• Bister, M., , and Emanuel K. A. , 1997: The genesis of Hurricane Guillermo: TEXMEX analyses and a modeling study. Mon. Wea. Rev., 125, 26622682, doi:10.1175/1520-0493(1997)125<2662:TGOHGT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bosart, L. F., , and Bartlo J. A. , 1991: Tropical storm formation in a baroclinic environment. Mon. Wea. Rev., 119, 19792013, doi:10.1175/1520-0493(1991)119<1979:TSFIAB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bosart, L. F., , Bracken W. E. , , Molinari J. , , Velden C. S. , , and Black P. G. , 2000: Environmental influences on the rapid intensification of Hurricane Opal (1995) over the Gulf of Mexico. Mon. Wea. Rev., 128, 322352, doi:10.1175/1520-0493(2000)128<0322:EIOTRI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bracken, W. E., , and Bosart L. F. , 2000: The role of synoptic-scale flow during tropical cyclogenesis over the North Atlantic Ocean. Mon. Wea. Rev., 128, 353376, doi:10.1175/1520-0493(2000)128<0353:TROSSF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Braun, S. A., , Montgomery M. T. , , Mallen K. J. , , and Reasor P. D. , 2010: Simulation and interpretation of the genesis of Tropical Storm Gert (2005) as part of the NASA Tropical Cloud Systems and Processes Experiment. J. Atmos. Sci., 67, 9991025, doi:10.1175/2009JAS3140.1.

    • Search Google Scholar
    • Export Citation
  • Briegel, L. M., , and Frank W. M. , 1997: Large-scale influences on tropical cyclogenesis in the western North Pacific. Mon. Wea. Rev., 125, 13971413, doi:10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Davis, C. A., , and Bosart L. F. , 2001: Numerical simulations of the genesis of Hurricane Diana (1984). Part I: Control simulation. Mon. Wea. Rev., 129, 18591881, doi:10.1175/1520-0493(2001)129<1859:NSOTGO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Davis, C. A., , and Bosart L. F. , 2002: Numerical simulations of the genesis of Hurricane Diana (1984). Part II: Sensitivity of track and intensity prediction. Mon. Wea. Rev., 130, 11001124, doi:10.1175/1520-0493(2002)130<1100:NSOTGO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Davis, C. A., , and Bosart L. F. , 2003: Baroclinically induced tropical cyclogenesis. Mon. Wea. Rev., 131, 27302747, doi:10.1175/1520-0493(2003)131<2730:BITC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Davis, C. A., , and Ahijevych D. A. , 2013: Thermodynamic environments of deep convection in Atlantic tropical disturbances. J. Atmos. Sci., 70, 19121928, doi:10.1175/JAS-D-12-0278.1.

    • Search Google Scholar
    • Export Citation
  • DeMaria, M., , and Pickle J. D. , 1988: A simplified system of equations for simulation of tropical cyclones. J. Atmos. Sci., 45, 15421554, doi:10.1175/1520-0469(1988)045<1542:ASSOEF>2.0.CO;2.

    • 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, doi:10.1175/1520-0469(1989)046<3077:NSOCOD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dunkerton, T. J., , Montgomery M. T. , , and Wang Z. , 2009: Tropical cyclogenesis in a tropical wave critical layer: Easterly waves. Atmos. Chem. Phys., 9, 55875646, doi:10.5194/acp-9-5587-2009.

    • Search Google Scholar
    • Export Citation
  • Frank, W. M., , and Ritchie E. A. , 2001: Effects of vertical wind shear on the intensity and structure of numerically simulated hurricanes. Mon. Wea. Rev., 129, 22492269, doi:10.1175/1520-0493(2001)129<2249:EOVWSO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Gamache, J. F., , Dodge P. P. , , and Griffin N. F. , 2008: Automatic quality control and analysis of airborne Doppler data: Real-time applications, and automatically post-processed analyses for research. 28th Conf. on Hurricanes and Tropical Meteorology, Orlando, FL, Amer. Meteor. Soc., P2B.12. [Available online at http://ams.confex.com/ams/pdfpapers/137969.pdf.]

  • Gao, J., , Xue M. , , Shapiro A. , , Xu Q. , , and Droegemeier K. K. , 2001: Three-dimensional simple adjoint velocity retrievals from single-Doppler radar. J. Atmos. Oceanic Technol., 18, 2638, doi:10.1175/1520-0426(2001)018<0026:TDSAVR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Gray, W. M., 1968: Global view on the origins of tropical cyclones. Mon. Wea. Rev., 96, 669700, doi:10.1175/1520-0493(1968)096<0669:GVOTOO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hennon, C. C., , and Hobgood J. S. , 2003: Forecasting tropical cyclogenesis over the Atlantic basin using large-scale data. Mon. Wea. Rev., 131, 29272940, doi:10.1175/1520-0493(2003)131<2927:FTCOTA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., , and Lim J.-O. J. , 2006: The WRF single-moment 6-class microphysics scheme (WSM6). J. Korean Meteor. Soc., 42, 129151.

  • Hong, S.-Y., , Noh Y. , , and Dudhia J. , 2006: A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Wea. Rev., 134, 23182341, doi:10.1175/MWR3199.1.

    • Search Google Scholar
    • Export Citation
  • Jin, Y., , Peng M. S. , , and Jin H. , 2008: Simulating the formation of Hurricane Katrina (2005). Geophys. Res. Lett., 35, L11802, doi:10.1029/2008GL033168.

    • Search Google Scholar
    • Export Citation
  • Jorgensen, D. P., , LeMone M. A. , , and Trier S. B. , 1997: Structure and evolution of the 22 February 1993 TOGA COARE squall line: Aircraft observations of precipitation, circulation, and surface energy fluxes. J. Atmos. Sci., 54, 19611985, doi:10.1175/1520-0469(1997)054<1961:SAEOTF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kain, J. S., , and Fritsch J. M. , 1993: Convective parameterization for mesoscale models: The Kain–Fritsch scheme. The Representation of Cumulus Convection in Numerical Models, Meteor. Monogr., No. 46, Amer. Meteor. Soc., 165–170.

  • Kerns, B., , Greene K. , , and Zipser E. , 2008: Four years of tropical ERA-40 vorticity maxima tracks. Part I: Climatology and vertical vorticity structure. Mon. Wea. Rev., 136, 43014319, doi:10.1175/2008MWR2390.1.

    • Search Google Scholar
    • Export Citation
  • Kieu, C. Q., , and Zhang D.-L. , 2010: Genesis of Tropical Storm Eugene (2005) from merging vortices associated with ITCZ breakdowns. Part III: Sensitivity to various genesis parameters. J. Atmos. Sci., 67, 17451758, doi:10.1175/2010JAS3227.1.

    • Search Google Scholar
    • Export Citation
  • Knaff, J. A., , Seseske S. A. , , DeMaria M. , , and Demuth J. L. , 2004: On the influences of vertical wind shear on symmetric tropical cyclone structure derived from AMSU. Mon. Wea. Rev., 132, 25032510, doi:10.1175/1520-0493(2004)132<2503:OTIOVW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Komaromi, W., 2013: An investigation of composite dropsonde profiles for developing and nondeveloping tropical waves during the 2010 PREDICT field campaign. J. Atmos. Sci., 70, 542558, doi:10.1175/JAS-D-12-052.1.

    • Search Google Scholar
    • Export Citation
  • Li, Z., 2013: Studying the genesis of Typhoon Nuri (2008) with numerical simulations and data assimilation. Ph.D. dissertation, University of Utah, 164 pp. [Available from Dept. of Atmospheric Sciences, University of Utah, Salt Lake City, UT 84112.]

  • McBride, J. L., , and Zehr R. , 1981: Observational analysis of tropical cyclone formation. Part II: Comparison of nondeveloping versus developing systems. J. Atmos. Sci., 38, 11321151, doi:10.1175/1520-0469(1981)038<1132:OAOTCF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Merrill, R. T., , and Velden C. S. , 1996: A three-dimensional analysis of the outflow layer of Supertyphoon Flo (1990). Mon. Wea. Rev., 124, 4763, doi:10.1175/1520-0493(1996)124<0047:ATDAOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Mlawer, E. J., , Taubman S. J. , , Brown P. D. , , Iacono M. J. , , and Clough S. A. , 1997: Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102, 16 66316 682, doi:10.1029/97JD00237.

    • Search Google Scholar
    • Export Citation
  • Möller, J. D., , and Montgomery M. T. , 1999: Vortex Rossby waves and hurricane intensification in a barotropic model. J. Atmos. Sci., 56, 16741687, doi:10.1175/1520-0469(1999)056<1674:VRWAHI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Montgomery, M. T., , and Enagonio J. , 1998: Tropical cyclogenesis via convectively forced vortex Rossby waves in a three-dimensional quasigeostrophic model. J. Atmos. Sci., 55, 31763207, doi:10.1175/1520-0469(1998)055<3176:TCVCFV>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Montgomery, M. T., , Nicholls M. E. , , Cram T. A. , , and Saunders A. B. , 2006: A vortical hot tower route to tropical cyclogenesis. J. Atmos. Sci., 63, 355386, doi:10.1175/JAS3604.1.

    • Search Google Scholar
    • Export Citation
  • Montgomery, M. T., , Lussier L. III, , Moore R. W. , , and Wang Z. , 2010: The genesis of Typhoon Nuri as observed during the Tropical Cyclone Structure 2008 (TCS-08) field experiment—Part 1: The role of the easterly wave critical layer. Atmos. Chem. Phys., 10, 98799900, doi:10.5194/acp-10-9879-2010.

    • Search Google Scholar
    • Export Citation
  • Musgrave, K. D., , Davis C. A. , , and Montgomery M. T. , 2008: Numerical simulations of the formation of Hurricane Gabrielle (2001). Mon. Wea. Rev., 136, 31513167, doi:10.1175/2007MWR2110.1.

    • Search Google Scholar
    • Export Citation
  • Oye, R., , Mueller C. , , and Smith S. , 1995: Software for translation, visualization, editing and interpolation. Preprints, 27th Conf. on Radar Meteorology, Vail, CO, Amer. Meteor. Soc., 359–361.

  • Peng, M. S., , Fu B. , , Li T. , , and Stevens D. E. , 2012: Developing versus nondeveloping disturbances for tropical cyclone formation. Part I: North Atlantic. Mon. Wea. Rev., 140, 10471066, doi:10.1175/2011MWR3617.1.

    • Search Google Scholar
    • Export Citation
  • Pu, Z., , and Zhang L. , 2010: Validation of AIRS temperature and moisture profiles over tropical oceans and their impact on numerical simulations of tropical cyclones. J. Geophys. Res., 115, D24114, doi:10.1029/2010JD014258.

    • Search Google Scholar
    • Export Citation
  • Ramsay, H. A., , and Sobel A. H. , 2011: Effects of relative and absolute sea surface temperature on tropical cyclone potential intensity using a single-column model. J. Climate, 24, 183193, doi:10.1175/2010JCLI3690.1.

    • Search Google Scholar
    • Export Citation
  • Rappin, E. D., , and Nolan D. S. , 2012: The effect of vertical shear orientation on tropical cyclogenesis. Quart. J. Roy. Meteor. Soc., 138, 10351054, doi:10.1002/qj.977.

    • Search Google Scholar
    • Export Citation
  • Rappin, E. D., , Nolan D. S. , , and Emanuel K. A. , 2010: Thermodynamic control of tropical cyclogenesis in environments of radiative–convective equilibrium with shear. Quart. J. Roy. Meteor. Soc., 136, 19541971, doi:10.1002/qj.706.

    • Search Google Scholar
    • Export Citation
  • Raymond, D. J., , and Sessions S. L. , 2007: Evolution of convection during tropical cyclogenesis. Geophys. Res. Lett., 34, L06811, doi:10.1029/2006GL028607.

    • Search Google Scholar
    • Export Citation
  • Raymond, D. J., , Sessions S. L. , , and López Carrillo C. L. , 2011: Thermodynamics of tropical cyclogenesis in the northwest Pacific. J. Geophys. Res., 116, D18101, doi:10.1029/2011JD015624.

    • Search Google Scholar
    • Export Citation
  • Reed, R. J., , Norquist D. C. , , and Recker E. E. , 1977: The structure and properties of African wave disturbances as observed during phase III of GATE. Mon. Wea. Rev., 105, 317333, doi:10.1175/1520-0493(1977)105<0317:TSAPOA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Ritchie, E. A., , and Holland G. J. , 1997: Scale interactions during the formation of Typhoon Irving. Mon. Wea. Rev., 125, 13771396, doi:10.1175/1520-0493(1997)125<1377:SIDTFO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Rodgers, E., , Olson W. , , Halverson J. , , Simpson J. , , and Pierce H. , 2000: Environmental forcing of Supertyphoon Paka’s (1997) latent heat structure. J. Appl. Meteor., 39, 19832006, doi:10.1175/1520-0450(2001)040<1983:EFOSPS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Shen, B.-W., , Tao W.-K. , , Lau W. K. , , and Atlas R. , 2010: Predicting tropical cyclogenesis with a global mesoscale model: Hierarchical multiscale interactions during the formation of Tropical Cyclone Nargis (2008). J. Geophys. Res., 115, D14102, doi:10.1029/2009JD013140.

    • Search Google Scholar
    • Export Citation
  • Skamarock, W. C., and et al. , 2008: A description of the Advanced Research WRF version 3. Tech. Note NCAR/TN-475+STR, 113 pp. [Available online at http://www2.mmm.ucar.edu/wrf/users/docs/arw_v3.pdf.]

  • Thatcher, L., , and Pu Z. , 2013: Evaluation of tropical cyclone genesis precursors with relative operating characteristics (ROC) in high-resolution ensemble forecasts: Hurricane Ernesto. Trop. Cyclone Res. Rev., 2, 131148, doi:10.6057/2013TCRR03.01.

    • Search Google Scholar
    • Export Citation
  • Vecchi, G. A., , and Soden B. J. , 2007: Effect of remote sea surface temperature change on tropical cyclone potential intensity. Nature, 450, 10661070, doi:10.1038/nature06423.

    • Search Google Scholar
    • Export Citation
  • Wang, Z., 2012: Thermodynamic aspects of tropical cyclone formation. J. Atmos. Sci., 69, 24332451, doi:10.1175/JAS-D-11-0298.1.

  • Wang, Z., , Montgomery M. T. , , and Dunkerton T. J. , 2010a: Genesis of pre–Hurricane Felix (2007). Part I: The role of the easterly wave critical layer. J. Atmos. Sci., 67, 17111729, doi:10.1175/2009JAS3420.1.

    • Search Google Scholar
    • Export Citation
  • Wang, Z., , Montgomery M. T. , , and Dunkerton T. J. , 2010b: Genesis of pre-Hurricane Felix (2007). Part II: Warm core formation, precipitation evolution, and predictability. J. Atmos. Sci., 67, 17301744, doi:10.1175/2010JAS3435.1.

    • Search Google Scholar
    • Export Citation
  • Willoughby, H. E., , and Black P. G. , 1996: Hurricane Andrew in Florida: Dynamics of a disaster. Bull. Amer. Meteor. Soc., 77, 543549, doi:10.1175/1520-0477(1996)077<0543:HAIFDO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Zhang, D.-L., , and Zhu L. , 2012: Roles of upper-level processes in tropical cyclogenesis. Geophys. Res. Lett., 39, L17804, doi:10.1029/2012GL053140.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Locations of model domains for WRF simulations. Domains D01 and D02 are used for experiments ERA and FNL, and D03 is used for sensitivity experiments (see Table 1 for details).

  • View in gallery

    Time series of the minimum center sea level pressure (hPa) from numerical simulations (ERA and FNL) compared with JTWC best-track data.

  • View in gallery

    Nuri’s track (every 6 h) between 0000 UTC 16 Aug and 0000 UTC 18 Aug 2008 from numerical simulations (ERA and FNL), compared with the best-track data.

  • View in gallery

    Horizontal distribution of hourly precipitation for (a) FNL at 23-h simulation (1700 UTC 16 Aug 2008), (b) ERA at 23-h simulation (1700 UTC 16 Aug 2008), (c) AMSR-E-derived rain rate at 1619 UTC 16 Aug 2008, (d) FNL at 53-h simulation (2300 UTC 17 Aug 2008), (e) ERA at 53-h simulation (2300 UTC 17 Aug 2008), (f) SSM/IS-derived rain rate at 2240 UTC 17 Aug 2008, (g) FNL at 63-h simulation (0900 UTC 18 Aug 2008), (h) ERA at 63-h simulation (0900 UTC 18 Aug 2008), and (i) WindSat-derived rain rate at 0854 UTC 18 Aug 2008. The contour lines and vectors in (a),(b),(d),(e),(g), and (h) are the simulated sea level pressure (contour interval: 1 hPa) and the horizontal winds at the lowest model level, respectively. The rain-rate plots in (c),(f), and (i) are obtained from the NRL’s TC website (http://www.nrlmry.navy.mil/tc_pages/tc_home.html).

  • View in gallery

    Comparisons of sounding profiles from numerical simulation (FNL, red curves; ERA, blue curves) with the dropsondes (black curves) near the center of Nuri at (a) 0241 UTC 16 Aug 2008 at 11.81°N, 144.98°E and (b) 0047 UTC 17 Aug 2008 at 14.0°N, 140.0°E. The temperature and dewpoint temperature profiles are shown with solid and dashed curves, respectively.

  • View in gallery

    Time series of the minimum center sea level pressure (hPa) from numerical simulations compared with JTWC best-track data. Sensitivity to (a) grid spacings (ERA, ERA_d03, FNL, and FNL_d03) and (b) forecast leading times (ERA_E24h, FNL_E24h, ERA_E24h_d03, and FNL_E24h_d03).

  • View in gallery

    Radar reflectivity (color shading) observed by (a) Guam NEXRAD at 0031 UTC Aug 16 2008 and (b) airborne ELDORA at 1.5 km AGL. The white box in (a) denotes the approximate coverage of (b). The plot in (a) is obtained from data online (http://catalog.eol.ucar.edu/tparc_2008/index.html). The wind vectors in (b) include Doppler radar wind analysis and dropsondes (with black dots at the tails of those vectors).

  • View in gallery

    The 3-h accumulated precipitation (contours) and 850-hPa wind vectors valid at 0100 UTC 16 Aug in the control and sensitivity experiments: (left) ERA and (right) FNL. In (a), “ERA at 7 h” means 7-h forecast in ERA. Others follow the same pattern: (b) 7-h forecast in FNL, (c) 7-h forecast in ERA_d03, (d) 7-h forecast in FNL_d03, (e) 31-h forecast in ERA_E24h, (f) 31-h forecast in FNL_E24h, (g) 31-h forecast in ERA_E24h_d03, and (h) 31-h forecast in FNL_E24h_d03.

  • View in gallery

    Hovmöller plots of the meridional wind speed (m s−1) at 850 hPa along 13°N in (a) ERA and (b) FNL between 1800 UTC 15 Aug and 1800 UTC 16 Aug 2008 (0–24-h forecast). The black solid line indicates the propagation of the trough axes in both simulations.

  • View in gallery

    The convergence (color shading) at 850 hPa and sea level pressure (blue contours; contour interval: 2 hPa) in ERA. The wind vectors present the 850-hPa horizontal winds in the co-moving frame. The purple lines indicate the critical layers (isopleths of zero relative zonal wind). The black lines indicate the wave’s trough axis. The black dots indicate the pouch center. The blue arrows indicate the direction of the convection movement: (a) 2100 UTC 15 Aug, (b) 0000 UTC 16 Aug, (c) 0300 UTC 16 Aug, and (d) 1800 UTC 16 Aug 2008. “ERA 3 h” means 3-h forecast in ERA from 1800 UTC 15 Aug 2008. Others follow the same pattern.

  • View in gallery

    As in Fig. 10, but for relative vorticity (color shading). The two vortices discussed in the text are indicated by dark red arrows and letters (A and B): (a) 0300 UTC 16 Aug, (b) 0500 UTC 16 Aug, (c) 0700 UTC 16 Aug, and (d) 1400 UTC 16 Aug 2008.

  • View in gallery

    As in Fig. 10, but for the FNL.

  • View in gallery

    Vertical cross sections of the initial relative vorticity (contour interval: 1 × 10−5 s−1) and the initial relative humidity (contour interval: 5%) at 1800 UTC 15 Aug 2008: (a),(b) the relative vorticity in FNL and ERA, respectively; (c),(d) the relative humidity in FNL and ERA, respectively.

  • View in gallery

    Comparisons of the saturation fraction (color shading; %) and the horizontal winds at 500 hPa in the co-moving frame between (left) ERA and (right) FNL at (a),(b) 1800 UTC 15 Aug; (c),(d) 0300 UTC 16 Aug; and (e),(f) 2100 UTC 16 Aug 2008.

  • View in gallery

    Comparisons of the relative vorticity (color shading; 10−5 s−1), relative humidity (red contour lines; contour interval: 5%) and geopotential height (blue contours; contour interval: 10 m), and the horizontal winds in the co-moving frame at 500 hPa between (left) ERA and (right) FNL at (a),(b) 1800 UTC 15 Aug; (c),(d) 0300 UTC 16 Aug; and (e),(f) 1800 UTC 16 Aug 2008.

  • View in gallery

    Time series of simulated MSLP (hPa) between 1800 UTC 15 Aug and 1800 UTC 18 Aug 2008 in the sensitivity experiments with swapped initial moisture conditions, compared with JTWC best-track data.

  • View in gallery

    Radius–pressure cross sections of the azimuthally averaged values of temperature perturbations (color shading; shading interval: 0.25 K), calculated as differences between the temperature near the simulated circulation center (0–350 km) and the environmental temperature that is determined by averaging the temperature within a radial band between 500–900 km, and the vertical mass flux (contour interval: 10−5 kg m−2 s−1) in (left) ERA and (right) FNL at (a),(b) 1800 UTC 15 Aug; (c),(d) 0000 UTC 16 Aug; (e),(f) 1800 UTC 16 Aug; and (g),(h) 0000 UTC 17 Aug 2008.

  • View in gallery

    Radius–pressure cross sections of the azimuthally averaged relative humidity (contour interval: 5%) and secondary circulation (represented by uw vectors; u unit: m s−1; w unit: cm s−1; u is the radial velocity; w is the vertical velocity) in (left) ERA and (right) FNL at (a),(b) 1800 UTC 15 Aug; (c),(d) 0000 UTC 16 Aug; (e),(f) 1800 UTC 16 Aug; and (g),(h) 0000 UTC 17 Aug 2008.

  • View in gallery

    Time series of vertical profiles of area-averaged (over the area of 100 km of radius from the center of Nuri) diabatic heating (contour interval: 0.05 K h−1) in (a) ERA and (b) FNL from 1800 UTC 15 Aug to 0000 UTC 17 Aug 2008.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 32 32 9
PDF Downloads 28 28 18

Numerical Simulations of the Genesis of Typhoon Nuri (2008): Sensitivity to Initial Conditions and Implications for the Roles of Intense Convection and Moisture Conditions

View More View Less
  • 1 Department of Atmospheric Sciences, University of Utah, Salt Lake City, Utah
© Get Permissions
Full access

Abstract

The sensitivity of numerical simulations of the genesis of Typhoon Nuri (2008) to initial conditions is examined using the Advanced Research core of the Weather Research and Forecasting (WRF) Model. The initial and boundary conditions are derived from two different global analyses at different lead times. One simulation successfully captures the processes of Nuri’s genesis and early intensification, whereas other simulations fail to predict the genesis of Nuri. Discrepancies between simulations with and without Nuri’s development are diagnosed. Significant differences are found in the development and organization of the intense convection during Nuri’s pregenesis phase. In the developing case, convection evolves and organizes into a “pouch” center of a westward-propagating wavelike disturbance. In the nondeveloping case, the convection fails to develop and organize. Favorable conditions for the development of deep convection include strong closed circulation patterns with high humidity, especially at the middle levels. An additional set of sensitivity experiments is performed to examine the impact of the moisture field on numerical simulations of Nuri’s genesis. Results confirm that the enhancement of mid- to upper-level moisture is favorable for Nuri’s genesis, mainly because moist conditions benefit deep convection, which produces diabatic heating from latent heat release when vertical airmass flux maxima occur in the mid- to upper-level atmosphere. The substantial warming at upper levels induced by latent heat release from persistent deep convection contributes to the drop in Nuri’s minimum central sea level pressure. Overall, results from this study demonstrate that it is essential to accurately represent the initial conditions in numerical predictions of tropical cyclone genesis.

Corresponding author address: Dr. Zhaoxia Pu, Dept. of Atmospheric Sciences, Rm. 819, University of Utah, 135 S 1460 E, Salt Lake City, UT 84112. E-mail: Zhaoxia.Pu@utah.edu

Abstract

The sensitivity of numerical simulations of the genesis of Typhoon Nuri (2008) to initial conditions is examined using the Advanced Research core of the Weather Research and Forecasting (WRF) Model. The initial and boundary conditions are derived from two different global analyses at different lead times. One simulation successfully captures the processes of Nuri’s genesis and early intensification, whereas other simulations fail to predict the genesis of Nuri. Discrepancies between simulations with and without Nuri’s development are diagnosed. Significant differences are found in the development and organization of the intense convection during Nuri’s pregenesis phase. In the developing case, convection evolves and organizes into a “pouch” center of a westward-propagating wavelike disturbance. In the nondeveloping case, the convection fails to develop and organize. Favorable conditions for the development of deep convection include strong closed circulation patterns with high humidity, especially at the middle levels. An additional set of sensitivity experiments is performed to examine the impact of the moisture field on numerical simulations of Nuri’s genesis. Results confirm that the enhancement of mid- to upper-level moisture is favorable for Nuri’s genesis, mainly because moist conditions benefit deep convection, which produces diabatic heating from latent heat release when vertical airmass flux maxima occur in the mid- to upper-level atmosphere. The substantial warming at upper levels induced by latent heat release from persistent deep convection contributes to the drop in Nuri’s minimum central sea level pressure. Overall, results from this study demonstrate that it is essential to accurately represent the initial conditions in numerical predictions of tropical cyclone genesis.

Corresponding author address: Dr. Zhaoxia Pu, Dept. of Atmospheric Sciences, Rm. 819, University of Utah, 135 S 1460 E, Salt Lake City, UT 84112. E-mail: Zhaoxia.Pu@utah.edu

1. Introduction

Identifying necessary conditions for the genesis of tropical cyclones (TCs) is a challenging problem. Despite a lack of conventional observations over the ocean, several factors are recognized as necessary for the genesis of a TC. According to research within the last 40 years, these factors include warm sea surface temperatures (SSTs), weak environmental vertical wind shear, a preexisting synoptic-scale disturbance with high relative humidity, and abundant deep convective clouds and precipitation (latent heat release near the incipient center) (e.g., Gray 1968; McBride and Zehr 1981; DeMaria and Pickle 1988; Merrill and Velden 1996; Willoughby and Black 1996; Bosart et al. 2000; Montgomery et al. 2006).

The SST has been considered to be one of several factors influencing the genesis of TCs. Because SSTs directly affect the amount of water vapor available for latent heat release and storm intensification, Davis and Bosart (2002) found strong sensitivities with varying SSTs in numerical simulations of the genesis of Hurricane Diana in 1984. Peng et al. (2012) studied TC formation over the North Atlantic and identified SST as an important parameter in the box difference index (BDI), which distinguishes developing and nondeveloping disturbances. They also found that SSTs account for the seasonal variation in TC genesis: developing disturbances are more frequent in the late season, coinciding with increasing SSTs. Many studies have suggested a threshold SST value of around 26.5°C (e.g., Gray 1968; Briegel and Frank 1997; Rodgers et al. 2000). Alternatively, Vecchi and Soden (2007) revealed that TC activity is influenced more by the SST relative to the overlying atmospheric thermal conditions than by the absolute value of the SST. Their study, using a single-column model by Ramsay and Sobel (2011), demonstrated that TC potential intensity is more sensitive to relative SST than to absolute SST since relative SST represents the air–sea thermodynamic disequilibrium.

Strong, deep tropospheric vertical wind shear is believed to be detrimental to TC genesis because it can deform the vertical structure of a vortex and displace deep convection away from the low-level center (Musgrave et al. 2008). However, the role of vertical wind shear is currently under debate. For instance, Bracken and Bosart (2000) pointed out that a small amount of vertical wind shear is necessary for TC development. Rappin and Nolan (2012) found that TC genesis is influenced not only by the magnitude of the vertical wind shear but also by its direction relative to the direction of the mean surface wind. The process of TC genesis is enhanced when the mean surface wind and shear are counteraligned.

A favorable preexisting disturbance is also critical to TC genesis. Such precursor disturbances can originate from easterly waves (Reed et al. 1977), and monsoon troughs can provide a source for TC genesis (Gray 1968; Briegel and Frank 1997; Ritchie and Holland 1997). In addition, baroclinic systems provide a source of disturbances for TC genesis (Bosart and Bartlo 1991; Davis and Bosart 2001, 2003). Furthermore, mesoscale convective vortices (MCVs) have been considered precursor disturbances in some studies: TC genesis involves the transformation of MCVs through top–down development into a surface-concentrated vortex (Bister and Emanuel 1997). Another theory has suggested that TCs are generated from an initial MCV via bottom–up development (Montgomery and Enagonio 1998; Möller and Montgomery 1999; Montgomery et al. 2006).

Recently, a new way to present the flow structure of the preexisting disturbance has been adopted (Dunkerton et al. 2009). In a comoving frame of reference with the disturbance propagation speed, a persistently protected region, or “pouch,” with a quasi-closed circulation in the low to middle troposphere, prevents dry air from entering the circulation and preserves the concentrated moisture inside the circulation pattern. In addition to the rich moisture in the pouch, mesoscale convection is important for TC genesis (Montgomery et al. 2006; Wang et al. 2010b; Komaromi 2013), but few studies have demonstrated how the convection interacts with the preexisting disturbance or the pouch. So one of our goals is to reveal the role the convection in the pouch plays in the ensuing TC genesis.

Owing to the complexity of the conditions that contribute to TC genesis, numerical prediction of TC genesis is a challenging problem (e.g., Hennon and Hobgood 2003; Kerns et al. 2008; Thatcher and Pu 2013). Previous studies have shown that initial conditions have substantial impacts on the numerical prediction of tropical cyclone genesis (e.g., Davis and Bosart 2002; Kieu and Zhang 2010; Pu and Zhang 2010). In light of previous findings, we perform numerical simulations of the genesis of Typhoon Nuri (2008) with the mesoscale community Weather Research and Forecasting (WRF) Model using initial and boundary conditions derived from two different global analyses at different lead times. Through analyzing numerical simulations with different initial conditions that result in the development and nondevelopment of Typhoon Nuri, we intend to demonstrate the influence of initial conditions on predicting TC genesis. In addition, by diagnosing the results, we will investigate the major fields controlling the genesis of Typhoon Nuri. Specifically, by comparing the differences in the initial conditions, convective development, and the evolution of circulation structures prior to Nuri’s genesis, we will reveal how the persistent convection in the pouch contributes to the reinforcement of the vortex. In addition, considering the large discrepancies in the initial moisture conditions from two global analyses, sensitivity experiments are performed to investigate the role of the moisture field in Nuri’s genesis. The effect of the moisture field on thermodynamic processes and their impacts on the genesis of Typhoon Nuri are also analyzed. Overall, the essential role of accurate initial conditions in predicting TC geneses will be demonstrated.

This paper is organized as follows. Section 2 describes the observational data and a brief overview of Typhoon Nuri. Section 3 presents the model configuration, simulation validation of the control experiments, and the sensitivity of numerical simulations of Nuri’s genesis to model horizontal grid size and forecast lead times. Section 4 interprets the numerical simulation results. The roles of intense convection, pouch circulation, and the moisture field, as well as SSTs and vertical wind shear, are discussed. The evolution of thermodynamic and dynamic processes associated with Nuri’s genesis is also diagnosed. A summary and concluding remarks are given in section 5.

2. Descriptions of Typhoon Nuri (2008) and observations

a. A brief overview of Typhoon Nuri

Similar to many other TCs that develop in the western North Pacific, Typhoon Nuri was initiated in the form of an easterly wave over the ocean with high SSTs (above 26°C). The Joint Typhoon Warning Center (JTWC) began tracking the pre-Nuri disturbance at 0000 UTC 16 August 2008 with an initial location of 13.2°N, 146.8°E. At that time, a low-level circulation center formed to the east of Guam as the easterly wave increased its low-level vorticity. With a well-organized convective system and low vertical wind shear, the system reached tropical depression intensity with a maximum surface wind speed (MSW) of 12.8 m s−1 and a minimum sea level pressure (MSLP) of 1004 hPa at 1800 UTC 16 August. Then it moved farther westward and formed a tropical storm with an MSW of 18.0 m s−1 and an MSLP of 996 hPa at 1200 UTC 17 August 2008. Twenty-four hours later (1200 UTC 18 August), it developed rapidly, achieving typhoon strength and acquiring the name Nuri. The central MSLP reached 974 hPa.

In this study, the genesis time of Nuri is defined as the time at which the JTWC designated it as a tropical depression, namely, at 1800 UTC 16 August. Numerical simulations are conducted between 1800 UTC 15 August and 1800 UTC 18 August 2008, which covers the period of Nuri’s genesis, early intensification, and development. The analysis of the simulation results emphasizes Nuri’s pregenesis and genesis, namely, between 1800 UTC 15 August and 1800 UTC 16 August 2008 and slightly beyond, with the goal of understanding the processes that influence the genesis of Nuri.

b. Observations

The JTWC best-track data are used to verify the intensity of the simulated Typhoon Nuri. Rainfall rates derived from a host of instruments on multiple satellites [e.g., Advanced Microwave Scanning Radiometer for Earth Observing System (EOS) (AMSR-E)/Aqua, Special Sensor Microwave Imager/Sounder (SSM/IS), WindSat] are used to validate the simulated rainband structures. Next Generation Weather Radar (NEXRAD) reflectivity is employed to compare the convective systems in the pre-Nuri disturbance. The temperature and dewpoint temperature profiles from dropsondes obtained by U.S. Air Force (USAF) C-130J aircraft during The Observing System Research and Predictability Experiment (THORPEX) Pacific Asian Regional Campaign (TPARC) and Office of Naval Research’s (ONR) Tropical Cyclone Structure (TCS-08) field experiments between August and October 2008 are useful for validating the model’s simulated results.

During TPARC–TCS-08, the U.S. Naval Research Laboratory’s (NRL) P-3 aircraft carried the National Center for Atmospheric Research’s (NCAR) Electra Doppler Radar (ELDORA) to obtain the radar reflectivity and radial velocity. The research flight was operated during the pre-Nuri disturbance from 2300 UTC 15 August to 0350 UTC 16 August near Guam (13.45°N, 144.78°E). A quality control (Oye et al. 1995) process and a three-dimensional wind analysis scheme (Gao et al. 2001; Gamache et al. 2008) were applied for ELDORA radial velocity to obtain the three-dimensional wind analyses. These analyses are used to present the observed wind structures during the pre-Nuri disturbance.

3. Numerical simulations

a. Control simulations and validation

1) Simulation description

An Advanced Research version of the WRF Model, version 3.2.1 (Skamarock et al. 2008), is employed in this study to simulate the genesis of Nuri. A two-way interactive nested-domain configuration is used. A total of 40 terrain-following σ levels in the vertical coordinate are used, with the top of the model at 5 hPa. Physics schemes used for both domains include a WRF single-moment six-class (WSM6) cloud microphysics scheme (Hong and Lim 2006), the Rapid Radiative Transfer Model (RRTM) for longwave radiation with six molecular species (Mlawer et al. 1997), the Dudhia shortwave radiation scheme (Dudhia 1989), a modified version of the Kain–Fritsch cumulus parameterization scheme (Kain and Fritsch 1993), and the Yonsei University (YSU) planetary boundary layer parameterization with the Monin–Obukhov surface layer scheme (Hong et al. 2006). These physical parameterization schemes, especially the cumulus, PBL, and microphysics, are chosen because they produce the best possible simulation based on the early sensitivity studies.

Using two-level nested domains (D01 with grid spacing of 36 km and D02 with grid spacing of 12 km), we first conduct two sets of numerical simulations, with two different sets of initial conditions: the FNL experiment uses initial and lateral boundary conditions derived from the National Centers for Environmental Prediction (NCEP) Global Forecast System (GFS) Final (FNL) analysis at 1° × 1° horizontal resolution; the ERA experiment obtains initial and lateral boundary conditions from the Interim European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-Interim; also referred to as ERA) data product at 0.75° × 0.75° horizontal resolution. Both simulations start at 1800 UTC 15 August 2008. FNL and ERA are set up as the control experiments. All forecasts extend to 1800 UTC 18 August 2008. The basic model configurations for the control experiments (FNL and ERA) are summarized in Table 1. Figure 1 illustrates the positions and sizes of the model domains (D01 and D02 for ERA and FNL).

Table 1.

Model configurations for numerical simulation experiments.

Table 1.
Fig. 1.
Fig. 1.

Locations of model domains for WRF simulations. Domains D01 and D02 are used for experiments ERA and FNL, and D03 is used for sensitivity experiments (see Table 1 for details).

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

First, 72-h integrations are performed. As shown by the time series of the MSLP in Fig. 2a, it is apparent that FNL does not predict the genesis of Nuri. The simulated MSLP is higher than 1005 hPa throughout the entire simulation. Significant improvements in the simulated storm intensity are obtained from ERA. ERA successfully predicts Nuri’s genesis, with the simulated MSLP of Nuri in ERA almost identical to the best-track data.

Fig. 2.
Fig. 2.

Time series of the minimum center sea level pressure (hPa) from numerical simulations (ERA and FNL) compared with JTWC best-track data.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

2) Validation of control simulations

Figure 3 compares Nuri’s track from the two numerical simulations with the JTWC best-track data. Overall, Nuri moves west-northwest. Although the initial locations of the simulated Nuri in both experiments are almost the same at 0000 UTC 16 August, the disturbance moves more slowly in FNL and the track is farther north compared with ERA and the best-track data. Simulated tracks from both experiments show a northeast bias during most of the simulation. Nevertheless, the track errors in ERA are much smaller than those in FNL.

Fig. 3.
Fig. 3.

Nuri’s track (every 6 h) between 0000 UTC 16 Aug and 0000 UTC 18 Aug 2008 from numerical simulations (ERA and FNL), compared with the best-track data.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

Simulated rain rates from both experiments are compared with the satellite-derived rain rates (Fig. 4). Near the time of Nuri’s genesis (1619 UTC 16 August), the National Aeronautics and Space Administration (NASA) AMSR-E satellite-derived rainfall data show modest precipitation around the circulation center. ERA captures a similar pattern of precipitation near this time. FNL, however, produces much weaker and more scattered precipitation, along with weaker circulation in terms of wind speed and sea level pressure. During Nuri’s development, ERA produces a much stronger circulation pattern and heavier precipitation. At both 53 and 63 h, maximum rain rates are located on the southern side of the circulation center in ERA (Figs. 4e,h), similar to the locations of maximum precipitation in the SSM/IS and WindSat satellite-derived rainfall rates (Figs. 4f,i). Furthermore, ERA shows TC rainbands similar to those in the satellite-derived data. In contrast, FNL completely misses both the intensities and patterns of the satellite-derived precipitation. Overall, compared with FNL, ERA produces a better simulation in terms of precipitation intensity and distribution.

Fig. 4.
Fig. 4.

Horizontal distribution of hourly precipitation for (a) FNL at 23-h simulation (1700 UTC 16 Aug 2008), (b) ERA at 23-h simulation (1700 UTC 16 Aug 2008), (c) AMSR-E-derived rain rate at 1619 UTC 16 Aug 2008, (d) FNL at 53-h simulation (2300 UTC 17 Aug 2008), (e) ERA at 53-h simulation (2300 UTC 17 Aug 2008), (f) SSM/IS-derived rain rate at 2240 UTC 17 Aug 2008, (g) FNL at 63-h simulation (0900 UTC 18 Aug 2008), (h) ERA at 63-h simulation (0900 UTC 18 Aug 2008), and (i) WindSat-derived rain rate at 0854 UTC 18 Aug 2008. The contour lines and vectors in (a),(b),(d),(e),(g), and (h) are the simulated sea level pressure (contour interval: 1 hPa) and the horizontal winds at the lowest model level, respectively. The rain-rate plots in (c),(f), and (i) are obtained from the NRL’s TC website (http://www.nrlmry.navy.mil/tc_pages/tc_home.html).

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

To further validate the vertical profiles of thermal conditions in the simulations, Fig. 5 compares the model-simulated soundings with dropsondes obtained from the C-130J aircraft during TPARC–TCS-08. At 8 h (0200 UTC 16 August; Fig. 5a), in Nuri’s pregenesis phase, the simulated temperature profile from ERA agrees with the dropsonde observations. Specifically, ERA captures a moist layer around 950 hPa that is indicated by the dropsondes. Meanwhile, the FNL soundings are generally less compatible with the observed soundings. Figure 5b compares the dropsondes with the simulated soundings at 31 h (0100 UTC 17 August), several hours after Nuri’s genesis. Compared with the dropsondes, ERA predicts realistic temperature and dewpoint temperature profiles, whereas FNL produces much warmer and drier biases compared with the dropsondes in most of the troposphere. ERA produces much better forecasts both before and after Nuri’s genesis.

Fig. 5.
Fig. 5.

Comparisons of sounding profiles from numerical simulation (FNL, red curves; ERA, blue curves) with the dropsondes (black curves) near the center of Nuri at (a) 0241 UTC 16 Aug 2008 at 11.81°N, 144.98°E and (b) 0047 UTC 17 Aug 2008 at 14.0°N, 140.0°E. The temperature and dewpoint temperature profiles are shown with solid and dashed curves, respectively.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

The above results clearly demonstrate that ERA produces a reasonable simulation of Typhoon Nuri’s genesis, whereas the FNL simulation does not. Explaining the reasons why these two analyses differ in this particular case is beyond the scope of this study. Meanwhile, the simulation results cannot be used to draw general conclusions regarding the relative quality of the two global analyses, as the selection of a single case study is completely arbitrary. However, based on these two sets of numerical experiments, we should ask the following fundamental questions: 1) What are the major differences between the two sets of initial conditions that cause the discrepancies in the two simulations? 2) Would numerical simulation results change if we used different model resolutions and initial times (viz., forecast lead times)? More importantly, 3) what are the implications of the simulated results for Typhoon Nuri’s genesis? In the following sections, additional numerical experiments are performed and results are analyzed in detail in order to address these questions.

b. Sensitivity experiments

Sensitivity experiments are first conducted to explore the impact of model grid spacing and initialization time (i.e., the lead time relative to the time of TC genesis) on simulating Nuri’s genesis.

1) Grid spacing

It is commonly believed that high-resolution simulation can lead to better numerical forecasts of TCs (e.g., Jin et al. 2008; Shen et al. 2010). Two sets of experiments are first conducted to examine the sensitivity of numerical simulations of Typhoon Nuri to model grid spacing: ERA_d03 and FNL_d03 are performed at 4-km horizontal grid spacing, nested into the D02 (12-km grid spacing) of ERA and FNL, respectively, using a two-way nested method. The simulations at the 4-km grid spacing use the same physical parameterization schemes as those at the 12-km grid spacing except that they exclude the cumulus scheme. The simulations also start at 1800 UTC 15 August. Figure 1 shows the locations of all domains. The basic model configurations for ERA_d03 and FNL_d03 are described in Table 1.

Similar to FNL, FNL_d03 fails to simulate Nuri’s genesis (Fig. 6a). Meanwhile, ERA_d03 produces a weaker simulated intensity than ERA (Fig. 2b). The final simulated MSLP at 1800 UTC 18 August in ERA_d03 is weaker than that in ERA by over 10 hPa. Overall, the higher-resolution grid spacing (4 km) does not help the model produce a better simulation of Nuri’s genesis.1 Similar results were documented in Davis and Bosart (2002), in which the higher-resolution simulation did not outperform the coarser resolution in the numerical simulation of the genesis of Hurricane Diana.

Fig. 6.
Fig. 6.

Time series of the minimum center sea level pressure (hPa) from numerical simulations compared with JTWC best-track data. Sensitivity to (a) grid spacings (ERA, ERA_d03, FNL, and FNL_d03) and (b) forecast leading times (ERA_E24h, FNL_E24h, ERA_E24h_d03, and FNL_E24h_d03).

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

2) Forecast lead time

To examine the sensitivity of numerical simulations to forecast lead time, four simulations with an early initialization time (1800 UTC 14 August 2008) are set up with a 48-h lead time relative to the time of Nuri’s genesis (1800 UTC 16 August 2008). Since the new initialization time (1800 UTC 14 August 2008) is 24 h earlier than the initialization time (1800 UTC 15 August 2008) in the control simulations (ERA and FNL), the new simulations are denoted as ERA_E24h and FNL_E24h, and the model physics configurations are the same as those used in ERA and FNL. Then, two simulations (ERA_E24h_d03 and FNL_E24h_d03) at the 4-km resolution domains (D03) are conducted with the initialization time at 1800 UTC 14 August 2008. The sizes of the model domains are expanded to ensure D03 covers the preexisting disturbance at the initial time. The model configurations for the four simulations (ERA_E24h, FNL_E24h, ERA_E24h_d03, and FNL_E24h_d03) are also listed in Table 1.

Figure 6b shows the MSLP from the four simulations with the early initialization time compared with the best-track data. In terms of MSLP, all four of the simulations (ERA_E24h, FNL_E24h, ERA_E24h_d03, and FNL_E24h_d03) fail to predict TC genesis at 1800 UTC 16 August (Nuri’s genesis time) and have much weaker intensities compared with the best-track data. Despite two simulations (FNL_E24h and ERA_E24h_d03) reaching the intensity of TC genesis at about 1800 UTC 17 August (72-h forecasts), the simulated genesis time is up to 24 h later than Nuri’s genesis time. Consequently, these simulations are also unable to capture the early intensification of Typhoon Nuri.

4. Interpretation of numerical simulation results

a. Development of mesoscale convective systems in Nuri’s genesis

The above sensitivity experiments indicate that the numerical simulation of Nuri is very sensitive to initial conditions, forecast lead time, and model grid size. To interpret the different outcomes from the above experiments, differences among these simulations are investigated.

Notable differences are found among these experiments in the convective initiation and developments during the first few hours of the simulations. Therefore, the simulated convective initiations from various experiments are compared along with the observations. The convective system in the pre-Nuri disturbance was noted by ground-based radar at Guam, as it captured a strong convective system developing in the pre-Nuri disturbance (Fig. 7a). This convective system presented as a squall line along 146°E at 0031 UTC 16 August and propagated westward. At the same time, the P-3 aircraft flew over the pre-Nuri disturbance, and the airborne ELDORA captured the convective system. Figure 7b shows the radar reflectivity and wind analysis at 1.5 km. Strong reflectivity was observed in this convective system around 14°N, 146°E (Fig. 7b). The wind structure shows that the convective system approached the wave axis along 146°E (Fig. 7b). The importance of the convective system in the pre-Nuri disturbance was also noted by Montgomery et al. (2010).

Fig. 7.
Fig. 7.

Radar reflectivity (color shading) observed by (a) Guam NEXRAD at 0031 UTC Aug 16 2008 and (b) airborne ELDORA at 1.5 km AGL. The white box in (a) denotes the approximate coverage of (b). The plot in (a) is obtained from data online (http://catalog.eol.ucar.edu/tparc_2008/index.html). The wind vectors in (b) include Doppler radar wind analysis and dropsondes (with black dots at the tails of those vectors).

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

Similar to these observation results, the successful ERA simulation also predicted the strong, organized convective system (Fig. 8a): the large rain-rate region indicated the convective system associated with the preexisting disturbance at 7 h (0100 UTC 16 August). In contrast, the convective systems in FNL were scattered on the eastern side of the circulation (Fig. 8b), and no intense convective system developed near the circulation center (Fig. 4b). The agreement between the observations and the ERA simulation and the disagreement between the observations and the FNL simulation imply that the development of the convective system in the disturbance played an important role in Nuri’s genesis.

Fig. 8.
Fig. 8.

The 3-h accumulated precipitation (contours) and 850-hPa wind vectors valid at 0100 UTC 16 Aug in the control and sensitivity experiments: (left) ERA and (right) FNL. In (a), “ERA at 7 h” means 7-h forecast in ERA. Others follow the same pattern: (b) 7-h forecast in FNL, (c) 7-h forecast in ERA_d03, (d) 7-h forecast in FNL_d03, (e) 31-h forecast in ERA_E24h, (f) 31-h forecast in FNL_E24h, (g) 31-h forecast in ERA_E24h_d03, and (h) 31-h forecast in FNL_E24h_d03.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

Further comparison is conducted for all other numerical experiments. Compared with ERA (Fig. 8a), the convective system in ERA_d03 (at 4-km grid spacing) is weaker, as indicated by the weaker rainfall and smaller precipitation area (Fig. 8c). This contrast is consistent with the results of Davis and Bosart (2002), in which fifth-generation Pennsylvania State University–National Center for Atmospheric Research Mesoscale Model (MM5) higher-resolution simulations produced a weaker TC in its genesis phase. They suggested that the weaker intensity with higher resolution could be ascribed to more downdrafts embedded in the convection. Although it is beyond the scope of this paper to provide an explicit explanation for the weaker convection produced by the high-resolution (~4 km) simulation and its corresponding weaker TC, this result further demonstrates that the development of convective systems is important for Nuri’s genesis.

Similar to FNL (Fig. 8b), the convective systems in FNL_d03 are scattered on the eastern side of the circulation, far from the disturbance center (Fig. 8d). Since neither simulation (FNL or FNL_d03) produces intense convection in the central disturbance region, the preexisting disturbances in FNL and FNL_d03 do not intensify to become a TC.

The simulated convection around the preexisting disturbance is also examined at the same corresponding time (0100 UTC 16 August 2008, 31-h forecast) for the experiments ERA_E24h (Fig. 8e) and FNL_E24h (Fig. 8f). They both produce much weaker rainfall around the disturbance, indicating that no intense convective systems are produced in the simulations with earlier lead times. Similarly, ERA_E24h_d03 and FNL_E24h_d03 predict only scattered convective systems around the disturbance and no convection initiated near the center of the disturbance (Figs. 8g,h). Overall, ERA presents widespread convection over the circulation center (Fig. 8a), while four other simulations with FNL initial conditions, early lead times, or higher resolution do not predict such a convective system in the disturbance center (Figs. 8e–h), and they do not predict Nuri’s genesis.

b. Interaction between convection and pouch flow

Among all numerical experiments, ERA and FNL present typical developing and nondeveloping tropical cyclone cases with clear contrasts. In the following diagnoses, we emphasize these two experiments to further investigate the initial conditions and their evolution, which influence the prediction of the genesis of Nuri.

The above results indicate that convective system development is a key factor for Nuri’s genesis. However, convection occurs commonly in tropical disturbances. Only part of these disturbances become developing systems to generate TCs. Montgomery and his colleagues identified two processes that are necessary for a developing system (Dunkerton et al. 2009; Montgomery et al. 2010; Wang et al. 2010a,b). First, the pregenesis disturbance (e.g., an easterly wave) presents a circulation pattern in the comoving frame and the center of the circulation (referred to as a pouch) is the favored region for TC genesis. Second, the aggregation of convection-induced vorticity anomalies into the “sweet spot” (viz., the pouch center) enhances the pouch’s circulation. To examine these two processes in numerical simulations of Nuri’s genesis, the evolution of convective systems and flow structures in the pregenesis disturbance are illustrated.

To present the pouch flow (disturbance) structure during the pregenesis period, the comoving frame of reference shows the horizontal relative wind fields. By subtracting the propagation speed of the large-scale wave from the wind fields, the comoving frame presents a closed circulation, a so-called wave pouch (Dunkerton et al. 2009; Wang et al. 2010a). In this study, the wave propagation is indicated by the Hovmöller plots of the meridional wind at 850 hPa along 13°N (Fig. 9). The westward speed of the wave propagation is estimated by the zonal movement of the wave axes (shown as solid black lines in Fig. 9). The propagation speeds are calculated: −5.79 m s−1 in ERA and −3.22 m s−1 in FNL. Then, the relative winds in the comoving frame are obtained by subtracting the propagation speeds from the wind speeds in ERA and FNL, respectively.

Fig. 9.
Fig. 9.

Hovmöller plots of the meridional wind speed (m s−1) at 850 hPa along 13°N in (a) ERA and (b) FNL between 1800 UTC 15 Aug and 1800 UTC 16 Aug 2008 (0–24-h forecast). The black solid line indicates the propagation of the trough axes in both simulations.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

First, the evolution of the convection in the pouch is examined. Figure 10 shows the divergence and wind vectors at 850 hPa in the first 24 h in ERA. At 3 h (2100 UTC 15 August), the low-level confluent flow in the region southeast of the circulation leads to strong convergence and further enhancement of upward motion (Fig. 10a). From 6 to 9 h, the convective system (represented by the strong convergence at 850 hPa) develops pronouncedly into the pouch center, which is the intersection of the critical layer (purple line) and the wave trough axis (black line). Note that the strong southeast flow in the comoving frame points from the convection region to the pouch center at 6 h (Fig. 10b). Following such flow, the convection enters the pouch center at 9 h (Fig. 10c). Then, the pouch circulation protects the convective system, and eventually it develops into Nuri (Fig. 10d). Meanwhile, the circulation itself also grows rapidly until Nuri’s genesis (Fig. 10d).

Fig. 10.
Fig. 10.

The convergence (color shading) at 850 hPa and sea level pressure (blue contours; contour interval: 2 hPa) in ERA. The wind vectors present the 850-hPa horizontal winds in the co-moving frame. The purple lines indicate the critical layers (isopleths of zero relative zonal wind). The black lines indicate the wave’s trough axis. The black dots indicate the pouch center. The blue arrows indicate the direction of the convection movement: (a) 2100 UTC 15 Aug, (b) 0000 UTC 16 Aug, (c) 0300 UTC 16 Aug, and (d) 1800 UTC 16 Aug 2008. “ERA 3 h” means 3-h forecast in ERA from 1800 UTC 15 Aug 2008. Others follow the same pattern.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

In addition, to reveal how convection-induced vorticity contributes to the evolution of the pouch circulation, Fig. 11 presents vorticity anomalies at 850 hPa in the pouch circulation. At 9 h (Fig. 11a), the pouch circulation presented by a large area of positive vorticity anomalies (group A in Fig. 11a) with a sweet spot (denoted by the black dot) is clearly observed. Meanwhile, the cyclonic vortex (group B in Fig. 11a) emerges in the southeast quadrant of the pouch circulation. This cyclonic vortex is initiated by the local convection (as shown in Fig. 10a). In the following several hours (Figs. 11b,c), the vortex (group B) merges into the pouch circulation (group A). As a consequence, the pouch circulation is enhanced with stronger positive vorticity. In the next few hours (Fig. 11d, near Nuri’s genesis at 20 h), the pouch circulation is accelerated by the persistent convection near the sweet spot. This process evidently demonstrates that the convection-induced vortex merges into the pouch circulation and leads to the growth of the pouch circulation. Montgomery et al. (2010) also found that the convection emerges in the southeast quadrant of the pouch and the convective band wraps into the pouch center prior to Nuri’s genesis.

Fig. 11.
Fig. 11.

As in Fig. 10, but for relative vorticity (color shading). The two vortices discussed in the text are indicated by dark red arrows and letters (A and B): (a) 0300 UTC 16 Aug, (b) 0500 UTC 16 Aug, (c) 0700 UTC 16 Aug, and (d) 1400 UTC 16 Aug 2008.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

Compared with ERA, FNL (Fig. 12) shows a different process of convection initiation and movement. At 3 h, the convergent and ascending motions associated with the initiated convection are located mostly on the eastern side of the circulation (Fig. 12a). Convective systems are also located away from the pouch center. Even during their development, they are scattered on the eastern side of the circulation core region (Figs. 12b,c). No obvious convective system has reached the pouch center by 9 h. Therefore, because the initial convergence region is far from the center and there is a lack of appropriate flow pointing from the convection to the pouch center, the FNL simulation eventually leads to a nondeveloping system (Fig. 11d).

Fig. 12.
Fig. 12.

As in Fig. 10, but for the FNL.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

The discrepancies between the ERA and FNL simulations in terms of the development of the initial convective systems and their interactions with the pouches demonstrate the importance of the flow structure of the pouch—the appropriate flow, namely, the southeast flow in ERA leads to the advection of the initial convection into the pouch center; by contrast, the initial convection in FNL develops far from the pouch center and without the appropriate flow.

c. Effects of environmental conditions

To better understand the influence of the initial conditions in predicting Nuri’s genesis, we revisit several factors that are recognized as necessary for the genesis of a TC (as mentioned in the introduction): warm SSTs, weak environmental vertical wind shear, a preexisting disturbance with high relative humidity, and abundant deep convective clouds and precipitation (latent heat release near the incipient center).

1) SSTs and vertical wind shear

SSTs are sufficiently warm in both ERA and FNL, and are generally greater than 28°C over the region of Nuri’s genesis. The distributions of SSTs are also similar in the two experiments, and the differences in Nuri’s track are minimal. To confirm the impact of SST differences on numerical simulations of Nuri’s genesis, a set of sensitivity experiments is conducted by swapping the SSTs between ERA and FNL. Results indicate very marginal differences from the control simulations in terms of both track and intensity forecasts (figure not shown), confirming that the SST is a necessary but not a determining condition for Nuri’s genesis.

Following Frank and Ritchie (2001) and Knaff et al. (2004), we checked the environmental vertical wind shear between 850 and 200 hPa over a circle within a 500-km radius centered at the circulation center (figure not shown). During the pregenesis phase (before 24 h), the vertical wind shear increases quickly at the beginning of both simulations and then decrease during 12–21 h. Although a slightly larger variation in vertical wind shear is found in FNL, the shear reaches similar values of 6–7 m s−1 at 24 h when Nuri’s genesis occurs. In spite of the differences of over 1 m s−1 in magnitude, the lower-tropospheric wind shear (850–500 hPa) show similar variations in the two experiments prior to genesis. Specifically, shear values decrease before genesis and reach their lowest values around the time of genesis. It is likely that this reduction in lower-tropospheric wind shear is favorable for Nuri’s genesis. However, this lower wind shear in the lower troposphere should be a necessary but not sufficient factor influencing genesis, since both developing (ERA) and nondeveloping (FNL) cases show a similar trend of lower-tropospheric wind shear around genesis time (24 h). An increasing trend of upper-tropospheric wind shear is found in both experiments, and the overall evolution of the upper-tropospheric wind shear shows a weak relationship with Nuri’s genesis.

2) Midlevel vorticity and moisture conditions

After convective systems move into the pouch center, continued convective development requires proper moisture conditions (near saturation) in the pouch because the entrainment of dry air into the convective clouds yields downdrafts, which definitely inhibit TC genesis (Rappin et al. 2010; Montgomery et al. 2010).

In the initial conditions (at 0 h), preexisting disturbances (easterly waves prior to Nuri) are found in both ERA and FNL to the east of Guam. Figure 13 shows a vertical cross section (through the circulation center) of the initial vorticity and relative humidity. In FNL, the positive relative vorticity is concentrated at low levels, and the vertical axis of the relative vorticity is tilted from 800 to 600 hPa. Above 600 hPa, the positive relative vorticity is separated and dissipated. In ERA, the relative vorticity extends from near the surface to much higher levels (up to 400–300 hPa), with a vorticity maximum at 850 hPa. According to Davis and Bosart (2002) and Musgrave et al. (2008), this type of initial disturbance at middle to higher levels is an important factor for TC genesis.

Fig. 13.
Fig. 13.

Vertical cross sections of the initial relative vorticity (contour interval: 1 × 10−5 s−1) and the initial relative humidity (contour interval: 5%) at 1800 UTC 15 Aug 2008: (a),(b) the relative vorticity in FNL and ERA, respectively; (c),(d) the relative humidity in FNL and ERA, respectively.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

Meanwhile, a concentrated moist column (RH > 85%) aligns well with midlevel vorticity in ERA, while relatively dry air is collocated with the center of vorticity in FNL. In particular, the middle levels (about 500 hPa) and the near-surface levels (below 900 hPa) are much drier in FNL than in ERA. Because the surface moisture flux leads to air–sea equilibrium (Jorgensen et al. 1997), the near-surface levels achieve near-saturation conditions quickly. Hence, the midlevel moisture difference between FNL and ERA during the subsequent simulation period is more significant. Previous studies (Dunkerton et al. 2009; Montgomery et al. 2010) demonstrated the importance of the midlevel vortex for protecting the moist air against the intrusion of environmental dry air. Thus, the evolution of the moisture fields in ERA and FNL in the middle levels is further compared.

Remarkable discrepancies in moisture conditions and midlevel vorticity are noted between ERA and FNL in the forecasts. Figure 14 compares the evolution of the saturation fraction, defined as the ratio of total precipitable water to saturated precipitable water through the troposphere. Figure 15 compares the midlevel (500 hPa) vorticity. The following points are apparent: in ERA, the moist air mass is concentrated inside the pouch, which is isolated from dry air outside (Figs. 14a,c,e); the strong midlevel relative vorticity is collocated with the moist air inside the pouch (Figs. 15a,c,e); and the pouch circulation wraps the moist air and propagates westward with it. Such moist conditions in the pouch benefit the development of convection. But in FNL, the saturation fraction (Figs. 14b,d,f) indicates much less moisture, while the relative vorticity (Figs. 15b,d,f) is much weaker and more scattered.

Fig. 14.
Fig. 14.

Comparisons of the saturation fraction (color shading; %) and the horizontal winds at 500 hPa in the co-moving frame between (left) ERA and (right) FNL at (a),(b) 1800 UTC 15 Aug; (c),(d) 0300 UTC 16 Aug; and (e),(f) 2100 UTC 16 Aug 2008.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

Fig. 15.
Fig. 15.

Comparisons of the relative vorticity (color shading; 10−5 s−1), relative humidity (red contour lines; contour interval: 5%) and geopotential height (blue contours; contour interval: 10 m), and the horizontal winds in the co-moving frame at 500 hPa between (left) ERA and (right) FNL at (a),(b) 1800 UTC 15 Aug; (c),(d) 0300 UTC 16 Aug; and (e),(f) 1800 UTC 16 Aug 2008.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

Two reasons account for the differences in moisture distribution between ERA and FNL. First, the FNL initial conditions are much less moist than are the ERA initial conditions. Second, the midlevel circulation, namely, the vorticity at 500 hPa, is much stronger in ERA than in FNL from pre-Nuri to Nuri’s genesis (0–24 h). In ERA, the strong circulation protects the moist air by preventing the intrusion of dry air into the center of the circulation. But in FNL, there is no closed circulation at 500 hPa until 24 h. The moisture dissipates significantly without the protection of the closed circulation. The important role of the midlevel circulation has been demonstrated in Dunkerton et al. (2009) and Wang et al. (2010a). Overall, the moisture inside the pouch and midlevel vortex favors the persistent convection and eventually leads to Nuri’s genesis.

3) Sensitivity to moisture conditions

To further solidify the impact of the initial moisture conditions in Nuri’s development, two new sets of simulations (ERA-WV and FNL-WV) are conducted by swapping the water vapor fields between the initial conditions of ERA and FNL. Namely, the experiment ERA_WV (FNL_WV) uses initial and boundary conditions derived from ERA-Interim (NCEP FNL) except that the moisture field comes from NCEP FNL (ERA-Interm). The basic model configurations for the two experiments (ERA-WV and FNL-WV) are described in Table 1. If moisture conditions are significant for Nuri’s genesis, the forecast of Nuri’s genesis in ERA-WV should be inhibited, while FNL-WV should simulate the genesis of Nuri, in contrast to the nondeveloping disturbance in the FNL experiment.

As expected, both the ERA-WV and FNL-WV simulations are significantly different from ERA and FNL (Fig. 16). In particular, the simulated Nuri is much weaker in ERA-WV than in ERA, with an MSLP difference of nearly 15 hPa at the end of the simulation. Compared with FNL, FNL-WV presents a significant increase in simulated intensity (Fig. 16). Although FNL-WV does not predict the same intensity as the best track, the simulation reaches the genesis intensity at 42 h. These results clearly demonstrate that the simulation of Nuri’s genesis is very sensitive to initial moisture conditions. Obviously, moisture conditions play an important role in Nuri’s genesis: ERA initial moisture conditions are necessary for successfully simulating Nuri’s genesis.

Fig. 16.
Fig. 16.

Time series of simulated MSLP (hPa) between 1800 UTC 15 Aug and 1800 UTC 18 Aug 2008 in the sensitivity experiments with swapped initial moisture conditions, compared with JTWC best-track data.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

d. Evolution of thermodynamic conditions and secondary circulations

Previous sections indicated that convection development in the pouch and moisture conditions play an important role in Nuri’s genesis. In this section, we further examine how thermodynamic and dynamic conditions evolve with convection and moisture conditions to result in the formation of Nuri.

Figure 17 shows the radius–pressure distribution of the temperature perturbation at the initial time in ERA; the temperature within the inner core (radius <300 km) presents weak warm anomalies between 600 and 700 hPa and cold anomalies below 800 hPa (Fig. 17a). With cold anomalies in the lower level and warm anomalies above, this temperature structure is consistent with previous results from observations (Bister and Emanuel 1997; Raymond et al. 2011; Davis and Ahijevych 2013) and from simulations [the idealized simulations by Montgomery et al. (2006), the simulation of Tropical Storm Gert’s genesis (2005) by Braun et al. (2010), and the simulation of Hurricane Felix’s (2007) genesis by Wang et al. (2010b)]. Wang (2012) ascribed this lower-level cooling to the widespread stratiform processes around the vortex center. Recent studies by Raymond and Sessions (2007) and Raymond et al. (2011) suggest that cold anomalies at low levels and warm anomalies aloft result in vertical mass flux profiles that are “bottom heavy” (e.g., the maximum vertical mass flux is present at 800 hPa in Fig. 17a). Such bottom-heavy vertical mass flux enhances the low-level inflow (e.g., as shown in Fig. 18a) and the low-level vorticity, both of which provide favorable conditions for TC genesis.

Fig. 17.
Fig. 17.

Radius–pressure cross sections of the azimuthally averaged values of temperature perturbations (color shading; shading interval: 0.25 K), calculated as differences between the temperature near the simulated circulation center (0–350 km) and the environmental temperature that is determined by averaging the temperature within a radial band between 500–900 km, and the vertical mass flux (contour interval: 10−5 kg m−2 s−1) in (left) ERA and (right) FNL at (a),(b) 1800 UTC 15 Aug; (c),(d) 0000 UTC 16 Aug; (e),(f) 1800 UTC 16 Aug; and (g),(h) 0000 UTC 17 Aug 2008.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

Fig. 18.
Fig. 18.

Radius–pressure cross sections of the azimuthally averaged relative humidity (contour interval: 5%) and secondary circulation (represented by uw vectors; u unit: m s−1; w unit: cm s−1; u is the radial velocity; w is the vertical velocity) in (left) ERA and (right) FNL at (a),(b) 1800 UTC 15 Aug; (c),(d) 0000 UTC 16 Aug; (e),(f) 1800 UTC 16 Aug; and (g),(h) 0000 UTC 17 Aug 2008.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

Owing to the development of convection, vertical mass fluxes increase significantly with time in subsequent simulation forecasts (Fig. 17). In ERA, the vertical airmass fluxes reach their maxima at 500–300 hPa at 6 h, 600–200 hPa at 24 h (the time of Nuri’s genesis), and 800–200 hPa (postgenesis), along with the formation and strengthening of the warm core as the warm temperature anomalies align with the maximum airmass fluxes.

To further explain the formation of the warm core, Fig. 18 shows the evolution of the secondary circulation (the vectors represent the azimuthally averaged inflow or radial velocity in the horizontal direction and the vertical velocity in the vertical direction) and the relative humidity. Figures 18a, 18c, 18e, and 18g clearly reveal the role of low-level inflow, along with the updraft due to the convection, in transporting the moist air from lower to middle and upper levels. In the early phase (0–6 h; Figs. 18a,c), the inflow and updraft bring moist air up to enhance the moisture field in the middle to upper levels (700–300 hPa). Due to the condensation of water vapor in the moist air at the upper levels, latent heat is released and causes warming at these levels (400–300 hPa). Then, when the convection is enhanced, the inflows combined with the updraft bring even more moist air (with upward airmass fluxes continuously strengthening) to the middle to upper levels (Figs. 18e,g), and the warm core forms before and near Nuri’s genesis and then further enhances afterward (Figs. 17e,g). Figure 19 demonstrates that the diabatic heating induced by convection contributes to the upper-level warming. The intense rainfall rate (shown in Fig. 4) also validates the considerable latent heat release associated with the convective precipitation. Zhang and Zhu (2012) found that the upper-level warming near the center of the tropical cyclone could account for almost the entire total drop in MSLP. Rapid decreases in Nuri’s MSLP after 18 h (Fig. 2) further illustrate the correspondence between upper-level warming and decreasing MSLP.

Fig. 19.
Fig. 19.

Time series of vertical profiles of area-averaged (over the area of 100 km of radius from the center of Nuri) diabatic heating (contour interval: 0.05 K h−1) in (a) ERA and (b) FNL from 1800 UTC 15 Aug to 0000 UTC 17 Aug 2008.

Citation: Weather and Forecasting 29, 6; 10.1175/WAF-D-14-00003.1

In contrast, no significant low-level cold anomalies appear in FNL throughout the entire simulation (Fig. 17). During this phase, the upper-level warm anomalies in FNL are much weaker than in ERA due to the lack of deep convection and latent heat release in the circulation core region (Fig. 17). Without the development of deep convection and low-level inflow, the midlevel moisture is much less enhanced in the core region (radius <100 km) in FNL compared with that in ERA (Fig. 18).

5. Summary and concluding remarks

Numerical simulations of the genesis of Typhoon Nuri (2008) are conducted using the WRF Model with initial and boundary conditions derived from two different global analyses: ERA-Interim and NCEP FNL. First, with 12-km horizontal resolution in two-level nested, two-way interacting WRF Model grids and a 24-h lead time (starting at 18 UTC 15 August 2008), it is found that the numerical simulation with FNL initial and boundary conditions fails to predict the genesis of Nuri, while the simulation using ERA-Interim successfully captures the processes of Nuri’s genesis at 1800 UTC 16 August 2008, as well as its early intensification, as it produces stronger cyclonic circulation and upward motion. The sensitivity of the numerical simulations to model horizontal grid size and forecast lead time is also examined. It is found that the higher-resolution grid spacing (4 km) does not help the model produce a better simulation of Nuri’s genesis. In addition, the earlier (48 h instead of 24 h) forecast lead time does not improve the prediction of Nuri’s genesis. The failure to predict Nuri’s genesis on 4-km model grids, as shown in Li (2013), is mostly related to the uncertainties in representing the physics process at higher-resolution grids and is thus beyond the emphasis of this study. Discrepancies between simulations with and without Nuri’s development are diagnosed to understand the influence of the initial conditions on numerical simulations of Nuri’s genesis. The dynamic and thermodynamic processes that contribute to Nuri’s genesis are also examined.

It was found that sufficiently warm SSTs and the moderate-to-weak vertical wind shear provide favorable environmental conditions in both developing and nondeveloping simulations. Since these two conditions appear in both cases, they are necessary but not deterministic factors that lead to Nuri’s genesis. Other processes, such as initial moisture conditions and development of intense convection, play important roles in the development or nondevelopment of Nuri. Specifically, significant differences are found in the development and organization of the intense convection during Nuri’s pregenesis phase. In the developing case, the convective systems advect, evolve, and organize into a “pouch” center of a wavelike disturbance. In the nondeveloping case, the convective systems either fail to develop and organize or do not advect into the pouch center.

The initial conditions, which strongly influence the evolution within the first few hours of the simulations (the initiation phases), are shown to be critical for the establishment of a favorable pregenesis circulation. In the developing case (ERA), the convection is initiated by low-level convergence southeast of the pouch. With the southeast flow pointing from the original convection to the pouch center, convective bursts develop downstream into the pouch center. Therefore, at the end of the initiation phase, such evolution leads to a favorable pouch with persistent moist convection in the center, which develops rapidly into Nuri’s genesis. By contrast, the nondeveloping case (FNL) presents the initial convection in the eastern region of the disturbance. But these convective bursts do not enter the circulation center because the closed circulation provides southerly winds, which prevent the convection from developing into the center. Consequently, the disturbance center from the low to middle levels presents rare convection, which eventually leads to the nondevelopment in FNL. As the above comparison demonstrates, the favorable pre-Nuri pouch requires that the convection, which is initiated by the convergence at low levels, enter the pouch center.

The initial moisture conditions are another key environmental condition for the accurate prediction of Nuri’s genesis. Compared with FNL, moist air in ERA is concentrated more at the middle levels (~500 hPa) at the initial time. The stronger midlevel vortex in ERA maintains the moist environment, which is favorable for the persistent development of deep convection and Nuri’s genesis. The impact of the moisture field on simulations of Nuri’s genesis is further confirmed by a set of sensitivity experiments (ERA-WV and FNL-WV), in which the water vapor fields in the two original numerical experiments (ERA and FNL) are swapped. Results confirm that the enhancement of mid- to upper-level moisture is favorable for Nuri’s genesis. The deep convection produces diabatic heating from latent heat release when vertical airmass flux maxima occur in the mid- to upper-level atmosphere. The substantial warming at upper levels induced by the latent heat release from persistent deep convection contributes to the drop in Nuri’s minimum central sea level pressure. Specifically, in ERA, the vertical airmass fluxes reach maxima at 500–300 hPa at 6 h, 600–200 hPa at 24 h (the time of Nuri’s genesis), and 800–200 hPa (postgenesis), along with the formation and strengthening of the warm core as the warm temperature anomalies align with the maximum airmass fluxes. As a consequence, a significant drop in minimum sea level pressure is achieved to form a tropical cyclone.

The overall results from this study demonstrate that it is essential to accurately represent the initial conditions, especially those conditions associated with the key process of TC genesis (e.g., moisture, large-scale steering flow, etc.) in the numerical prediction of tropical cyclone geneses.

Acknowledgments

The authors acknowledge the NCAR WRF Model development group for their efforts in developing the community model. Computer support from the Center for High Performance Computing (CHPC) at the University of Utah is appreciated. This study is supported by Office of Naval Research (ONR) Grant N000141310582.

REFERENCES

  • Bister, M., , and Emanuel K. A. , 1997: The genesis of Hurricane Guillermo: TEXMEX analyses and a modeling study. Mon. Wea. Rev., 125, 26622682, doi:10.1175/1520-0493(1997)125<2662:TGOHGT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bosart, L. F., , and Bartlo J. A. , 1991: Tropical storm formation in a baroclinic environment. Mon. Wea. Rev., 119, 19792013, doi:10.1175/1520-0493(1991)119<1979:TSFIAB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bosart, L. F., , Bracken W. E. , , Molinari J. , , Velden C. S. , , and Black P. G. , 2000: Environmental influences on the rapid intensification of Hurricane Opal (1995) over the Gulf of Mexico. Mon. Wea. Rev., 128, 322352, doi:10.1175/1520-0493(2000)128<0322:EIOTRI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bracken, W. E., , and Bosart L. F. , 2000: The role of synoptic-scale flow during tropical cyclogenesis over the North Atlantic Ocean. Mon. Wea. Rev., 128, 353376, doi:10.1175/1520-0493(2000)128<0353:TROSSF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Braun, S. A., , Montgomery M. T. , , Mallen K. J. , , and Reasor P. D. , 2010: Simulation and interpretation of the genesis of Tropical Storm Gert (2005) as part of the NASA Tropical Cloud Systems and Processes Experiment. J. Atmos. Sci., 67, 9991025, doi:10.1175/2009JAS3140.1.

    • Search Google Scholar
    • Export Citation
  • Briegel, L. M., , and Frank W. M. , 1997: Large-scale influences on tropical cyclogenesis in the western North Pacific. Mon. Wea. Rev., 125, 13971413, doi:10.1175/1520-0493(1997)125<1397:LSIOTC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Davis, C. A., , and Bosart L. F. , 2001: Numerical simulations of the genesis of Hurricane Diana (1984). Part I: Control simulation. Mon. Wea. Rev., 129, 18591881, doi:10.1175/1520-0493(2001)129<1859:NSOTGO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Davis, C. A., , and Bosart L. F. , 2002: Numerical simulations of the genesis of Hurricane Diana (1984). Part II: Sensitivity of track and intensity prediction. Mon. Wea. Rev., 130, 11001124, doi:10.1175/1520-0493(2002)130<1100:NSOTGO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Davis, C. A., , and Bosart L. F. , 2003: Baroclinically induced tropical cyclogenesis. Mon. Wea. Rev., 131, 27302747, doi:10.1175/1520-0493(2003)131<2730:BITC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Davis, C. A., , and Ahijevych D. A. , 2013: Thermodynamic environments of deep convection in Atlantic tropical disturbances. J. Atmos. Sci., 70, 19121928, doi:10.1175/JAS-D-12-0278.1.

    • Search Google Scholar
    • Export Citation
  • DeMaria, M., , and Pickle J. D. , 1988: A simplified system of equations for simulation of tropical cyclones. J. Atmos. Sci., 45, 15421554, doi:10.1175/1520-0469(1988)045<1542:ASSOEF>2.0.CO;2.

    • 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, doi:10.1175/1520-0469(1989)046<3077:NSOCOD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dunkerton, T. J., , Montgomery M. T. , , and Wang Z. , 2009: Tropical cyclogenesis in a tropical wave critical layer: Easterly waves. Atmos. Chem. Phys., 9, 55875646, doi:10.5194/acp-9-5587-2009.

    • Search Google Scholar
    • Export Citation
  • Frank, W. M., , and Ritchie E. A. , 2001: Effects of vertical wind shear on the intensity and structure of numerically simulated hurricanes. Mon. Wea. Rev., 129, 22492269, doi:10.1175/1520-0493(2001)129<2249:EOVWSO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Gamache, J. F., , Dodge P. P. , , and Griffin N. F. , 2008: Automatic quality control and analysis of airborne Doppler data: Real-time applications, and automatically post-processed analyses for research. 28th Conf. on Hurricanes and Tropical Meteorology, Orlando, FL, Amer. Meteor. Soc., P2B.12. [Available online at http://ams.confex.com/ams/pdfpapers/137969.pdf.]

  • Gao, J., , Xue M. , , Shapiro A. , , Xu Q. , , and Droegemeier K. K. , 2001: Three-dimensional simple adjoint velocity retrievals from single-Doppler radar. J. Atmos. Oceanic Technol., 18, 2638, doi:10.1175/1520-0426(2001)018<0026:TDSAVR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Gray, W. M., 1968: Global view on the origins of tropical cyclones. Mon. Wea. Rev., 96, 669700, doi:10.1175/1520-0493(1968)096<0669:GVOTOO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hennon, C. C., , and Hobgood J. S. , 2003: Forecasting tropical cyclogenesis over the Atlantic basin using large-scale data. Mon. Wea. Rev., 131, 29272940, doi:10.1175/1520-0493(2003)131<2927:FTCOTA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hong, S.-Y., , and Lim J.-O. J. , 2006: The WRF single-moment 6-class microphysics scheme (WSM6). J. Korean Meteor. Soc., 42, 129151.

  • Hong, S.-Y., , Noh Y. , , and Dudhia J. , 2006: A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Wea. Rev., 134, 23182341, doi:10.1175/MWR3199.1.

    • Search Google Scholar
    • Export Citation
  • Jin, Y., , Peng M. S. , , and Jin H. , 2008: Simulating the formation of Hurricane Katrina (2005). Geophys. Res. Lett., 35, L11802, doi:10.1029/2008GL033168.

    • Search Google Scholar
    • Export Citation
  • Jorgensen, D. P., , LeMone M. A. , , and Trier S. B. , 1997: Structure and evolution of the 22 February 1993 TOGA COARE squall line: Aircraft observations of precipitation, circulation, and surface energy fluxes. J. Atmos. Sci., 54, 19611985, doi:10.1175/1520-0469(1997)054<1961:SAEOTF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kain, J. S., , and Fritsch J. M. , 1993: Convective parameterization for mesoscale models: The Kain–Fritsch scheme. The Representation of Cumulus Convection in Numerical Models, Meteor. Monogr., No. 46, Amer. Meteor. Soc., 165–170.

  • Kerns, B., , Greene K. , , and Zipser E. , 2008: Four years of tropical ERA-40 vorticity maxima tracks. Part I: Climatology and vertical vorticity structure. Mon. Wea. Rev., 136, 43014319, doi:10.1175/2008MWR2390.1.

    • Search Google Scholar
    • Export Citation
  • Kieu, C. Q., , and Zhang D.-L. , 2010: Genesis of Tropical Storm Eugene (2005) from merging vortices associated with ITCZ breakdowns. Part III: Sensitivity to various genesis parameters. J. Atmos. Sci., 67, 17451758, doi:10.1175/2010JAS3227.1.

    • Search Google Scholar
    • Export Citation
  • Knaff, J. A., , Seseske S. A. , , DeMaria M. , , and Demuth J. L. , 2004: On the influences of vertical wind shear on symmetric tropical cyclone structure derived from AMSU. Mon. Wea. Rev., 132, 25032510, doi:10.1175/1520-0493(2004)132<2503:OTIOVW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Komaromi, W., 2013: An investigation of composite dropsonde profiles for developing and nondeveloping tropical waves during the 2010 PREDICT field campaign. J. Atmos. Sci., 70, 542558, doi:10.1175/JAS-D-12-052.1.

    • Search Google Scholar
    • Export Citation
  • Li, Z., 2013: Studying the genesis of Typhoon Nuri (2008) with numerical simulations and data assimilation. Ph.D. dissertation, University of Utah, 164 pp. [Available from Dept. of Atmospheric Sciences, University of Utah, Salt Lake City, UT 84112.]

  • McBride, J. L., , and Zehr R. , 1981: Observational analysis of tropical cyclone formation. Part II: Comparison of nondeveloping versus developing systems. J. Atmos. Sci., 38, 11321151, doi:10.1175/1520-0469(1981)038<1132:OAOTCF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Merrill, R. T., , and Velden C. S. , 1996: A three-dimensional analysis of the outflow layer of Supertyphoon Flo (1990). Mon. Wea. Rev., 124, 4763, doi:10.1175/1520-0493(1996)124<0047:ATDAOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Mlawer, E. J., , Taubman S. J. , , Brown P. D. , , Iacono M. J. , , and Clough S. A. , 1997: Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102, 16 66316 682, doi:10.1029/97JD00237.

    • Search Google Scholar
    • Export Citation
  • Möller, J. D., , and Montgomery M. T. , 1999: Vortex Rossby waves and hurricane intensification in a barotropic model. J. Atmos. Sci., 56, 16741687, doi:10.1175/1520-0469(1999)056<1674:VRWAHI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Montgomery, M. T., , and Enagonio J. , 1998: Tropical cyclogenesis via convectively forced vortex Rossby waves in a three-dimensional quasigeostrophic model. J. Atmos. Sci., 55, 31763207, doi:10.1175/1520-0469(1998)055<3176:TCVCFV>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Montgomery, M. T., , Nicholls M. E. , , Cram T. A. , , and Saunders A. B. , 2006: A vortical hot tower route to tropical cyclogenesis. J. Atmos. Sci., 63, 355386, doi:10.1175/JAS3604.1.

    • Search Google Scholar
    • Export Citation
  • Montgomery, M. T., , Lussier L. III, , Moore R. W. , , and Wang Z. , 2010: The genesis of Typhoon Nuri as observed during the Tropical Cyclone Structure 2008 (TCS-08) field experiment—Part 1: The role of the easterly wave critical layer. Atmos. Chem. Phys., 10, 98799900, doi:10.5194/acp-10-9879-2010.

    • Search Google Scholar
    • Export Citation
  • Musgrave, K. D., , Davis C. A. , , and Montgomery M. T. , 2008: Numerical simulations of the formation of Hurricane Gabrielle (2001). Mon. Wea. Rev., 136, 31513167, doi:10.1175/2007MWR2110.1.

    • Search Google Scholar
    • Export Citation
  • Oye, R., , Mueller C. , , and Smith S. , 1995: Software for translation, visualization, editing and interpolation. Preprints, 27th Conf. on Radar Meteorology, Vail, CO, Amer. Meteor. Soc., 359–361.

  • Peng, M. S., , Fu B. , , Li T. , , and Stevens D. E. , 2012: Developing versus nondeveloping disturbances for tropical cyclone formation. Part I: North Atlantic. Mon. Wea. Rev., 140, 10471066, doi:10.1175/2011MWR3617.1.

    • Search Google Scholar
    • Export Citation
  • Pu, Z., , and Zhang L. , 2010: Validation of AIRS temperature and moisture profiles over tropical oceans and their impact on numerical simulations of tropical cyclones. J. Geophys. Res., 115, D24114, doi:10.1029/2010JD014258.

    • Search Google Scholar
    • Export Citation
  • Ramsay, H. A., , and Sobel A. H. , 2011: Effects of relative and absolute sea surface temperature on tropical cyclone potential intensity using a single-column model. J. Climate, 24, 183193, doi:10.1175/2010JCLI3690.1.

    • Search Google Scholar
    • Export Citation
  • Rappin, E. D., , and Nolan D. S. , 2012: The effect of vertical shear orientation on tropical cyclogenesis. Quart. J. Roy. Meteor. Soc., 138, 10351054, doi:10.1002/qj.977.

    • Search Google Scholar
    • Export Citation
  • Rappin, E. D., , Nolan D. S. , , and Emanuel K. A. , 2010: Thermodynamic control of tropical cyclogenesis in environments of radiative–convective equilibrium with shear. Quart. J. Roy. Meteor. Soc., 136, 19541971, doi:10.1002/qj.706.

    • Search Google Scholar
    • Export Citation
  • Raymond, D. J., , and Sessions S. L. , 2007: Evolution of convection during tropical cyclogenesis. Geophys. Res. Lett., 34, L06811, doi:10.1029/2006GL028607.

    • Search Google Scholar
    • Export Citation
  • Raymond, D. J., , Sessions S. L. , , and López Carrillo C. L. , 2011: Thermodynamics of tropical cyclogenesis in the northwest Pacific. J. Geophys. Res., 116, D18101, doi:10.1029/2011JD015624.

    • Search Google Scholar
    • Export Citation
  • Reed, R. J., , Norquist D. C. , , and Recker E. E. , 1977: The structure and properties of African wave disturbances as observed during phase III of GATE. Mon. Wea. Rev., 105, 317333, doi:10.1175/1520-0493(1977)105<0317:TSAPOA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Ritchie, E. A., , and Holland G. J. , 1997: Scale interactions during the formation of Typhoon Irving. Mon. Wea. Rev., 125, 13771396, doi:10.1175/1520-0493(1997)125<1377:SIDTFO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Rodgers, E., , Olson W. , , Halverson J. , , Simpson J. , , and Pierce H. , 2000: Environmental forcing of Supertyphoon Paka’s (1997) latent heat structure. J. Appl. Meteor., 39, 19832006, doi:10.1175/1520-0450(2001)040<1983:EFOSPS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Shen, B.-W., , Tao W.-K. , , Lau W. K. , , and Atlas R. , 2010: Predicting tropical cyclogenesis with a global mesoscale model: Hierarchical multiscale interactions during the formation of Tropical Cyclone Nargis (2008). J. Geophys. Res., 115, D14102, doi:10.1029/2009JD013140.

    • Search Google Scholar
    • Export Citation
  • Skamarock, W. C., and et al. , 2008: A description of the Advanced Research WRF version 3. Tech. Note NCAR/TN-475+STR, 113 pp. [Available online at http://www2.mmm.ucar.edu/wrf/users/docs/arw_v3.pdf.]

  • Thatcher, L., , and Pu Z. , 2013: Evaluation of tropical cyclone genesis precursors with relative operating characteristics (ROC) in high-resolution ensemble forecasts: Hurricane Ernesto. Trop. Cyclone Res. Rev., 2, 131148, doi:10.6057/2013TCRR03.01.

    • Search Google Scholar
    • Export Citation
  • Vecchi, G. A., , and Soden B. J. , 2007: Effect of remote sea surface temperature change on tropical cyclone potential intensity. Nature, 450, 10661070, doi:10.1038/nature06423.

    • Search Google Scholar
    • Export Citation
  • Wang, Z., 2012: Thermodynamic aspects of tropical cyclone formation. J. Atmos. Sci., 69, 24332451, doi:10.1175/JAS-D-11-0298.1.

  • Wang, Z., , Montgomery M. T. , , and Dunkerton T. J. , 2010a: Genesis of pre–Hurricane Felix (2007). Part I: The role of the easterly wave critical layer. J. Atmos. Sci., 67, 17111729, doi:10.1175/2009JAS3420.1.

    • Search Google Scholar
    • Export Citation
  • Wang, Z., , Montgomery M. T. , , and Dunkerton T. J. , 2010b: Genesis of pre-Hurricane Felix (2007). Part II: Warm core formation, precipitation evolution, and predictability. J. Atmos. Sci., 67, 17301744, doi:10.1175/2010JAS3435.1.

    • Search Google Scholar
    • Export Citation
  • Willoughby, H. E., , and Black P. G. , 1996: Hurricane Andrew in Florida: Dynamics of a disaster. Bull. Amer. Meteor. Soc., 77, 543549, doi:10.1175/1520-0477(1996)077<0543:HAIFDO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Zhang, D.-L., , and Zhu L. , 2012: Roles of upper-level processes in tropical cyclogenesis. Geophys. Res. Lett., 39, L17804, doi:10.1029/2012GL053140.

    • Search Google Scholar
    • Export Citation
1

The failure to predict Nuri’s genesis at 4-km model grids, as shown in Li (2013), is mostly related to the uncertainties in representing the physics process on higher-resolution grids and is thus beyond the scope of this study.

Save