• Adamson, D. S., S. E. Belcher, B. J. Hoskins, and R. S. Plant, 2006: Boundary layer friction in mid-latitude cyclones. Quart. J. Roy. Meteor. Soc., 132 , 101124.

    • Search Google Scholar
    • Export Citation
  • Andreas, E. L., 1992: Sea spray and the turbulent air–sea heat fluxes. J. Geophys. Res., 97 , 1142911441.

  • Andreas, E. L., 1995: The temperature of evaporating sea spray droplets. J. Atmos. Sci., 52 , 852862.

  • Andreas, E. L., 1998: A new sea spray generation function for wind speeds up to 32 m s−1. J. Phys. Oceanogr., 28 , 21752184.

  • Andreas, E. L., 2003: An algorithm to predict the turbulent air–sea fluxes in high-wind, spray conditions. Preprints, 12th Conf. on Interaction of Sea and Atmosphere, Long Beach, CA, Amer. Meteor. Soc., CD-ROM, 3.4.

  • Andreas, E. L., 2004: Spray stress revisited. J. Phys. Oceanogr., 34 , 14291440.

  • Andreas, E. L., and J. DeCosmo, 1999: Sea spray production and influence on air–sea heat and moisture fluxes over the open ocean. Air–Sea Exchange: Physics, Chemistry and Dynamics, G. L. Geernaert, Ed., Kluwer, 327–362.

    • Search Google Scholar
    • Export Citation
  • Andreas, E. L., and K. A. Emanuel, 2001: Effects of sea spray on tropical cyclone intensity. J. Atmos. Sci., 58 , 37413751.

  • Andreas, E. L., and J. DeCosmo, 2002: The signature of sea spray in the HEXOS turbulent heat flux data. Bound.-Layer Meteor., 103 , 303333.

    • Search Google Scholar
    • Export Citation
  • Bao, J-W., J. M. Wilczak, J-K. Choi, and L. H. Kantha, 2000: Numerical simulations of air–sea interaction under high wind conditions using a coupled model: A study of hurricane development. Mon. Wea. Rev., 128 , 21902210.

    • Search Google Scholar
    • Export Citation
  • Bender, M. A., and I. Ginis, 2000: Real-case simulations of hurricane–ocean interaction using a high-resolution coupled model: Effects on hurricane intensity. Mon. Wea. Rev., 128 , 917946.

    • Search Google Scholar
    • Export Citation
  • Benoit, R., M. Desgagne, P. Pellerin, Y. Chartier, and S. Desjardins, 1997: The Canadian MC2: A semi-implicit semi-Lagrangian wide-band atmospheric model suited for fine-scale process studies and simulation. Mon. Wea. Rev., 125 , 23822415.

    • Search Google Scholar
    • Export Citation
  • Chalikov, D. V., and V. K. Makin, 1991: Models of the wave boundary layer. Bound.-Layer Meteor., 56 , 8399.

  • Charnock, H., 1955: Wind stress on a water surface. Quart. J. Roy. Meteor. Soc., 81 , 639640.

  • Chouinard, C., J. Mailhot, and A. Mitchell, 1994: The Canadian Regional Data Assimilation System: Operational and research applications. Mon. Wea. Rev., 122 , 13061325.

    • Search Google Scholar
    • Export Citation
  • Davis, C. A., and K. A. Emanuel, 1991: Potential vorticity diagnostics of cyclogenesis. Mon. Wea. Rev., 119 , 19291953.

  • DeCosmo, J., K. B. Katsaros, S. D. Smith, R. J. Anderson, W. A. Oost, K. Bumke, and H. Chadwick, 1996: Air–sea exchange of water vapor and sensible heat: The Humidity Exchange Over the Sea (HEXOS) Results. J. Geophys. Res., 101 , 1200112016.

    • Search Google Scholar
    • Export Citation
  • Desjardins, S., J. Mailhot, and R. Lalbeharry, 2000: Examination of the impact of a coupled atmospheric and ocean wave system. Part I: Atmospheric aspects. J. Phys. Oceanogr., 30 , 385401.

    • Search Google Scholar
    • Export Citation
  • Donelan, M. A., F. W. Dobson, S. D. Smith, and R. J. Anderson, 1993: On the dependence of sea surface roughness on wave development. J. Phys. Oceanogr., 23 , 21432149.

    • Search Google Scholar
    • Export Citation
  • Donelan, M. A., B. K. Haus, N. Reul, W. J. Plant, M. Stiassnie, H. C. Graber, O. B. Brown, and E. S. Saltzman, 2004: On the limiting aerodynamic roughness of the ocean in very strong winds. Geophys. Res. Lett., 31 .L18306, doi:10.1029/2004GL019460.

    • Search Google Scholar
    • Export Citation
  • Doyle, J. D., 1995: Coupled ocean wave/atmosphere mesoscale model simulations of cyclogenesis. Tellus, 47A , 766778.

  • Doyle, J. D., 2002: Coupled atmosphere–ocean wave simulations under high wind conditions. Mon. Wea. Rev., 130 , 30873099.

  • Drennan, W. M., H. C. Graber, D. Hauser, and C. Quentin, 2003: On the wave age dependence of wind stress over pure wind seas. J. Geophys. Res., 108 .8062, doi:10.1029/2000JC000715.

    • Search Google Scholar
    • Export Citation
  • Edson, J. B., S. Anquentin, P. G. Mestayer, and J. F. Sini, 1996: Spray droplet modeling: 2. An interactive Eulerian–Lagrangian model of evaporating spray droplets. J. Geophys. Res., 101 , 12791293.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., 1995: Sensitivity of tropical cyclones to surface exchange coefficients and a revised steady-state model incorporating eye dynamics. J. Atmos. Sci., 52 , 39693976.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., 1999: Thermodynamic control of hurricane intensity. Nature, 401 , 665669.

  • Emanuel, K. A., 2003: A similarity hypothesis for air–sea exchange at extreme wind speeds. J. Atmos. Sci., 60 , 14201428.

  • Fairall, C. W., J. D. Kepert, and G. J. Holland, 1994: The effect of sea spray on the surface energy transports over the ocean. Global Atmos. Ocean Syst., 2 , 121142.

    • Search Google Scholar
    • Export Citation
  • Fairall, C. W., E. F. Bradley, D. P. Rogers, J. B. Edson, and G. S. Young, 1996: Bulk parameterization of air–sea fluxes for Tropical Ocean Global Atmosphere Coupled Ocean–Atmosphere Response Experiment. J. Geophys. Res., 101 , 37473764.

    • Search Google Scholar
    • Export Citation
  • Holton, J. R., 1979: An Introduction to Dynamic Meteorology. Academic Press, 391 pp.

  • Hoskins, B. J., M. E. Mcintyre, and A. W. Robertson, 1985: On the use and significance of isentropic potential vorticity maps. Quart. J. Roy. Meteor. Soc., 101 , 637655.

    • Search Google Scholar
    • Export Citation
  • Huo, Z., D. L. Zhang, and J. Gyakum, 1995: A diagnostic analysis of the superstorm of March 1993. Mon. Wea. Rev., 123 , 17401761.

  • Janssen, P. A. E. M., 1989: Wave-induced stress and the drag of air flow over sea waves. J. Phys. Oceanogr., 19 , 745754.

  • Janssen, P. A. E. M., 1991: Quasi-linear theory of wind-wave generation applied to wave forecasting. J. Phys. Oceanogr., 21 , 16311642.

    • Search Google Scholar
    • Export Citation
  • Janssen, P. A. E. M., 1997: Does wind stress depend on sea-state or not? A statistical error analysis of HEXMAX data. Bound.-Layer Meteor., 83 , 479503.

    • Search Google Scholar
    • Export Citation
  • Janssen, P. A. E. M., J. D. Doyle, J. Bidlot, B. Hansen, L. Isaksen, and P. Viterbo, 2002: Impact and feedback of ocean waves on the atmosphere. Atmosphere–Ocean Interactions, Vol. I, W. Perrie, Ed., Advances in Fluid Mechanics, Vol. 33, WIT Press, 155–198.

    • Search Google Scholar
    • Export Citation
  • Kain, J. S., and J. M. Fritsch, 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.

  • Kepert, J. D., C. W. Fairall, and J-W. Bao, 1999: Modelling the interaction between the atmospheric boundary layer and evaporating sea spray droplets. Air–Sea Exchange: Physics, Chemistry and Dynamics, G. L. Geernaert, Ed., Kluwer Academic, 363–410.

    • Search Google Scholar
    • Export Citation
  • Komen, G. J., L. Cavaleri, M. Donelan, K. Hasselmann, S. Hasselmann, and P. A. E. M. Janssen, 1994: Dynamics and Modeling of Ocean Waves. Cambridge University Press, 532 pp.

    • Search Google Scholar
    • Export Citation
  • Komen, G. J., P. A. E. M. Janssen, V. Makin, and W. Oost, 1998: On the sea state dependence of the Charnock parameter. Global Atmos. Ocean Syst., 5 , 367388.

    • Search Google Scholar
    • Export Citation
  • Kuo, Y-H., R. J. Reed, and S. Low-Nam, 1991: Effects of surface energy fluxes during the early development and rapid intensification stages of seven explosive cyclones in the western Atlantic. Mon. Wea. Rev., 119 , 457476.

    • Search Google Scholar
    • Export Citation
  • Large, W. G., and S. Pond, 1982: Sensible and latent heat flux measurements over the ocean. J. Phys. Oceanogr., 12 , 464482.

  • Lionello, P., P. Malguzzi, and A. Buzzi, 1998: Coupling between the atmospheric circulation and ocean wave field: An idealized case. J. Phys. Oceanogr., 28 , 161177.

    • Search Google Scholar
    • Export Citation
  • Liu, W. T., K. B. Katsaros, and J. A. Businger, 1979: Bulk parameterization of air–sea exchanges of heat and water vapor including the molecular constraints at the interface. J. Atmos. Sci., 36 , 17221735.

    • Search Google Scholar
    • Export Citation
  • Makin, V. K., 1998: Air–sea exchange of heat in the presence of wind waves and spray. J. Geophys. Res., 103 , 11371152.

  • Makin, V. K., and C. Mastenbroek, 1996: Impact of waves on air–sea exchange of sensible heat and momentum. Bound.-Layer Meteor., 79 , 279300.

    • Search Google Scholar
    • Export Citation
  • Meirink, J. F., and V. K. Makin, 2001: The impact of sea spray evaporation in a numerical weather prediction model. J. Atmos. Sci., 58 , 36263638.

    • Search Google Scholar
    • Export Citation
  • McTaggart-Cowan, R., J. R. Gyakum, and M. K. Yau, 2001: Sensitivity testing of extratropical transitions using potential vorticity inversions to modify initial conditions: Hurricane Earl case study. Mon. Wea. Rev., 129 , 16171636.

    • Search Google Scholar
    • Export Citation
  • Moon, I-J., I. Ginis, and T. Hara, 2004: Effect of surface waves on air–sea momentum exchange. Part II: Behavior of drag coefficient under tropical cyclones. J. Atmos. Sci., 61 , 23342348.

    • Search Google Scholar
    • Export Citation
  • Perrie, W., and L. M. Wang, 1995: A coupling mechanism for wind and waves. J. Phys. Oceanogr., 25 , 615630.

  • Powell, M. D., P. J. Vickery, and T. A. Reinhold, 2003: Reduced drag coefficient for high wind speeds in tropical cyclones. Nature, 422 , 279283.

    • Search Google Scholar
    • Export Citation
  • Smith, S. D., and Coauthors, 1992: Sea surface wind stress and drag coefficient: The HEXOS results. Bound.-Layer Meteor., 60 , 109142.

    • Search Google Scholar
    • Export Citation
  • Stoelinga, M. A., 1996: A potential vorticity-based study of the role of diabatic heating and friction in a numerically simulated baroclinic cyclone. Mon. Wea. Rev., 124 , 849874.

    • Search Google Scholar
    • Export Citation
  • Tolman, H. L., 2002: Alleviating the garden sprinkler effect in wind wave models. Ocean Modell., 4 , 269289.

  • Tolman, H. L., and D. V. Chalikov, 1996: Source terms in a third-generation wind wave model. J. Phys. Oceanogr., 26 , 24972518.

  • Wang, Y., J. D. Kepert, and G. J. Holland, 2001: The impact of sea spray evaporation on tropical cyclone boundary layer structure and intensity. Mon. Wea. Rev., 129 , 24812500.

    • Search Google Scholar
    • Export Citation
  • Zhang, Y., and W. Perrie, 2001: Feedback mechanisms for the atmosphere and ocean surface. Bound.-Layer Meteor., 100 , 321348.

  • View in gallery

    Comparison of the observed neutral drag coefficient Cd (circles, diamonds, squares, and triangles) from Powell et al. (2003) (vertical bars for 95% confidence), Large and Pond (1982) (light dashed line), Donelan et al. (2004), and Cd from the coupled model results from the MC2 grouped according to different wave ages [Eq. (5)], with the restriction that when U10 > 30 m s−1, we limit Z0 = 0.0034. Boldface dashed line and light solid line indicate Cd associated with young and fully developed sea states following Smith et al. (1992) and Drennan et al. (2003), respectively.

  • View in gallery

    Storm tracks of (a) Earl, (b) the bomb, and (c) the superbomb, using MC2 with and without spray and waves, as well as NHC and CMC analyses. Storm centers are plotted every 6 h. Simulations for Earl began at 0000 UTC 5 Sep 1998, for the bomb at 1200 UTC 12 Jan 2002, and for the superbomb at 1200 UTC 20 Jan 2000.

  • View in gallery

    (a) Control simulation and (b) QuikSCAT–NCEP blended wind fields, during maximum winds for the superbomb (0600 UTC 21 Jan). Units are kt.

  • View in gallery

    Minimum SLP and U10 time series for (a), (d) Earl, (b), (e) the bomb, and (c), (f) the superbomb, following the storm tracks, respectively. The Charnock parameter (dashed line with triangles) is indicated in (a)–(c). Here, U10 is averaged on an area of 200 km2 over each storm’s high-wind regions.

  • View in gallery

    Differences ΔSLP (hPa; contour) and ΔU10 (m s−1; shaded) at Earl’s peak intensity (1200 UTC 6 Sep 1998) for (a) MC2-spray minus control and (b) MC2-wave minus control. In (a) [(b)] ΔSLP contour starts at −1 (+1) mb, with ΔSLP intervals at −0.5 (+1) mb. Simulation winds U10 with (a) spray and (b) wave are superposed. Storm centers (cross-filled circles) are shown. For the bomb (0600 UTC 14 Jan 2002): (c) MC2-spray minus control and (d) MC2-wave minus control. In (c) [(d)] ΔSLP contour starts at −1 (+1) mb, with intervals at −1 (+1) mb. For the superbomb (1800 UTC 21 Jan 2000): (e) MC2-spray minus control and (f) MC2-wave minus control. In (e) [(f)] ΔSLP contour starts at −1 (+1) mb, with intervals at −2 (+1) mb.

  • View in gallery

    The Charnock parameter (shaded) and the wind field (arrows) from MC2-wave, with the same simulation hour as in Fig. 4, for (a) Earl at 1200 UTC 6 Sep 1998, (b) the bomb at 0600 UTC 14 Jan 2002, and (c) the superbomb at 1800 UTC 21 Jan 2000. Storm centers are shown with a times sign.

  • View in gallery

    Time series for 200 × 200 km2 area-averaged sensible (bottom four plotted time series) and latent (top four plotted time series) following maximal flux center storm with and without spray and waves, for (a) Earl, (b) the bomb, and (c) the superbomb. The difference of specific humidity q between the MC2-wave and MC2 runs are given for (d) Earl at 0600 UTC 6 Sep 1998, (e) the bomb at 0000 UTC 14 Jan 2002, and (f) the superbomb at 0000 UTC 21 Jan 2000.

  • View in gallery

    Trajectories of the storm center and latent heat flux center from the fully coupled simulation compared with the center for the difference of the LH flux, between the fully coupled minus MC2-only control simulations for (a) the bomb and (b) the superbomb. During the period 1800 UTC 21 Jan–1200 UTC 22 Jan 2000 the LH center for the superbomb does not move.

  • View in gallery

    (a) Total heat flux difference (ΔSH + ΔLH; shaded areas; W m−2) between the fully coupled (MC2-wave-spray) minus control (MC2-only) simulations at 0000 UTC 14 Jan 2002, superposed on control simulation heights at 850 hPa (thin solid lines) and wind speed at 1000 hPa. (b) Vertical cross section [for transect in (a)] for differences in potential temperature (shaded; negative Δθ with thick dashed contour), height (thin lines; dam), and along-plane vector flow (arrows). Associated results during the filling period at 1200 UTC 14 Jan 2002 are given in (c), (d). Storm center is shown with a cross-filled circle.

  • View in gallery

    As in Fig. 9 but for the superbomb. (a) Total heat flux difference (ΔSH + ΔLH; shaded areas; W m−2) for MC2-spray minus MC2-only control simulations at 0600 UTC 21 Jan 2000, superposed with control simulation of heights at 850 hPa (thin solid lines) and wind speed at 1000 hPa. (b) Height–latitude cross sections [for transect in (a)] showing Δθ (shaded; K), (c) Δq (g kg−1) and along-plane vector flow (arrows), and (d) cumulative precipitation (mm h−1). Control simulation contours for θ, q, and storm center (cross filled circle along the x axis) are indicated.

  • View in gallery

    As in Fig. 10 (MC2-spray minus MC2 only) for the superbomb at 0600 UTC 22 Jan 2000 during the superbomb’s filling phase. (a) Total ΔSH + ΔLH differences (shaded; W m−2), superposed with control simulation of heights at 850 hPa (thin solid lines) and wind speed at 1000 hPa. (b) Height–latitude cross sections [for transect in (a)] showing Δθ (K) differences.

  • View in gallery

    As in Figs. 10a–c (MC2-spray minus MC2 only) for Earl: (a)–(c) at 1800 UTC 5 Sep 1998 during the intensifying period and (d)–(e) at 1200 UTC 6 Sep 1998 during the filling period. Vertical cross sections are through Earl’s storm center and high-wind region. Dashed contour lines in (a) and (d) indicate total heat flux from MC2-only control run.

  • View in gallery

    The difference of SST minus screen temperature Ta (2 m) in the MC2-only control simulation for (a) Earl at 1800 UTC 5 Sep 1998 and (b) for the superbomb at 0600 UTC 21 Jan 2000. The storm center is shown with a times sign.

  • View in gallery

    Difference between MC2-wave minus MC2-only control simulations for (a) Δθ (K) and (b) Δq (1.0 × 10−3 kg kg−1). The time is 0600 UTC 21 Jan 2000 when winds and latent heat flux are maximal. The control simulation is as shown in Figs. 10b and 10c, and transect line is as show in Fig. 10a.

  • View in gallery

    Vertical motion (m s−1; positive is ascending motion) from (a) MC2-only control and (b) the difference between MC2-wave minus MC2-only simulations. The time is 0600 UTC 21 Jan 2000 when winds and latent heat flux are maximal. The transect line is as shown in Fig. 10a.

  • View in gallery

    As in Figs. 9a and 9b for Earl: (a), (b) differences between the fully coupled and control simulations (MC2-wave-spray minus MC2 only), (c) showing difference of Δq and along-plane vector flows at 1800 UTC 5 Sep 1998, and (d)–(f) at 1200 UTC 6 Sep 1998. Control simulations are shown in Figs. 12. Transect sections go through the storm center and the high-wind region, as indicated by solid black lines in (a) and (d).

  • View in gallery

    As in Fig. 16 but for the superbomb: (a)–(c) at 0600 UTC 21 Jan 2000 and (d)–(f) at 1800 UTC 21 Jan 2000.

  • View in gallery

    Simulated PV (PVU; 1 PVU = 10–6 m2 K kg−1 s−1; thick line) and θ (K; thin line) from MC2 only for (a) 1000 hPa and (b) vertical cross section for the superbomb [transect line in (c)] at 0600 UTC 21 Jan 2000. Differential PV (shaded areas) between MC2-wave and MC2 only are shown at (c) 1000 and (d) 900 hPa, with θ (thin line) superposed and wind vectors (arrows). The vertical PV difference cross section [transect line in (c)] between MC2-wave minus MC2 only for (e) and for (f) between MC2-spray-wave and MC2 only with 0.3-PVU intervals and where the thick solid line is the zero PV contour.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 44 42 5
PDF Downloads 21 21 5

Impacts of Waves and Sea Spray on Midlatitude Storm Structure and Intensity

View More View Less
  • 1 Fisheries and Oceans Canada, Bedford Institute of Oceanography, Dartmouth, and Department of Engineering Mathematics, Dalhousie University, Halifax, Nova Scotia, Canada
  • | 2 Department of Atmospheric Sciences, Zhongshan University, Guangzhou, China
© Get Permissions
Full access

Abstract

A coupled atmosphere–wave–sea spray model system is used to evaluate the combined impacts of spray evaporation and wave drag on midlatitude storms. The focus of this paper is on the role of air–sea fluxes on storm intensity and development, and related impacts on the structure of the atmospheric boundary layer. The composite model system consists of the Canadian Mesoscale Compressible Community atmospheric model coupled to the operational wave model WAVEWATCH III, and a recent bulk parameterization for heat fluxes due to sea spray. The case studies are extratropical Hurricane Earl (in 1998) and two intense winter storms from 2000 and 2002, hereafter denoted “superbomb” and “bomb,” respectively. The results show that sea spray tends to intensify storms, whereas wave-related drag tends to weaken storms. The mechanisms by which spray and wave-related drag can influence storm intensity are quite different. When wind speeds are high and sea surface temperatures warm, spray can significantly increase the surface heat fluxes. By comparison, momentum fluxes related to wave drag are important over regions of the storm where young, newly generated waves are prevalent, for example during the rapid development phase of the storm. These momentum fluxes decrease in areas where the storm waves reach maturity. The collective influence of spray and waves on storm intensity depends on their occurrence in the early stages of a storm’s rapid intensification phase, and their spatial distribution with respect to the storm center. Moreover, for the case of the superbomb, a potential vorticity framework is used to show the relative importance of these surface flux impacts compared with baroclinic processes.

Corresponding author address: Dr. W. Perrie, Bedford Institute of Oceanography, 1 Challenger Dr., Dartmouth NS B2Y 4A2, Canada. Email: perriew@dfo-mpo.gc.ca

Abstract

A coupled atmosphere–wave–sea spray model system is used to evaluate the combined impacts of spray evaporation and wave drag on midlatitude storms. The focus of this paper is on the role of air–sea fluxes on storm intensity and development, and related impacts on the structure of the atmospheric boundary layer. The composite model system consists of the Canadian Mesoscale Compressible Community atmospheric model coupled to the operational wave model WAVEWATCH III, and a recent bulk parameterization for heat fluxes due to sea spray. The case studies are extratropical Hurricane Earl (in 1998) and two intense winter storms from 2000 and 2002, hereafter denoted “superbomb” and “bomb,” respectively. The results show that sea spray tends to intensify storms, whereas wave-related drag tends to weaken storms. The mechanisms by which spray and wave-related drag can influence storm intensity are quite different. When wind speeds are high and sea surface temperatures warm, spray can significantly increase the surface heat fluxes. By comparison, momentum fluxes related to wave drag are important over regions of the storm where young, newly generated waves are prevalent, for example during the rapid development phase of the storm. These momentum fluxes decrease in areas where the storm waves reach maturity. The collective influence of spray and waves on storm intensity depends on their occurrence in the early stages of a storm’s rapid intensification phase, and their spatial distribution with respect to the storm center. Moreover, for the case of the superbomb, a potential vorticity framework is used to show the relative importance of these surface flux impacts compared with baroclinic processes.

Corresponding author address: Dr. W. Perrie, Bedford Institute of Oceanography, 1 Challenger Dr., Dartmouth NS B2Y 4A2, Canada. Email: perriew@dfo-mpo.gc.ca

1. Introduction

Tropical storm intensity is sensitive to the air–sea transfer rates of enthalpy and momentum, particularly in the storm’s high-wind core. Emanuel (1995, 1999) showed that if values of the exchange coefficients at 20 m s−1 are applied at higher wind speeds, maintaining a storm of greater than marginal hurricane intensity would be impossible. Thus, Andreas and Emanuel (2001) suggest that sea spray is a plausible mechanism to enhance the air–sea enthalpy exchange at high wind speeds. Strong winds eject large amounts of spray into the lower-atmospheric boundary layer. Heat fluxes can be significantly modified through spray droplets evaporation, and can ultimately affect cyclone dynamics. Fairall et al. (1994) and Kepert et al. (1999) suggest that spray droplet evaporation is needed, if the boundary layers of simulated tropical cyclones are to realistically evolve. Bao et al. (2000) used a coupled atmosphere–ocean–wave–spray model to show that sea spray can increase the intensity of Hurricane Opal simulations. Wang et al. (2001) showed that the boundary layer structure and convection in the cyclone’s near-core region are modified because of surface layer cooling from spray evaporation.

A competing mechanism is the impact of wind waves on air–sea momentum fluxes, as shown by observational and modeling studies (Janssen 1989, 1997; Chalikov and Makin 1991; Smith et al. 1992; Donelan et al. 1993). Waves can impact on the surface stress, contributing to the decay of the cyclone (Doyle 1995; Lionello et al. 1998; Bao et al. 2000; Desjardins et al. 2000; Doyle 2002). Smith et al. (1992) and Donelan et al. (1993) suggest that the roughness coefficient Z0 depends on the inverse wave age. However, no measurable wave age dependence has been observed for heat and moisture flux coefficients (DeCosmo et al. 1996).

Although studies of the impact of wave-induced drag and sea spray as separate factors on air–sea fluxes have made recent progress, their collective impact has received less attention. Makin and Mastenbroek (1996) suggest that while momentum flux is largely transported by wave-induced motion, heat flux is transported only by viscosity. Therefore, air–sea fluxes of sensible and latent heat are determined by the turbulent diffusivity, and only indirectly affected by waves. This supports the experimental result that the wind dependence of air–sea sensible heat and humidity exchange coefficients is much less important than the drag coefficient’s wind dependence. However, waves affect sensible heat and humidity fluxes through white-capping and spray. Spray droplets evaporate and change the boundary layer balance of moisture and sensible heat. Laboratory, numerical, and open-ocean studies all show that spray can redistribute enthalpy between the boundary layer temperature and humidity fields (Edson et al. 1996; Andreas 1992; Andreas and DeCosmo 2002). Using a one-dimensional stratified model, Makin (1998) showed that, for 10-m winds U10 above 25 m s−1, the spray impacts on heat and moisture fluxes depend on the atmospheric stratification. Moreover, Andreas and Emanuel (2001) suggest that the inclusion of wave drag and spray can result in acceptable hurricane intensity simulations.

Although recent studies have focused on the effects of sea spray on tropical storms, not only because of extreme wind speeds, but also because of the importance of the air–sea fluxes on their dynamics, comprehensive studies of the impact of waves and spray on midlatitude storms are less widespread. Meirink and Makin (2001) assessed the sensitivity of a high-resolution limited-area atmospheric model to spray evaporation over the North Sea. They reported that spray results in a significantly cooler and moister surface layer and indirectly enhances precipitation and the associated midlevel latent heat release. Janssen et al. (2002) suggested that enhanced wave-induced surface roughness may intensify a cyclone, because the heat fluxes may be enhanced due to vortex stretching, thereby deepening the low. Adamson et al. (2006) used idealized dry baroclinic life cycle simulations with a primitive equation model to investigate the mechanisms by which surface drag influences interior cyclone development. Using a potential vorticity analysis, they suggest that although surface drag is related to Ekman pumping in cyclone simulations, baroclinic processes along a warm front are evident, and that the latter are dominant.

In this paper, we construct a composite atmosphere–wave–spray model, consisting of a mesoscale weather model, and models for waves and sea spray. Case studies are extratropical Hurricane Earl (in 1998) and two intense winter storms: the “bomb” of 12–15 January 2002 and the “superbomb” of 20–22 January 2000. Our focus is on evaluating the combined impacts of spray and waves on midlatitude storm intensity and the structure of the lower atmosphere. The numerical models used are described in section 2, and the storm cases in section 3. Section 4 discusses the impacts of surface fluxes on storm track and intensity. In section 5 impacts on the structure of the lower part of atmosphere are discussed. Conclusions are given in section 6.

2. Model description and experimental design

The Mesoscale Compressible Community (MC2) atmospheric model, version 4.9.3, is used in all numerical simulations and is coupled to the WAVEWATCH III (WW3) ocean wave model and a bulk algorithm for turbulent air–sea fluxes, with a high-wind sea spray formulation.

a. Atmospheric model

The MC2 model is a nonhydrostatic, fully elastic, state-of-the-art model, solving the full Euler equations on a limited-area Cartesian domain (Benoit et al. 1997). Semi-Lagrangian advection and a semi-implicit time-differencing dynamic scheme are used, allowing MC2 to achieve excellent storm simulations over a wide range of conditions, ranging from extratropical storms (McTaggart-Cowan et al. 2001), to meso-γ nonhydrostatic studies (Benoit et al. 1997), comparable to other meteorological models. The model is set up on a latitude–longitude grid, on the domain 25°–58°N and 40°–80°W, with 30 vertical layers (11 in the boundary layer), a lowest-level depth of 18 m, 0.25° horizontal resolution, and a 600-s integration time step. Analysis data (Chouinard et al. 1994) from the Canadian Meteorological Center (CMC) are used to set the initial conditions. Surface heat and moisture fluxes over land are calculated from a force–restore scheme (Benoit et al. 1997). Surface fluxes above the sea use Monin–Obukhov similarity theory. The Kain–Fritsch scheme (Kain and Fritsch 1993) is used for deep cumulus convection, as well as related boundary layer, turbulent kinetic energy, and vertical diffusion parameterizations (Benoit et al. 1997). Boundary conditions at each time step are obtained from a 6-h nesting interval and interpolated linearly between analyses, along with weekly mean SSTs.

b. Ocean wave model

The WW3 wave model (Tolman and Chalikov 1996; Tolman 2002) was implemented on the same domain as the MC2 model, also with 0.25° horizontal resolution. The model simulates directional wave spectra in terms of wavenumber-direction bands by solving the well-known spectral action balance equation, and explicitly accounting for the energy input to waves by the wind as given by Sin, wave dissipation Sds, and wave–wave interactions Snl. All simulations assume Wave Action Model cycle 3 formulations for Sin and Sds (Komen et al. 1994).

c. Air–sea fluxes

Heat, moisture, and momentum can be transferred across the air–sea interface through the interfacial or turbulent fluxes, expressed as bulk turbulent flux parameterizations, and by sea spray. The MC2 model uses Monin–Obukhov similarity theory for the interfacial fluxes of momentum τ, sensible heat Hs, and latent heat HL:
i1520-0493-134-9-2418-e1
i1520-0493-134-9-2418-e2
i1520-0493-134-9-2418-e3
Here U, θ, and q are the mean wind speed, potential temperature, and specific humidity, respectively; ρa is the air density; cpa the specific heat of air at constant pressure; and Lυ the latent heat of the vaporization of water. Subscripts zl and o are the lowest atmospheric model level and the ocean surface, respectively. Positive heat fluxes are upward.
In uncoupled simulations, the atmospheric model parameterizes the momentum roughness length Z0m by the Charnock (1955) relation over the open sea:
i1520-0493-134-9-2418-e4
where the Charnock parameter β is 0.018, which represents a mature sea. In coupled simulations, described in detail in the following sections, Z0m is sea state dependent, following Smith et al.’s (1992) suggestion from Humidity Exchange Over the Sea (HEXOS) data that for younger waves,
i1520-0493-134-9-2418-e5
where β in Eq. (4) is generalized to include wave age CP/u*, and CP is the phase velocity at the wave spectrum peak, g is gravity, and u* is friction velocity. As waves mature, CP/u* increases and Z0m decreases. Desjardins et al. (2000), Zhang and Perrie (2001), and Janssen et al. (2002) show that Eq. (5) gives similar results to those of the wave-induced stress formulation of Janssen (1991) and Komen et al. (1998). In the limit when Cp/u* ≥ 26, the Charnock Z0m relation is imposed on Eq. (5). Motivated by Emanuel (2003), we set the limit Z0m = 0.0034 m as a high-wind constraint, when U10 > 30 m s−1. Figure 1 shows that for conditions experienced during the superbomb (described below), the Z0m formulation in Eq. (5), with Charnock and the high-wind constraint, implies a drag coefficient variation with U10 that is consistent with hurricane observations (Powell et al. 2003), theoretical estimates (Moon et al. 2004), and observations (Smith et al.1992; Drennan et al. 2003). When the high-wind condition is Z0m = 0.005 m when U10 > 30 m s−1, peak drag coefficient values increase very slightly (not shown), and are consistent with high-wind speed laboratory data of Donelan et al. (2004).

Roughness lengths for thermal fluxes Z0t and Z0q are simply fixed (4 × 10−5 m) in MC2 simulations, because there is no consensus that enthalpy transfer has sea state dependence, particularly under high-wind conditions.

d. Sea spray

Over the sea, MC2’s interfacial momentum and heat fluxes are calculated using Monin–Obukhov theory, leading to a bulk formulation based on turbulent transfer coefficients. These depend on empirical similarity functions ψm and ψh, and roughness lengths for wind speed, temperature, and humidity: Z0m, Z0t, and Z0q, respectively. However, in Andreas’s (2003) bulk spray flux algorithm, care is needed regarding the parameterizations for ψm, ψh, Z0m, Z0t, and Z0q, because the flux algorithm is derived by subtracting estimates of the interfacial heat fluxes from HEXOS measurements of the total heat fluxes. Andreas (2003) used the interfacial heat flux parameterization of Liu et al. (1979), which is the basis for the Coupled Ocean–Atmosphere Response Experiment (COARE), version 2.6, algorithm (Fairall et al. 1996). Thus, the spray flux is the residual and is specifically tuned to this COARE-type parameterization. While MC2 uses Monin–Obukhov theory to estimate interfacial fluxes, the spray algorithm estimates interfacial fluxes based on COARE-type parameterizations.

Total momentum τT, latent HL,T, and sensible HS,T fluxes, constituting the boundary conditions at the lowest model level, are obtained by adding the corresponding bulk interfacial (τ, HL, HS) and spray fluxes (τsp, QL,sp, QS,sp), following Andreas and DeCosmo (1999), Andreas and Emanuel (2001), Andreas (2003)
i1520-0493-134-9-2418-e6
i1520-0493-134-9-2418-e7
i1520-0493-134-9-2418-e8
Andreas’s (1992) spray flux model suggests that droplets whose initial radii are 100 μm carry most of the spray sensible heat, and droplets with 50-μm initial radii carry most of the spray latent heat (Andreas 1992, 1998; Andreas 1995). Andreas (2003) therefore assumed that these droplets are the bellwethers of the respective spray fluxes and modeled the heat fluxes as
i1520-0493-134-9-2418-e9
i1520-0493-134-9-2418-e10
where ρw is seawater density, cw the seawater specific heat, Teq,100 the equilibrium temperature of spray droplets with 100 μm initial radius (Andreas 1995; Kepert et al. 1999), and req,50 the equilibrium radius of droplets with 50-μm initial radius. Wind functions VS and VL depend on u* and tune Eqs. (9) and (10) to the HEXOS data (Andreas and DeCosmo 2002; Andreas 2003), yielding
i1520-0493-134-9-2418-e11
i1520-0493-134-9-2418-e12
The units for QS,sp, QL,sp, HS, and HL are watts per square meter, when u* is in units of meters per second. Spray-mediated fluxes, QS,sp and QL,sp, increase much more rapidly with U10 than do the interfacial fluxes; for example, when U10 exceeds 30 m s−1, QL,sp and QS,sp exceed HL and HS, respectively.
When spray droplets are ejected into the air, they quickly accelerate to the local wind speed, extracting momentum from the wind and slowing it. Falling back to the sea, they transfer momentum, affecting the wind speed profile, the atmospheric momentum profile, and the surface stress, which may be estimated as (Andreas 1998; Andreas and Emanuel 2001; Andreas 2004)
i1520-0493-134-9-2418-e13
Andreas and Emanuel (2001) show that because (13) goes as the fourth power of u*, it becomes larger than the interfacial stress (i.e., ρau2*) when u* is about 4 m s−1 or U10 is larger than about 65 m s−1. As our simulations feature wind speeds that are rarely above 40 m s−1, the spray stress is negligible.

e. Coupling formulation and experimental design

The coupling time step is 1800 s, because the model time steps for MC2 and WW3 are 600 and 900 s, respectively. Between coupling time steps, the Charnock parameter β in MC2 remains the same. At every coupling time step, information is exchanged between the atmosphere and the waves: wind speed and direction computed by MC2 are transferred to the wave model WW3, and in turn, CP computed by WW3 is passed to MC2 allowing for the computation of new Z0m values. Spray-mediated heat fluxes are passed back to MC2 at each MC2 time step. An iterative relaxation is included to produce an equilibrium state among the MC2 boundary layer winds, WW3 waves, and Z0m following Perrie and Wang (1995). This allows consistent values for u* and Z0m defined within both MC2 and WW3.

Four experiments were performed to study the wave drag and sea spray impacts on storms (Table 1). The control simulation uses MC2 atmospheric model winds to drive the WW3 waves, assuming the conventional Charnock roughness [Eq. (4)], with no MC2 feedback to MC2, and no spray (one-way coupling). The fully coupled (two-way coupled) MC2-wave–spray simulation includes spray-enhanced heat and momentum fluxes, and wave-modified stress feedbacks to MC2. Two partially coupled simulations study wave drag and spray competition: (a) a coupled MC2-wave run with wave-modified β [Eq. (5)] passed to MC2, but no spray-modified fluxes, and (b) a coupled MC2-spray run, driving WW3 waves, with no wave-modified β passed to MC2.

3. Storm cases

Tropical Storm Earl reached hurricane status on 2 September 1998 when it was 230 km south-southwest of New Orleans, rapidly intensified to category 2 (winds in excess of 42 m s−1), and made landfall near Panama City, Florida, on 3 September. Moving toward Georgia, it quickly weakened, becoming extratropical. It crossed the Carolinas, passed near Cape Hatteras, followed a northeastward trajectory, and intensified over the Maritime Provinces of eastern Canada. On 6 September, Earl crossed the Avalon Peninsula of Newfoundland and continued northeastward. On 8 September it was absorbed by the remnants of Hurricane Danielle, as it neared Ireland.

The 2002 January bomb followed the development pattern typical of North Atlantic winter storms. It originated off the coast of North Carolina at 1200 UTC 13 January 2002, as a rapidly intensifying low pressure system. It continued to deepen over the next 12 h, and moved northeastward, generating high winds and 18-m seas at National Data Buoy Center buoys in the Gulf of Maine and off eastern Canada. Nearing Nova Scotia at 0000 UTC 14 January, the maximum sustained winds were 60 kt (1 kt = 0.5144 m s−1), and its central sea level pressure (SLP) fell to 962 hPa. It continued northeastward, crossed eastern Nova Scotia and Newfoundland, and dissipated by 15 January.

The superbomb is similar to the bomb in that it developed off Cape Hatteras, and deepened explosively from 995 mb at 1200 UTC 20 January 2000 to 951 mb by 1200 UTC 21 January. At midlevels, CMC analysis suggests that a short wave traveled quickly around the base of a deepening larger-scale trough over Hudson Bay. The system was under the influence of a jet to its south and west, benefiting from the divergent upper-level forcing associated with the left-exit region. Phase locking occurred, and the vertically stacked system curled northward under the influence of midlevel flow. Propagating northeastward the superbomb’s peak U10 winds reached 45 m s−1 near Nova Scotia. It made landfall at 0000 UTC 22 January and continued weakening.

4. Surface fluxes, storm tracks, and intensity

Sea spray enhances heat fluxes and tends to intensify storms. Studies by Kuo et al. (1991) suggest that the occurrence of surface fluxes during and preceding a storm’s rapid intensification phase strongly influences development, whereas there is little impact if these fluxes occur later. By comparison, ocean waves are continuously generated by winds, enhance surface roughness, and tend to decrease the storm intensity (Desjardins et al. 2000; Doyle 2002).

a. Storm tracks

Figures 2a–c compare the modeled storm tracks of Earl, the bomb, and the superbomb, using the MC2 (with and without spray and waves), with the analysis. Modeled Earl tracks (Fig. 2a) decelerate over the course of the simulations, consistent with the analyzed National Hurricane Center (NHC) storm track. A slight initial displacement of Earl’s simulated low center at 0000 UTC 5 September results from analysis discrepancies between the NHC and CMC (used to initialize the MC2) low positions at that time. Although simulations of the bomb and superbomb tracks (Figs. 2b,c) are systematically biased to the right of the CMC analysis, storm propagation speeds are in overall agreement, capturing the basic evolution of the storms’ tracks and, moreover, showing little sensitivity to spray or waves.

As further verification that MC2 gives good baseline simulations of storm tracks and storm distributions, Figs. 3a,b compare the MC2 control winds (no spray or wave feedbacks) with the blended Quick Scatterometer (QuikSCAT)–National Centers for Environmental Prediction (NCEP) winds at the superbomb’s peak, 0600 UTC 21 January 2000. This shows that the overall structures of the MC2 control and the analysis winds are similar. Maximum QuikSCAT–NCEP winds are 42 m s−1, compared with control winds of 40 m s−1. Maximum U10 is at 38°N, 68°W in both the control and QuikSCAT–NCEP data. The asymmetric character of the wind structure due to the midlatitude baroclinic environment is evident.

b. Storm intensity

Figures 4a–c show the time series of central SLP from simulations with and without sea spray and waves, following the storm tracks of Earl, the bomb, and the superbomb. Spray results in intensification of the three storms: SLP deepening is about 1.2, 2.8, and 5.0 hPa at the peak of the respective storms, compared with the control simulation. By comparison, the combined impact of the spray and waves in the fully coupled MC2-waves-spray simulation lessens Earl’s peak intensity by 1.5 hPa, but deepens the bomb’s and the superbomb’s peak intensity by 1.3 and 2.5 hPa, respectively, relative to the control runs. Therefore, wave drag slightly dominates over spray in Earl, whereas spray is slightly dominant in the other two storms.

Maximum winds U10 are given in Figs. 4d–f, following each storm’s trajectory. For Earl, the bomb, and the superbomb, the maximum sea spray (positive) impacts on winds are 2, 7, and 10 kt, and the wave (negative) impacts are 5, 6, and 7 kt, respectively, compared with the control. In the fully coupled run, the maximum combined impacts of spray and waves are −3, +4, and +6 kt relative to the control, respectively. In each case, both the maximum deintensifying impact of waves and maximum intensifying impact of spray tend to become significant near the period of minimum SLPs, during the rapid intensification phase of the storm, when rough young waves and spray droplets abound, as represented in Figs. 4a–c by the Charnock parameter β [Eq. (4)]. Thereafter, as storms begin to attenuate, spray and wave impacts decrease and tend to balance.

The corresponding spatial distributions of the intensifying or deintensifying impacts of spray and waves on U10 and SLP are given in Figs. 5a–f, at the peak of each storm (defined by Figs. 4a–c). As suggested in Fig. 3, the storm structures are quite asymmetric, and the resulting spatial distributions of differences ΔU10 and ΔSLP are also quite asymmetric, whether mediated by spray (Figs. 5a,c,e) or by waves (Figs. 5b,d,f). The competition between spray and waves is also clear: to intensify or reduce storm intensity. As discussed later (Figs. 7 and 8), important factors in these processes are the magnitudes and spatial distributions of the spray- and wave-mediated impacts on heat and momentum fluxes, relative to the storm centers’ locations.

For Earl (Figs. 5a,b), the spray-enhanced ΔU10 values are generally rather weak (∼1–2 m s−1) and extend over a spiral-shaped high-wind region, and the ΔSLP values occur in the high-wind area west of Earl’s center. By comparison, larger wave-depressed ΔU10 (∼3 m s−1) values are smeared beyond the spray-enhanced ΔU10 areas (Fig. 5b), reflecting the tendency of the waves to integrate the U10 effects, while corresponding wave-depressed ΔSLP areas occur near Earl’s center, deintensifying the cyclone. For the bomb (Figs. 5c,d), weak (∼1–2 m s−1) spray-enhanced ΔU10 values are smeared over the high-wind areas, and spray impacts on ΔSLP are about the same magnitude as those of Earl, although more widely distributed, while stronger (∼3–5 m s−1) wave-depressed ΔU10 values cover an area similar to the spray-enhanced ΔU10 area. For the superbomb (Figs. 5e,f), stronger (∼2–4 m s−1) spray-enhanced ΔU10 values cover a widely distributed area, including high-wind areas, with peak values close to the storm center, and are almost collocated with the areas of maximum spray impacts on ΔSLP, whereas comparable wave-depressed impacts on ΔU10 cover an area similar to the spray-enhanced ΔU10 area. While peak MC2 control winds overestimate in situ buoy data by up to 5 m s−1, the MC2-wave results are improved by as much as 3 m s−1 (not shown). Similar results are obtained in comparing the observed and simulated turbulent kinetic energy results, suggesting that the model is responsive to wave-induced damping and does not excessively damp surface-induced perturbations.

To summarize Figs. 5a–f, spray and wave drag impacts are different because of their generating mechanisms. Maximum spray impacts occur near each storm’s high-wind side, where spray droplets abound. Maximum wave impacts occur in rapidly varying wind areas with roughened sea surfaces and high β, typically in right spiral bands (Figs. 6a–c), where winds are maximal and storm propagation gives an enhanced effective fetch (Doyle 2002). As discussed below, enhanced surface roughness and kinetic energy loss due to wave drag can induce dynamic compensation downdrafts contributing to downward mixing of upper-level dry cool air. This increases the lower-tropospheric static stability, suppresses convection, and reduces storm intensity.

c. Surface flux effects

The direct impact of sea spray and wave drag is to modify the heat fluxes and momentum flux across the air–sea interface (Andreas and Emanuel 2001; Andreas 2004). Section 5 will show that the dynamic compensation downdraft and entrainment associated with wave drag result in a cooler drier boundary layer. This is partially due to the fixed SST fields used in our simulations. Fixed SSTs, in addition to cooler and drier boundary layers, will increase the difference of the temperature and moisture (ΔT and Δq) between the air and sea. Thus, the wave drag’s combined effects, namely reduced wind speed (Figs. 4d–f) and increased ΔT and Δq, tend to result in small changes in surface heat fluxes. This is a thermal–dynamic effect resulting from the bulk formulations [Eqs. (2) and (3)], where the roughness lengths for thermal fluxes Z0t and Z0q are simply held constant. Accordingly, Figs. 7a–c show that the latent heat fluxes from the MC2-wave simulations tend to be only slightly larger than those from MC2 only. This occurs because wave drag results in a slightly drier boundary structure and is particularly evident when the difference in the latent heat flux between the two simulations is maximal, as is shown in Figs. 7d–f.

The impact of spray is variable, depending on the storm conditions, such as the SST, moisture, and wind field distributions. For example, Earl propagates quickly during the initial phase of its extratropical intensification (Fig. 2a), passing over the relatively warm Gulf Stream, and decelerating markedly over cold waters north of Newfoundland. Thus, the spray’s peak impact on Earl’s latent heat (Fig. 7a) is only about 30 W m−2 (5%), compared with 16% (Fig. 7b) for the bomb and 30% (Fig. 7c) for the superbomb, reflecting both their slow propagation over warm midlatitude waters, during their intensification phases, as well as their storm structures. For sensible heat flux, the maximum enhancements due to spray are about 2%, 12%, and 25% in Earl, the bomb, and the superbomb, respectively. This sensible heat flux enhancement is due to the spray-mediated increase in air–sea temperature difference ΔT, and related cooler boundary layers due to spray evaporation (discussed in the next section). Overall, the fixed SSTs used in our storm simulations tend to result in more sensible heat flux than would occur in most coupled atmosphere–ocean simulations, and in negative SST feedbacks from ocean mixing (Bao et al. 2000; Bender and Ginis 2000).

It is important to investigate the linkage between the surface flux distributions and the storm development and intensity. Spray impacts on heat fluxes that occur during the intensifying periods for the bomb and the superbomb act as positive feedbacks, by which more spray droplets are ejected into the atmosphere, providing more energy for storm development. The location of the flux center relative to the storm’s center is also important. Figures 8a and 8b show the latent heat (LH) flux centers, and the main spray-mediated LH difference centers for the bomb and the superbomb. Although the storms propagate north of 40°N as they mature, the LH flux centers do not. In the bomb, the maximal spray and wave enhancements (Fig. 4e) occur when the difference center of the surface heat fluxes (ΔLH + ΔSH) between the fully coupled and the control runs is near the storm center (Fig. 9a). The lower-atmospheric levels around the storm center are cooled due to spray evaporation (Δθ in Fig. 9b). The increased air–sea temperature difference tends to destabilize the atmospheric surface layer and thus increase the buoyancy production of the surface layer turbulence (Liu et al. 1979; Wang et al. 2001). Enhanced upward motion (Fig. 9b) coupled with moist fluxes from both the surface and the spray are transported upward, where latent heat release tends to warm the middle levels. These storm impacts due to moistening processes and heat release are similar to those of Huo et al. (1995), in which the upper-level geopotential height is increased and the lower-level height is decreased hydrostatically (Fig. 9b). Thus, the convergence of mass and moisture is increased in the lower levels, which favors storm intensification. Thereafter, during the bomb’s filling phase, the maximum heat flux region is relatively far from the storm center (Fig. 8a), resulting in the dominant mechanisms in the storm’s intensification being reduced, such as the impacts of local heat flux enhancements (Fig. 9c), moistening processes, and lower-level convergence (Fig. 9d). Moreover, because differential downward motion causes slight lower-level adiabatic warming, with an associated increase in geopotential height (Fig. 9d), the resulting enhancements in storm intensity and U10 are quickly diminished.

By comparison, the superbomb maintains a compact structure, with the storm center close to the location of maximum heat flux enhancement (ΔLH + ΔSH) during the intensifying period (Fig. 8b, also discussed later in Figs. 17a–e). During the filling phase, although the main heat flux region remains near the Gulf Stream and does not propagate farther north, the flux difference center propagates with the storm (Fig. 8b and Figs. 17a and 17d), which means that as the storm moves farther north, the heat flux near the storm’s active region increases locally due to spray. Thus, the moistening processes induced by spray are still important during the storm’s filling period and dominate over the wave impacts. Further discussion is given in section 5c.

5. Effects on the structure of the lower part of the atmosphere

The immediate impact of spray evaporation and wave drag is to modify the momentum and heat fluxes across the air–sea interface. These change the vertical boundary layer moisture and temperature structure and directly influence the surface layer turbulence and the vertical convection near the active region of the storm center. Heat for spray evaporation comes from upward transfers from the ocean and downward transfers from the atmosphere. Bao et al. (2000) suggest that these heat sources adjust internally until an equilibrium is reached, with spray acting as a source for moisture and heat.

a. Impacts of sea spray

To describe the impacts of spray on the lower-atmospheric layers, height–latitude cross sections through the superbomb’s storm center are given in Figs. 10b and 10c for the differences (MC2-spray minus MC2 only) in potential temperature Δθ, specific humidity Δq, and the along-plane flow vector. Contours for θ and q from the control run are included, showing the storm center’s structure, with high θ gradient values and a moisture ridge. This is during the storm’s intensification, when U10 and the latent heat flux regions are maximal (Figs. 4f and 7c) and are approximately collocated. It is notable that the maximum cooling (∼1.0 K) occurs at 750 hPa, rather than in the lower boundary layer. The cooling is tilted westward with height, following the tilting of the maximum winds to the southwest of the storm center. Moreover, the boundary layer on both sides of the storm center (particularly on the right side) becomes warmer and moister, due to the spray-enhanced sensible heat flux. Because the associated SST is considerably warmer than the 2-m atmospheric screen temperature Ta (discussed below, Fig. 13b), particularly during the intensification phase, when the storm is still over relatively warm water, more sensible heat flux from the surface is transported upward than in the control run. This increase in sensible heat flux can destabilize the surface layer, and promotes vertical mixing around the storm core (Fig. 10c), which facilitates lower-level mass and moisture convergence and results in uplift to the midtropospheric level by strong ascending motion.

The spray also impacts the superbomb’s boundary layer by providing a moisture source, as is shown in Fig. 10c. The boundary is systematically moistened on both sides of the storm center, by about 0.3–0.9 g kg−1. Maximum moistening (1.2 g kg−1) occurs at 750 hPa, about 300 km to the northeast of the storm center, and correlates with the spray-enhanced precipitation shown in Fig. 10d. Increased moisture is transported upward by the enhanced upward motion near the storm center (Fig. 10c), condenses at midtropospheric levels, and releases latent heat, contributing to the warming at these levels and above (Fig. 10b). This lowers the surface pressure and intensifies the storm. As the superbomb begins to fill and propagate over colder waters, the influence of the underlying sea surface decreases, although the moistening processes induced by spray can still be felt. For example, as the superbomb makes landfall on Cape Breton Island, a heat flux enhancement region develops near the high-wind region (Fig. 11a). The associated Δθ height–latitude section (Fig. 11b) shows boundary layer cooling by ∼0.4 K, because heat for spray evaporation is mainly from the atmosphere, and cooling is to the west of the storm center and extends throughout the boundary layer.

The weakest storm, Earl, initially propagates rapidly during its intensification phase (Fig. 2a), reaching Newfoundland when the main sea surface heat flux zone is far behind the storm center (dashed contours in Fig. 12a), while the spray-enhanced heat flux region is just to the east of the storm center (shaded area in Fig. 12a), within the high-wind zone. Because Earl has traveled over cooler waters than the surface atmospheric layer (Fig. 13a), heat for spray evaporation comes mainly from the atmosphere. This differs from the superbomb, which initially propagates slowly (Fig. 2c) over warmer SSTs than the atmospheric surface layer (Fig. 13b). Spray evaporation causes a reduction (∼−0.3 K) in low-level temperature that expands within the high-wind area east of the storm center, with maximal cooling (−0.5 K) at 700 hPa (Fig. 12b), due in part to rain evaporation. The boundary layer is systematically moistened by 0.2 g kg−1 (Fig. 12c). Although weaker than the superbomb, increased vertical motion related to the destabilized surface around Earl’s center is evident and, coupled with the latent heat release, facilitates its intensification.

When Earl begins to fill, the spray-enhanced heat flux zone (Figs. 12d) is still located to the northeast of the high-wind region, but relatively far from the storm center compared with its location during the intensifying period (Fig. 12a). Correspondingly, the moistening and cooling processes related to the spray’s evaporation and enhanced convection are also concentrated to the northeast of the storm center (Figs. 12e and 12f). Contrary to the intensifying phase, the differential spray-induced descending motion is dominant around the storm center (Fig. 12f), which suppresses the upward transport of mass and moisture. The storm loses its momentum and energy, and its development is suppressed.

b. Impacts of wave drag

Previous sections show that wave impacts compete with sea spray during storm development. Spray evaporation causes cooling and moistening in the boundary layer near the high-wind zone, destabilizing the surface layer and enhancing the turbulence at lower levels. This facilitates the upward transport of mass and moisture, which is coupled with latent heat release in the midtroposphere, favoring storm intensification. What is the impact of waves with no spray? Figures 14a and 14b show vertical cross sections through the superbomb’s storm center (transect shown in Fig. 10a) for the differences Δθ and Δq between the MC2-wave and MC2-only control simulations, while Figs. 15a and 15b show the corresponding vertical motions.

In the MC2-only control run (Fig. 15a), a weak downdraft occurs around the storm center near 700 hPa, but does not extend all the way to the sea surface, with updrafts related to the convection occurring just outside the storm center. The associated vertical motion is stronger on the left side of the storm than on the right side (Fig. 15a) because of the high-wind bias, typical of convective storm structures (Fig. 3a). Because of the overall wave-enhanced surface roughness (Fig. 4c) and kinetic energy loss in the MC2-wave simulation (compared with the control simulation in Fig. 4f), a dynamic compensation downdraft can be induced by wave drag. The corresponding difference in MC2-wave and the control (Fig. 15b) clearly shows the differential downdrafts outside the storm center, dynamically compensating for the lost kinetic energy, with the maximum just below 700 hPa. This compensating downdraft contributes to the downward mixing of upper-level drier and cool air (shown in Figs. 14a and 14b). The maximum cooling and drying occur near the storm center, with values up to −0.4 K and −0.02 g kg−1, respectively, which are smaller than those due to spray (Figs. 10b). This suggests that the lower-atmospheric boundary layer becomes relatively more stably stratified, and convection outside the storm center is suppressed in the MC2-wave simulation, compared with that of MC2 only. This is not favorable for storm development and significantly differs from the effects of the spray, in which the lower boundary is cooled and moistened (Fig. 10c) and the surface level is destabilized.

As discussed in section 4c, the overall impact of wave drag on heat fluxes is quite small as a thermal–dynamic effect resulting from the bulk formulations [Eqs. (2) and (3)]. This is because of the interaction among the reduced wind speed and increased ΔT and Δq and the fixed roughness lengths for thermal fluxes Z0t and Z0q. Thus, the linkage between wave drag and storm development is mainly through the dynamic processes, and the moistening processes play a very small role. By comparison, the spray-enhanced moistening processes generally dominate.

c. Combined impacts of sea spray and wave drag

The impact of spray is cooling and moistening in the boundary layer, whereas waves cause cooling and drying. Boundary layer impacts are determined by the balance of sea-spray-induced moistening processes and dynamic compensation processes relating to wave-induced drag. Boundary layer cooling has two potential effects: (a) a wave-drag-induced increase in the static stability in the lower troposphere, thus suppressing convection outside the storm center and possibly reducing storm intensity, or (b) a spray-induced increase in the air–sea temperature difference (SST − Ta), which may destabilize the surface layer and thus increase the buoyancy production of the surface layer turbulence (Wang et al. 2001), supporting the upward transportation of mass and moisture and increasing the storm intensity. Wave drag effects slightly dominate over those of spray in Earl’s intensity, whereas for the superbomb, the spray effects are dominant. To show the combined impacts of the spray and wave, Figs. 16 and 17 present an analysis of the physical processes of these two storms during their intensifying and peak intensity phases.

When the spray and waves are both included in the simulation (MC2-wave-spray), during Earl’s deepening phase, the enhanced wave-induced downward motion is evident to the northeast of the storm center in the high-wind region (Fig. 16c). This occurs because of the roughened sea surface related to the rapidly changing wind directions in this region of the storm (Fig. 16a). Dynamic divergence in the boundary layer (Fig. 16c) is slightly increased because of the enhanced descent to the northeast of the storm center, which tends to produce enhanced lower-level drying (Fig. 16c) and cooling (Fig. 16b) compared with the spray effects (Figs. 12b and 12c). By comparison, the ascending motion near the storm center shows little change due to wave-induced drag (Fig. 16c), and the midtropospheric region just to the southwest of the storm center is moister by about 0.2 g kg−1 because of this enhanced upward transport of moisture. When Earl reaches its peak intensity (Figs. 16d–f), the wave influences dominate over those of the spray, the ascending motion is replaced by descending motion under 600 hPa (Fig. 16f), and the lower atmosphere becomes drier around the storm. Because of an enhanced tendency for dynamic divergence within the boundary layer (Figs. 16c and 16f), the mass and moisture convergence are suppressed. The lower-level geopotential height rises and the upper level falls (Figs. 16b and 16e), decreasing the vertical ageostrophic circulation associated with storm development (Huo et al. 1995). Stability is increased and storm intensity reduced, compared with the control and MC2-spray runs.

The superbomb differs from Earl in that the underlying sea surface is comparatively warmer and, therefore, during its intensifying phase, the high winds facilitate greater spray evaporation and enhanced lower-level cooling (Fig. 17b). This destabilizes the surface layer and promotes enhanced vertical transportation (Fig. 17c) compared with the MC2-spray run (Figs. 10b and 10c). Consequently, the inclusion of waves only causes a slight change in the superbomb’s vertical structures for Δθ (Fig. 17b), Δq, and the along-plane flow vector (Fig. 17c). However, the downdraft to the southwest of the storm center is increased, and the updraft around the storm center is slightly decreased because of the dynamic compensation induced by wave drag. Correspondingly, the boundary layer and the high moisture center to the northeast of the storm center around 750 hPa (Fig. 17c), associated with precipitation (Fig. 10d), are about −0.2 to −0.4 g kg−1 drier compared with the MC2-spray run (Fig. 10c). At the superbomb’s peak intensity (Figs. 17d–f), the competition between waves and spray is evident. Descending motion occurs on both sides of the storm center, particularly to the southwest, and lower-level drying occurs (Fig. 17f) due to downward mixing of upper-level drier and cooler air. However, ascending motion and convergence around the storm center are maintained. This contributes to the transportation of locally spray-enhanced surface moisture (Fig. 17d) upward, where it warms the midtroposphere (Fig. 17e). Thus, although the moistening process relating to spray is partly counteracted by the waves’ influence, particularly to the southwest of the storm (see the Charnock parameter in Fig. 6d), the response of the differential geopotential height (Fig. 17e) to this midtroposphere warming is that the lower-level geopotential height falls and the upper level rises (similar to Fig. 9b), which tends to increase the lower-level convergence and upper-level divergence (Fig. 17f), and the maintenance of storm intensity is assisted. Therefore, when the superbomb starts to fill, the cooling and moistening processes related to spray are still relatively important and dominant over those due to waves. As mentioned in section 5a, because the sea surface is less influential compared with the previous intensification phase, the maximum cooling mainly occurs at lower levels.

d. Combined impacts of sea spray and wave drag from a potential vorticity perspective

The simplest conceptual model for the role of surface friction in cyclogenesis is the idea of Ekman pumping and vortex spindown, whereby positive relative geostrophic vorticity in the boundary layer results in frictional convergence, vertical motion, and reduction in vorticity above the boundary layer (Holton 1979). Except for vorticity, those features are shown and discussed in section 5b. Recently, Adamson et al. (2006) used potential vorticity (PV) analysis to consider the relative importance of baroclinic processes and Ekman pumping. In simulations of dry baroclinic cyclone life cycles, they suggest that baroclinic-generated PV anomalies along a warm front associated with surface friction are dominant, and can spiral upward and propagate westward toward the center of a low, thus greatly affecting the fate of a baroclinic cyclone.

To investigate the relation between Ekman pumping induced by wave drag and attendant baroclinic processes, Figs. 18a–f show the PV during the superbomb’s rapid development phase. In particular, the difference plots in Figs. 18c and 18d, showing MC2-wave minus MC2-only at 1000 and 900 hPa, respectively, suggest that a wave-drag-induced negative PV anomaly occurs in the vicinity and north of the storm center within the boundary layer (Fig. 18c), and is associated with the reduction in vorticity above the boundary layer (Fig. 18e). Moreover, a positive PV anomaly, produced by baroclinic processes, is evident to the east and northeast of the storm, in the region denoted the warm conveyor belt by Adamson et al. (2006), because of its association with potential temperature. Alternately, as is also evident in Figs. 18c and 18d, Stoelinga (1996) suggested that this positive anomaly was due to an inconsistent alignment of the surface wind and thermal wind (the thermal wind direction is the approximate mean isotherms). Figures 18c and 18d show that the positive PV anomaly is more evident at 900 hPa than at 1000 hPa, and that with increased height, it is advected westward toward the low center and spirals upward with the large-scale ascending flow along the warm conveyor belt. These features are generally consistent with idealized simulations by Adamson et al. (2006). Moreover, because of the attendant differential downdraft outside of the convergence (Fig. 15b) region, the upper-level positive PV (Fig. 18b) anomaly is propagated downward to ∼750 hPa, a process of dry intrusion of stratospheric air. Thus, a second positive PV region next to the negative PV also occurs at increased height in Fig. 18e, and the magnitude of this positive PV anomaly is slightly weaker than that of the negative PV.

This suggests that the wave-drag-induced Ekman pumping near the storm center, as well as the baroclinic processes along the warm front, are exhibited in our simulations. Wave drag is the dominant mechanism for storm damping in section 5b because of increased static stability. Moreover, under the assumption of PV conservation, we suggest that the negative PV around the storm center is converted to increased “static stability,” which is consistent with the discussion in section 5b. However, the impact on storm intensity is not as significant as that in studies by Adamson et al. (2006) or Doyle (1995), with a maximum decrease in wind speed of 5 m s−1 (∼13%), and in SLP of ∼3.0 hPa for the superbomb. A possible explanation for this difference is that we present a real midlatitude storm simulation with associated processes, whereas they present idealized simulations. Our overall results are similar in magnitude to simulations by Desjardins et al. (2000) and Zhang and Perrie (2001).

The impact of wave-induced drag is modified by sea spray in our simulations, which is a potential wet process associated with spray evaporation and latent heat release (section 5c). Latent heat release is conventionally assumed to be essential for storm development. According to Hoskins et al. (1985) diabatic heating tends to create (reduce) PV below (above) the level of maximum released latent heat. The modification of wave-drag-induced PV anomalies by diabatic heating can be clearly seen from the vertical cross section in Fig. 18f, showing the PV difference through the storm center between MC2-wave-spray and MC2-only. The transect extends through the storm center and the region of maximum lower-level baroclinic processes. Compared with PV anomalies due to wave drag (Fig. 18e), under 700 hPa (above 500 hPa), the negative PV anomaly pumped from the surface upward (downward-propagating positive PV) is suppressed by positive (negative) PV. This is mainly due to latent heat release associated with the maximum ascending region at approximately 700 hPa (see Fig. 10c). This creates (reduces) the PV below (above) this layer, as was suggested by Stoelinga (1996), and is favorable for coupling between the upper- and lower-level waves, by slowing the eastward propagation of the upper-level wave and by hastening the eastward propagation of the surface wave. Overall, the modification of the Ekman pumping effect due to latent heat release associated with sea spray is more obvious than its impact on the baroclinic-generated PV anomaly, which depends on the turning of the wind in the boundary layer and exhibits only a slight change.

6. Conclusions

A coupled atmosphere–wave–sea spray model system is used to evaluate the combined impacts of spray evaporation and wave drag on midlatitude storm development and intensity. We considered an extratropical hurricane (Earl, in 1998) and two intense winter storms from January 2000 and January 2002, denoted the superbomb and the bomb, respectively. The coupled model consists of the MC2 atmospheric model, the operational WW3 wave model, a bulk-type formulation for the heat and momentum effects of sea spray from Andreas and DeCosmo (1999, 2002) and Andreas (2003), and the HEXOS formulation for wave-induced stress.

While the impacts of sea spray and wave drag on storm tracks are small, the impacts on storm intensity are notable. The coupled MC2-wave-spray simulations reduced the peak error in area-averaged (2002 km2) U10 winds by as much as 25% compared with the control runs. This results from enhancements in air–sea fluxes. For latent (sensible) heat flux, the spray impact can be as much as 5% (2%), 16% (12%), and 30% (25%) for Earl, the bomb, and the superbomb, respectively, area averaged about peak heat flux regions. These peak flux enhancement values tend to occur during the storm’s period of intensification, when the sea state is young and rough, and winds are high, and spray droplet production rates are high, into the lower atmosphere.

Spray evaporation causes the lower part of the atmosphere to experience cooling and moistening. This cooling process increases the air–sea temperature difference, destabilizes the surface layer, and enhances the surface layer turbulence. Thus, the convergence of mass and moisture fluxes from the surface are enhanced, particularly when the local heat flux region is close to the active storm region, as when storms are over warm Gulf Stream waters. This results in the upward transport of moisture from both the surface and spray, coupled with latent heat release at the midtropospheric levels, which contributes to warming the midtroposphere air and lowering the surface pressure. This contribution (of moistening processes) to the thermal–dynamic structure is favorable for storm intensification. By comparison, the influence of wave drag on storm development is quite different from that of spray. Because of friction-induced kinetic energy dissipation associated with enhanced surface roughness, wave drag can induce anomalous convergence-generated upward motion around the storm center and an attendant dynamic compensation downdraft outside of the storm center, contributing to the downward mixing of upper-level dry cool air. This gives a slight increase in static stability in the lower troposphere, thus suppressing convection and reducing storm intensity, and is analogous to the classical Ekman spindown mechanism. A PV analysis is used to show that baroclinic processes associated with surface friction induced by wave drag along the warm front also occur, and also can be the dominant mechanism that affects storm development. The inclusion of sea spray can significantly dilute the Ekman pumping effect around the storm center, which relegates the wave drag effects to secondary importance for storm development. Further investigation of these processes is planned for the future.

Because of the time-invariant fixed SSTs used in our simulations, the differences between SSTs and TaT), and humidity (Δq), are slightly inflated, particularly when a storm moves over warmer SSTs. Thus, more sensible heat flux is provided to our simulated storms than would occur in comprehensively coupled atmosphere–ocean simulations, where the negative feedback between the cyclone and the upper ocean results in cooler SSTs and reduced heat fluxes from the sea surface. Increased sensible heat flux can cause the lower troposphere to potentially become unstable, which in turn can result in enhanced convection and precipitation, possibly reducing the cooling effect of the spray. On the other hand, the final impact of wave drag on heat fluxes is quite small because of the interaction among the reduced wind speed, increased ΔT and Δq, and the fixed roughness lengths for thermal fluxes Z0t and Z0q. Thus, the linkage between the wave drag and storm development is mainly through the dynamic processes, whereas the moistening processes play a very small role. By comparison, in our MC2-spray simulations, the moistening processes generally dominate.

The collective impacts of spray and waves on the structure of the lower part of the atmosphere depend on the competition between the spray’s thermal intensifying trend, compared with the wave drag’s dynamic deintensifying trend. Although both are important during a storm’s intensifying period, generally occurring over relatively warm Gulf Stream waters, as the storm propagates farther north, the spray impact is reduced because of cooler SSTs, and as the maximum heat flux region tends to lag behind, and to be far away from the storm’s active region. In this latter phase, the dynamic influence of waves tends to be important, as in the case of the bomb because of reduced local moistening processes. Only if the heat flux is locally increased within a storm’s active region because of spray, and if the moistening processes still dominate over the effects of wave drag, can the spray’s impacts be important during the whole storm life cycle, as in the case of the superbomb. To quantitatively assess the importance of wave-drag-induced friction and diabatic heating associated with spray, a piecewise PV inversion as suggested by Davis and Emanuel (1991) is necessary. This is planned for the future.

Acknowledgments

Funding was from the Panel on Energy Research and Development (PERD) of Canada, Petroleum Research Atlantic Canada (PRAC), the Natural Sciences and Engineering Research Council of Canada, the SURA Coastal Ocean Observing and Prediction (SCOOP) program, and the Canada Foundation for Climate and Atmospheric Sciences, under the SOLAS program. We thank John Gyakum and Ron McTaggart-Cowan for assistance with MC2, and Ed Andreas, for the sea spray model.

REFERENCES

  • Adamson, D. S., S. E. Belcher, B. J. Hoskins, and R. S. Plant, 2006: Boundary layer friction in mid-latitude cyclones. Quart. J. Roy. Meteor. Soc., 132 , 101124.

    • Search Google Scholar
    • Export Citation
  • Andreas, E. L., 1992: Sea spray and the turbulent air–sea heat fluxes. J. Geophys. Res., 97 , 1142911441.

  • Andreas, E. L., 1995: The temperature of evaporating sea spray droplets. J. Atmos. Sci., 52 , 852862.

  • Andreas, E. L., 1998: A new sea spray generation function for wind speeds up to 32 m s−1. J. Phys. Oceanogr., 28 , 21752184.

  • Andreas, E. L., 2003: An algorithm to predict the turbulent air–sea fluxes in high-wind, spray conditions. Preprints, 12th Conf. on Interaction of Sea and Atmosphere, Long Beach, CA, Amer. Meteor. Soc., CD-ROM, 3.4.

  • Andreas, E. L., 2004: Spray stress revisited. J. Phys. Oceanogr., 34 , 14291440.

  • Andreas, E. L., and J. DeCosmo, 1999: Sea spray production and influence on air–sea heat and moisture fluxes over the open ocean. Air–Sea Exchange: Physics, Chemistry and Dynamics, G. L. Geernaert, Ed., Kluwer, 327–362.

    • Search Google Scholar
    • Export Citation
  • Andreas, E. L., and K. A. Emanuel, 2001: Effects of sea spray on tropical cyclone intensity. J. Atmos. Sci., 58 , 37413751.

  • Andreas, E. L., and J. DeCosmo, 2002: The signature of sea spray in the HEXOS turbulent heat flux data. Bound.-Layer Meteor., 103 , 303333.

    • Search Google Scholar
    • Export Citation
  • Bao, J-W., J. M. Wilczak, J-K. Choi, and L. H. Kantha, 2000: Numerical simulations of air–sea interaction under high wind conditions using a coupled model: A study of hurricane development. Mon. Wea. Rev., 128 , 21902210.

    • Search Google Scholar
    • Export Citation
  • Bender, M. A., and I. Ginis, 2000: Real-case simulations of hurricane–ocean interaction using a high-resolution coupled model: Effects on hurricane intensity. Mon. Wea. Rev., 128 , 917946.

    • Search Google Scholar
    • Export Citation
  • Benoit, R., M. Desgagne, P. Pellerin, Y. Chartier, and S. Desjardins, 1997: The Canadian MC2: A semi-implicit semi-Lagrangian wide-band atmospheric model suited for fine-scale process studies and simulation. Mon. Wea. Rev., 125 , 23822415.

    • Search Google Scholar
    • Export Citation
  • Chalikov, D. V., and V. K. Makin, 1991: Models of the wave boundary layer. Bound.-Layer Meteor., 56 , 8399.

  • Charnock, H., 1955: Wind stress on a water surface. Quart. J. Roy. Meteor. Soc., 81 , 639640.

  • Chouinard, C., J. Mailhot, and A. Mitchell, 1994: The Canadian Regional Data Assimilation System: Operational and research applications. Mon. Wea. Rev., 122 , 13061325.

    • Search Google Scholar
    • Export Citation
  • Davis, C. A., and K. A. Emanuel, 1991: Potential vorticity diagnostics of cyclogenesis. Mon. Wea. Rev., 119 , 19291953.

  • DeCosmo, J., K. B. Katsaros, S. D. Smith, R. J. Anderson, W. A. Oost, K. Bumke, and H. Chadwick, 1996: Air–sea exchange of water vapor and sensible heat: The Humidity Exchange Over the Sea (HEXOS) Results. J. Geophys. Res., 101 , 1200112016.

    • Search Google Scholar
    • Export Citation
  • Desjardins, S., J. Mailhot, and R. Lalbeharry, 2000: Examination of the impact of a coupled atmospheric and ocean wave system. Part I: Atmospheric aspects. J. Phys. Oceanogr., 30 , 385401.

    • Search Google Scholar
    • Export Citation
  • Donelan, M. A., F. W. Dobson, S. D. Smith, and R. J. Anderson, 1993: On the dependence of sea surface roughness on wave development. J. Phys. Oceanogr., 23 , 21432149.

    • Search Google Scholar
    • Export Citation
  • Donelan, M. A., B. K. Haus, N. Reul, W. J. Plant, M. Stiassnie, H. C. Graber, O. B. Brown, and E. S. Saltzman, 2004: On the limiting aerodynamic roughness of the ocean in very strong winds. Geophys. Res. Lett., 31 .L18306, doi:10.1029/2004GL019460.

    • Search Google Scholar
    • Export Citation
  • Doyle, J. D., 1995: Coupled ocean wave/atmosphere mesoscale model simulations of cyclogenesis. Tellus, 47A , 766778.

  • Doyle, J. D., 2002: Coupled atmosphere–ocean wave simulations under high wind conditions. Mon. Wea. Rev., 130 , 30873099.

  • Drennan, W. M., H. C. Graber, D. Hauser, and C. Quentin, 2003: On the wave age dependence of wind stress over pure wind seas. J. Geophys. Res., 108 .8062, doi:10.1029/2000JC000715.

    • Search Google Scholar
    • Export Citation
  • Edson, J. B., S. Anquentin, P. G. Mestayer, and J. F. Sini, 1996: Spray droplet modeling: 2. An interactive Eulerian–Lagrangian model of evaporating spray droplets. J. Geophys. Res., 101 , 12791293.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., 1995: Sensitivity of tropical cyclones to surface exchange coefficients and a revised steady-state model incorporating eye dynamics. J. Atmos. Sci., 52 , 39693976.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., 1999: Thermodynamic control of hurricane intensity. Nature, 401 , 665669.

  • Emanuel, K. A., 2003: A similarity hypothesis for air–sea exchange at extreme wind speeds. J. Atmos. Sci., 60 , 14201428.

  • Fairall, C. W., J. D. Kepert, and G. J. Holland, 1994: The effect of sea spray on the surface energy transports over the ocean. Global Atmos. Ocean Syst., 2 , 121142.

    • Search Google Scholar
    • Export Citation
  • Fairall, C. W., E. F. Bradley, D. P. Rogers, J. B. Edson, and G. S. Young, 1996: Bulk parameterization of air–sea fluxes for Tropical Ocean Global Atmosphere Coupled Ocean–Atmosphere Response Experiment. J. Geophys. Res., 101 , 37473764.

    • Search Google Scholar
    • Export Citation
  • Holton, J. R., 1979: An Introduction to Dynamic Meteorology. Academic Press, 391 pp.

  • Hoskins, B. J., M. E. Mcintyre, and A. W. Robertson, 1985: On the use and significance of isentropic potential vorticity maps. Quart. J. Roy. Meteor. Soc., 101 , 637655.

    • Search Google Scholar
    • Export Citation
  • Huo, Z., D. L. Zhang, and J. Gyakum, 1995: A diagnostic analysis of the superstorm of March 1993. Mon. Wea. Rev., 123 , 17401761.

  • Janssen, P. A. E. M., 1989: Wave-induced stress and the drag of air flow over sea waves. J. Phys. Oceanogr., 19 , 745754.

  • Janssen, P. A. E. M., 1991: Quasi-linear theory of wind-wave generation applied to wave forecasting. J. Phys. Oceanogr., 21 , 16311642.

    • Search Google Scholar
    • Export Citation
  • Janssen, P. A. E. M., 1997: Does wind stress depend on sea-state or not? A statistical error analysis of HEXMAX data. Bound.-Layer Meteor., 83 , 479503.

    • Search Google Scholar
    • Export Citation
  • Janssen, P. A. E. M., J. D. Doyle, J. Bidlot, B. Hansen, L. Isaksen, and P. Viterbo, 2002: Impact and feedback of ocean waves on the atmosphere. Atmosphere–Ocean Interactions, Vol. I, W. Perrie, Ed., Advances in Fluid Mechanics, Vol. 33, WIT Press, 155–198.

    • Search Google Scholar
    • Export Citation
  • Kain, J. S., and J. M. Fritsch, 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.

  • Kepert, J. D., C. W. Fairall, and J-W. Bao, 1999: Modelling the interaction between the atmospheric boundary layer and evaporating sea spray droplets. Air–Sea Exchange: Physics, Chemistry and Dynamics, G. L. Geernaert, Ed., Kluwer Academic, 363–410.

    • Search Google Scholar
    • Export Citation
  • Komen, G. J., L. Cavaleri, M. Donelan, K. Hasselmann, S. Hasselmann, and P. A. E. M. Janssen, 1994: Dynamics and Modeling of Ocean Waves. Cambridge University Press, 532 pp.

    • Search Google Scholar
    • Export Citation
  • Komen, G. J., P. A. E. M. Janssen, V. Makin, and W. Oost, 1998: On the sea state dependence of the Charnock parameter. Global Atmos. Ocean Syst., 5 , 367388.

    • Search Google Scholar
    • Export Citation
  • Kuo, Y-H., R. J. Reed, and S. Low-Nam, 1991: Effects of surface energy fluxes during the early development and rapid intensification stages of seven explosive cyclones in the western Atlantic. Mon. Wea. Rev., 119 , 457476.

    • Search Google Scholar
    • Export Citation
  • Large, W. G., and S. Pond, 1982: Sensible and latent heat flux measurements over the ocean. J. Phys. Oceanogr., 12 , 464482.

  • Lionello, P., P. Malguzzi, and A. Buzzi, 1998: Coupling between the atmospheric circulation and ocean wave field: An idealized case. J. Phys. Oceanogr., 28 , 161177.

    • Search Google Scholar
    • Export Citation
  • Liu, W. T., K. B. Katsaros, and J. A. Businger, 1979: Bulk parameterization of air–sea exchanges of heat and water vapor including the molecular constraints at the interface. J. Atmos. Sci., 36 , 17221735.

    • Search Google Scholar
    • Export Citation
  • Makin, V. K., 1998: Air–sea exchange of heat in the presence of wind waves and spray. J. Geophys. Res., 103 , 11371152.

  • Makin, V. K., and C. Mastenbroek, 1996: Impact of waves on air–sea exchange of sensible heat and momentum. Bound.-Layer Meteor., 79 , 279300.

    • Search Google Scholar
    • Export Citation
  • Meirink, J. F., and V. K. Makin, 2001: The impact of sea spray evaporation in a numerical weather prediction model. J. Atmos. Sci., 58 , 36263638.

    • Search Google Scholar
    • Export Citation
  • McTaggart-Cowan, R., J. R. Gyakum, and M. K. Yau, 2001: Sensitivity testing of extratropical transitions using potential vorticity inversions to modify initial conditions: Hurricane Earl case study. Mon. Wea. Rev., 129 , 16171636.

    • Search Google Scholar
    • Export Citation
  • Moon, I-J., I. Ginis, and T. Hara, 2004: Effect of surface waves on air–sea momentum exchange. Part II: Behavior of drag coefficient under tropical cyclones. J. Atmos. Sci., 61 , 23342348.

    • Search Google Scholar
    • Export Citation
  • Perrie, W., and L. M. Wang, 1995: A coupling mechanism for wind and waves. J. Phys. Oceanogr., 25 , 615630.

  • Powell, M. D., P. J. Vickery, and T. A. Reinhold, 2003: Reduced drag coefficient for high wind speeds in tropical cyclones. Nature, 422 , 279283.

    • Search Google Scholar
    • Export Citation
  • Smith, S. D., and Coauthors, 1992: Sea surface wind stress and drag coefficient: The HEXOS results. Bound.-Layer Meteor., 60 , 109142.

    • Search Google Scholar
    • Export Citation
  • Stoelinga, M. A., 1996: A potential vorticity-based study of the role of diabatic heating and friction in a numerically simulated baroclinic cyclone. Mon. Wea. Rev., 124 , 849874.

    • Search Google Scholar
    • Export Citation
  • Tolman, H. L., 2002: Alleviating the garden sprinkler effect in wind wave models. Ocean Modell., 4 , 269289.

  • Tolman, H. L., and D. V. Chalikov, 1996: Source terms in a third-generation wind wave model. J. Phys. Oceanogr., 26 , 24972518.

  • Wang, Y., J. D. Kepert, and G. J. Holland, 2001: The impact of sea spray evaporation on tropical cyclone boundary layer structure and intensity. Mon. Wea. Rev., 129 , 24812500.

    • Search Google Scholar
    • Export Citation
  • Zhang, Y., and W. Perrie, 2001: Feedback mechanisms for the atmosphere and ocean surface. Bound.-Layer Meteor., 100 , 321348.

Fig. 1.
Fig. 1.

Comparison of the observed neutral drag coefficient Cd (circles, diamonds, squares, and triangles) from Powell et al. (2003) (vertical bars for 95% confidence), Large and Pond (1982) (light dashed line), Donelan et al. (2004), and Cd from the coupled model results from the MC2 grouped according to different wave ages [Eq. (5)], with the restriction that when U10 > 30 m s−1, we limit Z0 = 0.0034. Boldface dashed line and light solid line indicate Cd associated with young and fully developed sea states following Smith et al. (1992) and Drennan et al. (2003), respectively.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 2.
Fig. 2.

Storm tracks of (a) Earl, (b) the bomb, and (c) the superbomb, using MC2 with and without spray and waves, as well as NHC and CMC analyses. Storm centers are plotted every 6 h. Simulations for Earl began at 0000 UTC 5 Sep 1998, for the bomb at 1200 UTC 12 Jan 2002, and for the superbomb at 1200 UTC 20 Jan 2000.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 3.
Fig. 3.

(a) Control simulation and (b) QuikSCAT–NCEP blended wind fields, during maximum winds for the superbomb (0600 UTC 21 Jan). Units are kt.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 4.
Fig. 4.

Minimum SLP and U10 time series for (a), (d) Earl, (b), (e) the bomb, and (c), (f) the superbomb, following the storm tracks, respectively. The Charnock parameter (dashed line with triangles) is indicated in (a)–(c). Here, U10 is averaged on an area of 200 km2 over each storm’s high-wind regions.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 5.
Fig. 5.

Differences ΔSLP (hPa; contour) and ΔU10 (m s−1; shaded) at Earl’s peak intensity (1200 UTC 6 Sep 1998) for (a) MC2-spray minus control and (b) MC2-wave minus control. In (a) [(b)] ΔSLP contour starts at −1 (+1) mb, with ΔSLP intervals at −0.5 (+1) mb. Simulation winds U10 with (a) spray and (b) wave are superposed. Storm centers (cross-filled circles) are shown. For the bomb (0600 UTC 14 Jan 2002): (c) MC2-spray minus control and (d) MC2-wave minus control. In (c) [(d)] ΔSLP contour starts at −1 (+1) mb, with intervals at −1 (+1) mb. For the superbomb (1800 UTC 21 Jan 2000): (e) MC2-spray minus control and (f) MC2-wave minus control. In (e) [(f)] ΔSLP contour starts at −1 (+1) mb, with intervals at −2 (+1) mb.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 6.
Fig. 6.

The Charnock parameter (shaded) and the wind field (arrows) from MC2-wave, with the same simulation hour as in Fig. 4, for (a) Earl at 1200 UTC 6 Sep 1998, (b) the bomb at 0600 UTC 14 Jan 2002, and (c) the superbomb at 1800 UTC 21 Jan 2000. Storm centers are shown with a times sign.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 7.
Fig. 7.

Time series for 200 × 200 km2 area-averaged sensible (bottom four plotted time series) and latent (top four plotted time series) following maximal flux center storm with and without spray and waves, for (a) Earl, (b) the bomb, and (c) the superbomb. The difference of specific humidity q between the MC2-wave and MC2 runs are given for (d) Earl at 0600 UTC 6 Sep 1998, (e) the bomb at 0000 UTC 14 Jan 2002, and (f) the superbomb at 0000 UTC 21 Jan 2000.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 8.
Fig. 8.

Trajectories of the storm center and latent heat flux center from the fully coupled simulation compared with the center for the difference of the LH flux, between the fully coupled minus MC2-only control simulations for (a) the bomb and (b) the superbomb. During the period 1800 UTC 21 Jan–1200 UTC 22 Jan 2000 the LH center for the superbomb does not move.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 9.
Fig. 9.

(a) Total heat flux difference (ΔSH + ΔLH; shaded areas; W m−2) between the fully coupled (MC2-wave-spray) minus control (MC2-only) simulations at 0000 UTC 14 Jan 2002, superposed on control simulation heights at 850 hPa (thin solid lines) and wind speed at 1000 hPa. (b) Vertical cross section [for transect in (a)] for differences in potential temperature (shaded; negative Δθ with thick dashed contour), height (thin lines; dam), and along-plane vector flow (arrows). Associated results during the filling period at 1200 UTC 14 Jan 2002 are given in (c), (d). Storm center is shown with a cross-filled circle.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 10.
Fig. 10.

As in Fig. 9 but for the superbomb. (a) Total heat flux difference (ΔSH + ΔLH; shaded areas; W m−2) for MC2-spray minus MC2-only control simulations at 0600 UTC 21 Jan 2000, superposed with control simulation of heights at 850 hPa (thin solid lines) and wind speed at 1000 hPa. (b) Height–latitude cross sections [for transect in (a)] showing Δθ (shaded; K), (c) Δq (g kg−1) and along-plane vector flow (arrows), and (d) cumulative precipitation (mm h−1). Control simulation contours for θ, q, and storm center (cross filled circle along the x axis) are indicated.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 11.
Fig. 11.

As in Fig. 10 (MC2-spray minus MC2 only) for the superbomb at 0600 UTC 22 Jan 2000 during the superbomb’s filling phase. (a) Total ΔSH + ΔLH differences (shaded; W m−2), superposed with control simulation of heights at 850 hPa (thin solid lines) and wind speed at 1000 hPa. (b) Height–latitude cross sections [for transect in (a)] showing Δθ (K) differences.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 12.
Fig. 12.

As in Figs. 10a–c (MC2-spray minus MC2 only) for Earl: (a)–(c) at 1800 UTC 5 Sep 1998 during the intensifying period and (d)–(e) at 1200 UTC 6 Sep 1998 during the filling period. Vertical cross sections are through Earl’s storm center and high-wind region. Dashed contour lines in (a) and (d) indicate total heat flux from MC2-only control run.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 13.
Fig. 13.

The difference of SST minus screen temperature Ta (2 m) in the MC2-only control simulation for (a) Earl at 1800 UTC 5 Sep 1998 and (b) for the superbomb at 0600 UTC 21 Jan 2000. The storm center is shown with a times sign.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 14.
Fig. 14.

Difference between MC2-wave minus MC2-only control simulations for (a) Δθ (K) and (b) Δq (1.0 × 10−3 kg kg−1). The time is 0600 UTC 21 Jan 2000 when winds and latent heat flux are maximal. The control simulation is as shown in Figs. 10b and 10c, and transect line is as show in Fig. 10a.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 15.
Fig. 15.

Vertical motion (m s−1; positive is ascending motion) from (a) MC2-only control and (b) the difference between MC2-wave minus MC2-only simulations. The time is 0600 UTC 21 Jan 2000 when winds and latent heat flux are maximal. The transect line is as shown in Fig. 10a.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 16.
Fig. 16.

As in Figs. 9a and 9b for Earl: (a), (b) differences between the fully coupled and control simulations (MC2-wave-spray minus MC2 only), (c) showing difference of Δq and along-plane vector flows at 1800 UTC 5 Sep 1998, and (d)–(f) at 1200 UTC 6 Sep 1998. Control simulations are shown in Figs. 12. Transect sections go through the storm center and the high-wind region, as indicated by solid black lines in (a) and (d).

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 17.
Fig. 17.

As in Fig. 16 but for the superbomb: (a)–(c) at 0600 UTC 21 Jan 2000 and (d)–(f) at 1800 UTC 21 Jan 2000.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Fig. 18.
Fig. 18.

Simulated PV (PVU; 1 PVU = 10–6 m2 K kg−1 s−1; thick line) and θ (K; thin line) from MC2 only for (a) 1000 hPa and (b) vertical cross section for the superbomb [transect line in (c)] at 0600 UTC 21 Jan 2000. Differential PV (shaded areas) between MC2-wave and MC2 only are shown at (c) 1000 and (d) 900 hPa, with θ (thin line) superposed and wind vectors (arrows). The vertical PV difference cross section [transect line in (c)] between MC2-wave minus MC2 only for (e) and for (f) between MC2-spray-wave and MC2 only with 0.3-PVU intervals and where the thick solid line is the zero PV contour.

Citation: Monthly Weather Review 134, 9; 10.1175/MWR3191.1

Table 1.

List of numerical experiments.

Table 1.
Save