible, nonhydrostatic model—TCM4. J. Atmos. Sci., 65, 15051527.

    • Search Google Scholar
    • Export Citation
  • Wang, Y., 2009: How do outer spiral rainbands affect tropical cyclone structure and intensity? J. Atmos. Sci., 66, 12501273.

  • Wang, Y., , and J. Xu, 2010: Energy production, frictional dissipation, and maximum intensity of a numerically simulated tropical cyclone. J. Atmos. Sci., 67, 97116.

    • Search Google Scholar
    • Export Citation
  • Wang, Y., , and H. Wang, 2013: The inner-core size increase of Typhoon Megi (2010) during its rapid intensification phase. Trop. Cyclone Res. Rev., 2, 6580.

    • Search Google Scholar
    • Export Citation
  • Wu, C.-C., , K.-H. Chou, , H.-J. Cheng, , and Y. Wang, 2003: Eyewall contraction, breakdown and reformation in a landfalling typhoon. Geophys. Res. Lett., 30, 1887, doi:10.1029/2003GL017653.

    • Search Google Scholar
    • Export Citation
  • Wu, C.-C., , H.-J. Cheng, , Y. Wang, , and K.-H. Chou, 2009: A numerical investigation of the eyewall evolution of a landfalling typhoon. Mon. Wea. Rev., 137, 2140.

    • Search Google Scholar
    • Export Citation
  • Xu, J., , and Y. Wang, 2010a: Sensitivity of the simulated tropical cyclone inner-core size to the initial vortex size. Mon. Wea. Rev., 138, 41354157.

    • Search Google Scholar
    • Export Citation
  • Xu, J., , and Y. Wang, 2010b: Sensitivity of tropical cyclone inner-core size and intensity to the radial distribution of surface entropy flux. J. Atmos. Sci., 67, 18311852.

    • Search Google Scholar
    • Export Citation
  • Yuter, S. E., , and R. A. Houze Jr., 1995: Three-dimensional kinematic and microphysical evolution of Florida cumulonimbus. Part II: Frequency distributions of vertical velocity, reflectivity, and differential reflectivity. Mon. Wea. Rev., 123, 19411963.

    • Search Google Scholar
    • Export Citation
  • Zhang, D.-L., , and H. Chen, 2012: Importance of the upper-level warm core in the rapid intensification of a tropical cyclone. Geophys. Res. Lett., 39, L02806, doi:10.1029/2011GL050578.

    • Search Google Scholar
    • Export Citation
1

RI was defined as the deepening rate of greater than 42 hPa day−1 in the central SLP by Holliday and Thompson (1979) for western Pacific TCs and as 15 m s−1 day−1 in the maximum 10-m wind speed by Kaplan and DeMaria (2003) for Atlantic TCs.

2

Inertial stability is defined as for an axisymmetric vortex, where f is the Coriolis parameter at the storm center, is the azimuthal mean tangential wind, and r is radius. Here, we show the normalized inertial stability, which is defined as I2/f2.

3

There are different definitions of CBs in the literature: 1) Rogers (2010) defined CBs as the vertical velocity averaged between 700 and 300 hPa exceeding 5 m s−1; 2) Reasor et al. (2009) defined CBs as the layer-averaged vertical velocity between 2- and 6-km altitudes larger than 5 m s−1; 3) Montgomery et al. (2006) defined CBs as the vertical velocity through the deep layer between 1 and 15 km, all greater than 1 m s−1; 4) Chen and Zhang (2013) defined CBs as updrafts of at least 15 m s−1 in the upper troposphere (at 11-km height); and 5) Kelley et al. (2005) defined CBs as the extremely tall convective cells with 20 dBZ at least up to 14.5 km. In numerical simulations, the effect of model resolution needs to be considered when CBs are defined. Here, with a 2-km horizontal resolution, we found that 7.5 m s−1 or higher in the upper troposphere (or the maximum updraft velocity in the layer between 2 and 12 km greater than 7.5 m s−1) can be a good indicator of deep convective cells.

Save