• Anthes, R. A., , R. W. Corell, , G. Holland, , J. W. Hurrell, , M. C. MacCracken, , and K. E. Trenberth, 2005: Hurricanes and global warming—Potential linkages and consequences. Bull. Amer. Meteor. Soc., 88 , 623628.

    • Search Google Scholar
    • Export Citation
  • Barnett, T. P., , D. W. Pierce, , K. M. AchutaRao, , P. J. Gleckler, , B. D. Santer, , J. M. Gregory, , and W. M. Washington, 2005: Penetration of human-induced warming into the World’s Ocean. Science, 309 , 284287.

    • Search Google Scholar
    • Export Citation
  • Bister, M., , and K. A. Emanuel, 1998: Dissipative heating and hurricane intensity. Meteor. Atmos. Phys., 65 , 233240.

  • Bister, M., , and K. A. Emanuel, 2002: Low frequency variability of tropical cyclone potential intensity 1. Interannual to interdecadal variability. J. Geophys. Res., 107 , 4801. doi:10.1029/2001JD000776.

    • Search Google Scholar
    • Export Citation
  • Camp, J. P., , and M. T. Montgomery, 2001: Hurricane maximum intensity: Past and present. Mon. Wea. Rev., 129 , 17041717.

  • DeMaria, M., 1996: The effect of vertical shear on tropical cyclone intensity change. J. Atmos. Sci., 53 , 20762087.

  • DeMaria, M., , and J. Kaplan, 1994: Sea surface temperature and the maximum intensity of Atlantic tropical cyclones. J. Climate, 7 , 13251334.

    • Search Google Scholar
    • Export Citation
  • Elsner, J. B., , J. P. Kossin, , and T. H. Jagger, 2008: The increasing intensity of the strongest tropical cyclones. Nature, 455 , 9295.

  • Emanuel, K. A., 1986: An air–sea interaction theory for tropical cyclones. Part I: Steady-state maintenance. J. Atmos. Sci., 43 , 586604.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., 1987: The dependence of hurricane intensity on climate. Nature, 42 , 483485.

  • Emanuel, K. A., 1995: Sensitivity of tropical cyclone 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., 2000: A statistical analysis of tropical cyclone intensity. Mon. Wea. Rev., 128 , 11391152.

  • Emanuel, K. A., 2007: Environmental factors affecting tropical cyclone power dissipation. J. Climate, 20 , 54975509.

  • Emanuel, K. A., , C. DesAutels, , C. Holloway, , and R. Korty, 2004: Environmental control of tropical cyclone intensity. J. Atmos. Sci., 61 , 843858.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., , R. Sundararajan, , and J. Williams, 2008: Hurricane and global warming: Results from downscaling IPCC AR4 simulations. Bull. Amer. Meteor. Soc., 89 , 347367.

    • Search Google Scholar
    • Export Citation
  • Frank, W. M., , and E. A. Ritchie, 2001: Effects of vertical wind shear on hurricane intensity and structure. Mon. Wea. Rev., 129 , 22492269.

    • Search Google Scholar
    • Export Citation
  • Free, M., , M. Bister, , and K. A. Emanuel, 2004: Potential intensity of tropical cyclones: Comparison of results from radiosonde and reanalysis data. J. Climate, 17 , 17221727.

    • Search Google Scholar
    • Export Citation
  • Goldenberg, S. B., , and L. J. Shapiro, 1996: Physical mechanisms for the association of El Niño and West African rainfall with Atlantic major hurricane activity. J. Climate, 9 , 11691187.

    • Search Google Scholar
    • Export Citation
  • Goldenberg, S. B., , C. W. Landsea, , A. M. Mestas-Nuñez, , and W. M. Gray, 2001: The recent increase in Atlantic hurricane activity: Causes and implications. Science, 293 , 474479.

    • Search Google Scholar
    • Export Citation
  • Henderson-Sellers, A., and Coauthors, 1998: Tropical cyclones and global climate change: A post-IPCC assessment. Bull. Amer. Meteor. Soc., 79 , 1938.

    • Search Google Scholar
    • Export Citation
  • Holland, G. J., 1997: The maximum potential intensity of tropical cyclones. J. Atmos. Sci., 54 , 25192541.

  • Houghton, J. T., , Y-H. Ding, , D. J. Griggs, , M. Noguer, , P. J. van der Linden, , X. Dai, , K. Maskell, , and C. A. Johnson, Eds.,. 2001: Climate Change 2001: The Scientific Basis. Cambridge University Press, 881 pp.

    • Search Google Scholar
    • Export Citation
  • Knutson, T. R., , and R. E. Tuleya, 2004: Impact of CO2-induced warming on simulated hurricane intensity and precipitation: Sensitivity to the choice of climate model and convective parameterization. J. Climate, 17 , 34773495.

    • Search Google Scholar
    • Export Citation
  • Knutson, T. R., , R. E. Tuleya, , and Y. Kurihara, 1998: Simulated increase of hurricane intensities in a CO2-warmed climate. Science, 279 , 10181021.

    • Search Google Scholar
    • Export Citation
  • Knutson, T. R., , J. J. Sirutis, , S. T. Garner, , G. A. Vecchi, , and I. M. Held, 2008: Simulated reduction in Atlantic hurricane frequency under twenty-first-century warming conditions. Nat. Geosci., 1 , 359364. doi:10.1038/ngeo202.

    • Search Google Scholar
    • Export Citation
  • Landsea, C. W., 2005: Hurricanes and global warming. Nature, 438 , E11E13.

  • Landsea, C. W., , B. A. Harper, , K. Hoarau, , and J. A. Kanff, 2006: Can we detect trends in extreme tropical cyclones? Science, 313 , 452454.

    • Search Google Scholar
    • Export Citation
  • Latif, M., , N. Keenlyside, , and J. Bader, 2007: Tropical sea surface temperature, vertical shear, and hurricane development. Geophys. Res. Lett., 34 , L01710. doi:10.1029/2006GL027969.

    • Search Google Scholar
    • Export Citation
  • Malkus, B. S., , and H. Riehl, 1960: On the dynamics and energy transformation in steady-state hurricanes. Tellus, 12 , 120.

  • Meehl, G. A., , C. Covey, , T. Delworth, , M. Latif, , B. McAvaney, , J. F. B. Mitchell, , R. J. Stouffer, , and K. E. Taylor, 2007: THE WCRP CMIP3 multimodel dataset: A new era in climate change research. Bull. Amer. Meteor. Soc., 88 , 13831394.

    • Search Google Scholar
    • Export Citation
  • Nguyen, K. C., , and K. J. E. Walsh, 2001: Interannual, decadal, and transient greenhouse simulation of tropical cyclone–like vortices in a regional climate model of the South Pacific. J. Climate, 14 , 30433054.

    • Search Google Scholar
    • Export Citation
  • Oouchi, K., , J. Yoshimura, , H. Yoshimura, , R. Mizuta, , S. Kusunoki, , and A. Noda, 2006: Tropical cyclone climatology in a global-warming climate as simulated in a 20 km-mesh global atmospheric model: Frequency and wind intensity analyses. J. Meteor. Soc. Japan, 84 , 259276.

    • Search Google Scholar
    • Export Citation
  • Paterson, L. A., , B. N. Hanstrum, , and N. E. Davidson, 2005: Influence of environmental vertical wind shear on the intensity of hurricane-strength tropical cyclones in the Australian region. Mon. Wea. Rev., 133 , 36443660.

    • Search Google Scholar
    • Export Citation
  • Peng, M. S., , B-F. Jeng, , and R. T. Williams, 1999: A numerical study on tropical cyclone intensification. Part I: Beta effect and mean flow effect. J. Atmos. Sci., 56 , 14041423.

    • Search Google Scholar
    • Export Citation
  • Pielke Jr., R. A., , C. Landsea, , M. Mayfield, , J. Laver, , and R. Pasch, 2005: Hurricane and global warming. Bull. Amer. Meteor. Soc., 80 , 20272033.

    • Search Google Scholar
    • Export Citation
  • Rayner, N. A., , D. E. Parker, , E. B. Horton, , C. K. Folland, , L. V. Alexander, , D. P. Rowell, , E. C. Kent, , and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108 , 4407. doi:10.1029/2002JD002670.

    • Search Google Scholar
    • Export Citation
  • Rotunno, R., , and K. A. Emanuel, 1987: An air–sea interaction theory for tropical cyclones. Part II: Evolutionary study using a nonhydrostatic axisymmetric numerical model. J. Atmos. Sci., 44 , 542561.

    • Search Google Scholar
    • Export Citation
  • Santer, B. D., and Coauthors, 2006: Forced and unforced ocean temperature changes in Atlantic and Pacific tropical cyclogenesis regions. Proc. Natl. Acad. Sci. USA, 103 , 1390513910. doi:10.1073/PNAS.0602861103.

    • Search Google Scholar
    • Export Citation
  • Saunders, M. A., , and A. R. Harris, 1997: Statistical evidence links of exceptional 1995 Atlantic hurricane season to record sea warming. Geophys. Res. Lett., 24 , 12551258.

    • Search Google Scholar
    • Export Citation
  • Schade, L. R., 2000: Tropical cyclone intensity and sea surface temperature. J. Atmos. Sci., 57 , 31223130.

  • Schade, L. R., , and K. A. Emanuel, 1999: The ocean’s effect on the intensity of tropical cyclones: Results from a simple coupled atmosphere–ocean model. J. Atmos. Sci., 56 , 642651.

    • Search Google Scholar
    • Export Citation
  • Shen, W., , R. E. Tuleya, , and I. Ginis, 2000: A sensitivity study of thermodynamic environment on GFDL model hurricane intensity: Implications for global warming. J. Climate, 13 , 109121.

    • Search Google Scholar
    • Export Citation
  • Smith, T. M., , and R. W. Reynolds, 2004: Improved extended reconstruction of SST (1954–1997). J. Climate, 17 , 24662477.

  • Sobel, A. H., , I. M. Held, , and C. S. Bretherton, 2002: The ENSO signal in tropical tropospheric temperature. J. Climate, 15 , 27022706.

    • Search Google Scholar
    • Export Citation
  • Stowasser, M., , Y. Wang, , and K. Hamilton, 2007: Tropical cyclone changes in the western North Pacific in a global warming scenario. J. Climate, 20 , 23782396.

    • Search Google Scholar
    • Export Citation
  • Sugi, M., , A. Noda, , and N. Sato, 2002: Influence of the global warming on tropical cyclone climatology: An experiment with the JMA global model. J. Meteor. Soc. Japan, 80 , 249272.

    • Search Google Scholar
    • Export Citation
  • Tang, B. H., , and J. D. Neelin, 2004: ENSO influence on Atlantic hurricanes via tropospheric warming. Geophys. Res. Lett., 31 , L24204. doi:10.1029/2004GL021072.

    • Search Google Scholar
    • Export Citation
  • Tonkin, H., , G. J. Holland, , N. Holbrook, , and A. Henderson-Sellers, 2000: An evaluation of thermodynamic estimates of climatological maximum potential tropical cyclone intensity. Mon. Wea. Rev., 128 , 746762.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K., 2005: Uncertainty in hurricanes and global warming. Science, 308 , 17531754.

  • Vecchi, G. A., 2007: Effect of remote sea surface temperature change on potential tropical cyclone intensity. Nature, 450 , 10661070.

  • Vecchi, G. A., , and B. J. Soden, 2007: Increased tropical Atlantic wind shear in model projections of global warming. Geophys. Res. Lett., 34 , L08702. doi:10.1029/2006GL028905.

    • Search Google Scholar
    • Export Citation
  • Vecchi, G. A., , B. J. Soden, , A. T. Wittenberg, , I. M. Held, , A. Leetmaa, , and M. J. Harrison, 2006: Weakening of tropical Pacific atmospheric circulation due to anthropogenic forcing. Nature, 441 , 7376.

    • Search Google Scholar
    • Export Citation
  • Wang, Y., 2002: Vortex Rossby waves in a numerically simulated tropical cyclone. Part II: The role in tropical cyclone structure and intensity changes. J. Atmos. Sci., 59 , 12391262.

    • Search Google Scholar
    • Export Citation
  • Wang, Y., , and G. J. Holland, 1996: Tropical cyclone motion and evolution in vertical shear. J. Atmos. Sci., 53 , 33133332.

  • Wang, Y., , and C-C. Wu, 2004: Current understanding of tropical cyclone structure and intensity changes—A review. Meteor. Atmos. Phys., 87 , 257278.

    • Search Google Scholar
    • Export Citation
  • Wang, Y., , M. T. Montgomery, , and B. Wang, 2004: How much vertical shear can a tropical cyclone resist? Bull. Amer. Meteor. Soc., 85 , 661662.

    • Search Google Scholar
    • Export Citation
  • Wong, M. L. M., , and J. C. L. Chan, 2004: Tropical cyclone intensity in vertical wind shear. J. Atmos. Sci., 61 , 18591876.

  • Yang, B., , Y. Wang, , and B. Wang, 2007: The effect of internally generated inner core asymmetric structure on tropical cyclone intensity. J. Atmos. Sci., 64 , 11651188.

    • Search Google Scholar
    • Export Citation
  • Zeng, Z., , Y. Wang, , and C-C. Wu, 2007: Environmental dynamical control of tropical cyclone intensity—An observational study. Mon. Wea. Rev., 135 , 3859.

    • Search Google Scholar
    • Export Citation
  • Zeng, Z., , L-S. Chen, , and Y. Wang, 2008: An observational study of environmental dynamical control of tropical cyclone intensity in the North Atlantic. Mon. Wea. Rev., 136 , 33073322.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Time series of the annual-mean SST averaged over the topical oceans (30°S–30°N, 0°–360°) for the twentieth-century runs (20C3M) and the NOAA ERSST. Line 19: ERSST; line 20: all-model ensemble; line 21: ensemble with CNRM-CM3, CSIRO Mk3.0, and GISS-ER removed. The order from lines 1 to 18 is the same as that in Table 1. Ensemble mean from models (a) 25.31°C for 1900–99 and (b) 25.44°C for 1948–99 vs the NOAA ERSST mean 25.64° and 25.76°C (Table 1).

  • View in gallery

    (a) The 70-yr changes associated with the linear trends in the annual mean SST (°C) of the model ensemble mean in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario; and (b) the composite mean SST anomaly (°C) in June in an El Niño year. Contours in (a) show the ensemble-mean background (initial) annual-mean SST.

  • View in gallery

    Time series of CO2 concentration (dashed) used in the IPCC AR4 1pctto2× scenario and the annual-mean SST (solid) of the 15 CGCM ensembles averaged over the tropical oceans (30°S–30°N, 0°–360°).

  • View in gallery

    Spatial distribution of the 70-yr changes associated with the linear trends in the outflow layer temperature (°C) in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario from the 15-CGCM ensemble mean for the (a) JASO mean and (b) JFM season. Contours show the initial background outflow layer temperature (°C).

  • View in gallery

    The 70-yr relative changes (%) associated with the linear trends in thermodynamic efficiency in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario for the (a) JASO mean and (b) JFM mean based on the 15-CGCM ensemble mean. Contours show the corresponding initial background thermodynamic efficiency.

  • View in gallery

    The 70-yr relative changes (%) in the root of disequilibrium enthalpy between the ocean and atmosphere in response to the increasing CO2 concentration in the IPCC AR4 ipctto2× scenario for (a) JASO and (b) JFM from the 15-CGCM ensemble mean. Contours show the initial background fields (k*0k, m s−1).

  • View in gallery

    The 70-yr changes associated with the linear trends in the thermodynamic potential intensity of TCs (m s−1, THPI) in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario for the (a) JASO mean and (b) JFM mean based on the 15-CGCM ensemble. Contours in (a) and (b) show the corresponding initial background THPI (m s−1). Boxes indicate the major TC activity basins (see Table 1).

  • View in gallery

    Time series of THPI changes from the corresponding initial values: (a) lines 1 and 2 for the south Indian Ocean; lines 3 and 4 for the north Indian Ocean; line 5 and 6 for the eastern Pacific; lines 7 and 8 for the western North Pacific; (b) Lines 1 and 2 are the same as lines 7 and 8 in (a); lines 3 and 4 for the southwest Pacific; lines 5 and 6 for the North Atlantic. The THPI change averaged over all basins (solid curve) and the linear regression of THPI relative to the initial value (dot curve).

  • View in gallery

    The 70-yr changes (%) associated with the linear trends in the model ensemble-mean vertical shear in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario for the (a) JASO mean and (b) JFM mean. Contours show the initial background vertical shear (m s−1).

  • View in gallery

    The 70-yr changes associated with the linear trends in model ensemble mean translational speed (Vtrans) in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario for the (a) JASO mean and (b) corresponding changes in percentage of the absolute (Vtrans − 5). Contours show the (a) initial background Vtrans (m s−1) and (b) absolute (Vtrans − 5; m s−1).

  • View in gallery

    The 70-yr changes (%) associated with the linear trends in model ensemble mean dynamical efficiency in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario for the (a) JASO mean and (b) JFM mean. Contours show the initial background dynamical efficiency.

  • View in gallery

    The 70-yr changes (m s−1) associated with the linear trends in the TC PI with dynamical control in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario from the 15-CGCM ensemble for the (a) JASO mean and (c) JFM mean. (b) The difference between the changes in the modified PI and the THPI for JASO mean. Contours show the initial background modified PI (m s−1).

  • View in gallery

    The 70-yr changes (%) associated with the linear trends in various control parameters and PI of TCs averaged in six individual TC activity basins in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario from the 15-CGCM ensemble: the thermodynamic potential intensity (THPI, black), thermodynamic efficiency (green), dynamical efficiency (blue), and modified potential intensity (PI, red).

  • View in gallery

    The 70-yr change (%) in the PI modified by the empirical dynamical efficiency in six individual TC basins in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario from 15 CGCMs. The different color bars correspond to the different CGCMs as given in the legends. The values of the percentage changes averaged across all models are also given.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 98 98 13
PDF Downloads 58 58 5

Response of Tropical Cyclone Potential Intensity to a Global Warming Scenario in the IPCC AR4 CGCMs

View More View Less
  • 1 Pacific Typhoon Research Center, KLME, Nanjing University of Information Science and Technology, Nanjing, China, and International Pacific Research Center, and Department of Meteorology, School of Ocean and Earth Science and Technology, University of Hawaii at Manoa, Honolulu, Hawaii
  • | 2 International Pacific Research Center, and Department of Meteorology, School of Ocean and Earth Science and Technology, University of Hawaii at Manoa, Honolulu, Hawaii, and Pacific Typhoon Research Center, KLME, Nanjing University of Information Science and Technology, Nanjing, China
  • | 3 International Pacific Research Center, and Department of Meteorology, School of Ocean and Earth Science and Technology, University of Hawaii at Manoa, Honolulu, Hawaii
© Get Permissions
Full access

Abstract

This paper reports on an analysis of the tropical cyclone (TC) potential intensity (PI) and its control parameters in transient global warming simulations. Specifically, the TC PI is calculated for phase 3 of the Coupled Model Intercomparison Project (CMIP3) integrations during the first 70 yr of a transient run forced by a 1% yr−1 CO2 increase. The linear trend over the period is used to project a 70-yr change in relevant model parameters. The results for a 15-model ensemble-mean climate projection show that the thermodynamic potential intensity (THPI) increases on average by 1.0% to ∼3.1% over various TC basins, which is mainly attributed to changes in the disequilibrium in enthalpy between the ocean and atmosphere in the transient response to increasing CO2 concentrations. This modest projected increase in THPI is consistent with that found in other recent studies.

In this paper the effects of evolving large-scale dynamical factors on the projected TC PI are also quantified, using an empirical formation that takes into account the effects of vertical shear and translational speed based on a statistical analysis of present-day observations. Including the dynamical efficiency in the formulation of PI leads to larger projected changes in PI relative to that obtained using just THPI in some basins and smaller projected changes in others. The inclusion of the dynamical efficiency has the largest relative effect in the main development region (MDR) of the North Atlantic, where it leads to a 50% reduction in the projected PI change. Results are also presented for the basin-averaged changes in PI for the climate projections from each of the 15 individual models. There is considerable variation among the results for individual model projections, and for some models the projected increase in PI in the eastern Pacific and south Indian Ocean regions exceeds 10%.

Corresponding author address: Dr. Yuqing Wang, Post Bldg. 409G, 1680 East-West Road, IPRC/SOEST, University of Hawaii at Manoa, Honolulu, HI 96822. Email: yuqing@hawaii.edu

Abstract

This paper reports on an analysis of the tropical cyclone (TC) potential intensity (PI) and its control parameters in transient global warming simulations. Specifically, the TC PI is calculated for phase 3 of the Coupled Model Intercomparison Project (CMIP3) integrations during the first 70 yr of a transient run forced by a 1% yr−1 CO2 increase. The linear trend over the period is used to project a 70-yr change in relevant model parameters. The results for a 15-model ensemble-mean climate projection show that the thermodynamic potential intensity (THPI) increases on average by 1.0% to ∼3.1% over various TC basins, which is mainly attributed to changes in the disequilibrium in enthalpy between the ocean and atmosphere in the transient response to increasing CO2 concentrations. This modest projected increase in THPI is consistent with that found in other recent studies.

In this paper the effects of evolving large-scale dynamical factors on the projected TC PI are also quantified, using an empirical formation that takes into account the effects of vertical shear and translational speed based on a statistical analysis of present-day observations. Including the dynamical efficiency in the formulation of PI leads to larger projected changes in PI relative to that obtained using just THPI in some basins and smaller projected changes in others. The inclusion of the dynamical efficiency has the largest relative effect in the main development region (MDR) of the North Atlantic, where it leads to a 50% reduction in the projected PI change. Results are also presented for the basin-averaged changes in PI for the climate projections from each of the 15 individual models. There is considerable variation among the results for individual model projections, and for some models the projected increase in PI in the eastern Pacific and south Indian Ocean regions exceeds 10%.

Corresponding author address: Dr. Yuqing Wang, Post Bldg. 409G, 1680 East-West Road, IPRC/SOEST, University of Hawaii at Manoa, Honolulu, HI 96822. Email: yuqing@hawaii.edu

1. Introduction

The potential intensity (PI) of tropical cyclones (TCs) is defined as an upper bound of intensity that a TC may reach under a given suite of environmental thermodynamic and dynamical conditions. Thermodynamic PI (THPI) is a theoretically possible peak TC intensity that is limited by thermodynamic factors only. Somewhat different approaches have been advocated for computing the THPI from an environmental state (Emanuel 1986; Holland 1997; Bister and Emanuel 2002), but different approaches lead to fairly similar values (Tonkin et al. 2000). The TCs simulated in simple axisymmetric models generally reach the THPI (Rotunno and Emanuel 1987). By analyzing real data, Emanuel (2000) showed that the probability that the maximum intensity (as measured by the maximum near-surface wind speed) attained by TCs of hurricane strength is roughly equal up to a value close to (but less than) the THPI determined from monthly-mean climatological values of the background environment. Holland (1997) noted that the very strongest TCs observed in recent history seem to have attained very nearly the THPI.

Previous studies have indicated that THPI of TCs agrees well with the observed maximum intensities of most severe tropical storms, and also that the spatial and seasonal variability of real TC intensity is highly correlated with the variability of a THPI determined by the environment (Emanuel 1986; Tonkin et al. 2000; Bister and Emanuel 2002; Free et al. 2004). Assuming that these observed connections between realized TC intensity and THPI hold in a changing climate, the predicted THPI should be a useful measure of how TC intensities will change with global warming (Emanuel 1987, 2007).

Previous studies have identified both thermodynamic and dynamic environmental control parameters that determine the TC intensity. The former include the sea surface temperature (SST), vertical temperature, and humidity structure of the atmosphere, while the latter include the vertical shear of large-scale horizontal winds and TC translational speed (e.g., Emanuel 2000; Zeng et al. 2007, 2008). The TC intensity may also be affected significantly by the internal dynamics of a TC itself (Camp and Montgomery 2001; Wang 2002; Wang and Wu 2004; Yang et al. 2007). Although there are considerable uncertainties in examining the TC intensity change in response to a changing climate based on past observations resulting from the uncertainties in observational databases (Landsea 2005; Landsea et al. 2006; Anthes et al. 2005; Pielke et al. 2005; Trenberth 2005), the THPI of TCs can be calculated given the large-scale thermodynamic structure of the atmosphere and the underlying SST (Emanuel 1995; Holland 1997).

The local SST determines the energy input available to TC development and maintenance (Malkus and Riehl 1960; Schade 2000; Saunders and Harris 1997), and thus is a key factor controlling the PI of a TC. SSTs over tropical oceans have displayed a warming from 0.5° to ∼0.6°C since the mid-nineteenth century (Houghton et al. 2001), with warming from 0.25° to ∼0.5°C during the past several decades (Rayner et al. 2003; Santer et al. 2006). State-of-the-art coupled general circulation models (CGCMs) with realistic anthropogenic forcing can simulate these warming trends reasonably well (Houghton et al. 2001; Barnett et al. 2005).

A TC can be regarded to the first approximation as a natural Carnot heat engine (Emanuel 1986, 1987, 1999). The thermodynamic efficiency of such a Carnot engine is determined by both the SST and outflow layer temperature in the upper troposphere (Emanuel 1995; Bister and Emanuel 2002). From this perspective global warming can influence the PI of TCs through changing the surface energy input or the upper-tropospheric energy exhaust or both (Emanuel 1987, 1999; Holland 1997; Henderson-Sellers et al. 1998). Because global warming can potentially affect both SST and the outflow layer temperature, the response of TC PI to global warming cannot be determined purely by the changes in local SST (Shen et al. 2000; Knutson and Tuleya 2004; Vecchi and Soden 2007).

In addition to the effect of local changes to SST and atmospheric thermodynamic structure, global warming may affect the development of TCs through changing the large-scale atmospheric circulation (Latif et al. 2007; Goldenberg and Shapiro 1996; Goldenberg et al. 2001; Vecchi and Soden 2007; Sobel et al. 2002; Shen et al. 2000; Tang and Neelin 2004). For example, Vecchi and Soden (2007) and Emanuel et al. (2008) commented on the possible effects of projected changes in the large-scale vertical wind shear on future TC intensity. In this paper we investigate the possible changes in TC PI from projected large-scale climate trends using a quantitative formulation of PI that explicitly includes both local thermodynamic fields and dynamical fields, namely, those derived by Zeng et al. (2007, 2008).

A dominant dynamical factor that controls the TC intensity is the vertical shear of the large-scale horizontal winds through the depth of the troposphere. Strong vertical shear inhibits the formation and intensification of TCs primarily by preventing the axisymmetric organization of deep convection and ventilating the TC warm core (Wang and Holland 1996; DeMaria 1996; Frank and Ritchie 2001; Wong and Chan 2004). Observational studies show consistent negative correlation between vertical shear and TC intensity change (DeMaria and Kaplan 1994; Paterson et al. 2005; Zeng et al. 2007, 2008). Goldenberg et al. (2001) attributed the active phase of North Atlantic hurricanes between 1995 and 2000, related to the weak phase between 1971 and 1994, to simultaneous increases in local SST and decreases in vertical wind shear. Vecchi et al. (2006) found that an increasing linear trend in vertical shear over the tropical North Atlantic is reproduced well by the CGCMs that participated in the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4) and is attributed to the anthropogenic forcing. Vecchi and Soden (2007) show that there is also a substantial increase in vertical shear over the tropical Atlantic and eastern Pacific in model projections for global warming scenarios. This is related to the model-projected decrease in the Pacific Walker circulation and the temperature difference between the tropical North Atlantic and the tropical Indian and Pacific Oceans that controls the vertical wind shear across the North Atlantic (Latif et al. 2007).

Another environmental dynamical control of TC intensity is the translational speed of the TC itself resulting from the steering effect of the large-scale environmental flow. Based on observational analyses, Wang and Wu (2004) and Zeng et al. (2007, 2008) found that both very intense TCs and TCs with rapid intensification rates only occur in a narrow range of translational speeds between 3 and 8 m s−1. If a TC moves too slowly, oceanic cooling induced by turbulent mixing generated by surface wind stress curl under the TC will disrupt the intensification (Schade and Emanuel 1999; Schade 2000), while if it moves too quickly the resulting asymmetric structure will also inhibit intensification (Peng et al. 1999). In this regard, the storm translation can add a wavenumber-1 wind asymmetry to the axisymmetric cyclone. In view of energetics (Emanuel 2000), the contribution by the asymmetric component to the volume-integrated entropy flux, which depends linearly on the ground-relative wind speed, tends to be zero if the exchange coefficient and boundary layer entropy are quasi symmetric about the TC center. However, the asymmetric component in the ground-relative wind field can have a net contribution to the volume-integrated surface frictional dissipation rate, which varies as the cube of the ground-relative wind speed. As a result, the net frictional dissipation rate implies a weaker TC with faster movement than that implied from the axisymmetric theoretical THPI (Emanuel 1995, 2000).

Zeng et al. (2007, 2008) studied the observed distribution of maximum TC intensities and their connection to large-scale environmental fields in 25 yr of real data in both the western North Pacific and the North Atlantic. They found that the storms rarely reach full THPI and that the upper bound of the observed distribution of peak intensities is better characterized as the THPI that is multiplied by an empirical “dynamical efficiency,” which involves the mean shear and TC translational speed.

The effect of global warming on the climatology of TC frequencies, tracks, and intensities is a subject of much interest. One approach to project the global warming impacts on TCs is to perform long integrations of regional or global climate models with spatial resolutions that are sufficient to explicitly represent at least the broad features of individual TCs (e.g., Nguyen and Walsh 2001; Sugi et al. 2002; Oouchi et al. 2006; Knutson et al. 2008; Stowasser et al. 2007). Unfortunately, limitations of computer power prevent such models from being run for long periods at a resolution that would allow them to adequately represent the full spectrum of observed TC intensities. Knutson et al. (1998) adopt a different approach in which regional models are run embedded in global model fields for short periods and locations where storms are formed in the global model. Emanuel et al. (2008) and Knutson et al. (2008) use different statistical–dynamical and dynamical techniques to downscale TC activity both from National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalyses for the present day (as a test of the method) and from IPCC AR4/phase 3 of the Coupled Model Intercomparison Project (CMIP3) global climate model projections (Meehl et al. 2007) for the late twenty-first century.

Here we examine possible changes in TC intensity through the twenty-first century by examining the THPI and the modified PI accounting for the dynamical efficiency using a large suite of the AR4/CMIP3 integrations. This approach, of course, is free of significant computational cost, and allows one to examine the results in a wide variety of models. This study aims to explore the projected changes in various control parameters, especially contributions of the environmental thermodynamic and dynamic factors that control TC PI over six individual active TC basins using ensemble simulations from 15 CGCMs forced by the 1% yr−1 CO2 increase to doubling emission scenario (1pctto2×) in the AR4/CMIP3 experiments. Unlike previous studies that focus mainly on the equilibrium response of increased CO2 concentration, this study analyzes the transient response of TC PI to the warming climate driven by increasing CO2 concentrations.

The rest of the paper is organized as follows. Section 2 describes the data and methodology used in this study. The projected changes in thermodynamic factors are analyzed in section 3, and the projected THPI is discussed in section 4. Section 5 discusses the projected changes in dynamical factors and the modified TC PI that includes both thermodynamic and dynamical controls. Our major results are summarized in the last section.

2. Data and methodology

The data analyzed in this study include both atmospheric variables, such as temperature, specific humidity, and horizontal winds, and SST from 18 CGCMs (Table 1) downloaded from the Web site of the Program of Climate Model Diagnosis and Intercomparison (PCMDI) for the IPCC AR4/CMIP3 archive. We focus mainly on the 1pctto2× scenario, in which the CO2 concentration is increased at 1% yr−1 to doubling in the 70th year and then fixed for the remainder of the run, while all other greenhouse gase (such as CH4 and N2O) concentrations were fixed at their initial values. The runs were initialized from the output of a preindustrial control run. In addition, we analyzed results for 1900–99 and 1948–99 from the scenario 20c3m, in which CO2 and other greenhouse gases concentration are taken from observations, to evaluate the performance of these CGCMs in reproducing the variability and trend in SST during the twentieth century. In this scenario, changing values of CO2 and other climate forcings were specified based on observations for a period that includes at least the entire twentieth century.

Because of differences in physical parameterizations, initial conditions, and effective climate forcings, these models produced different details in the mean climate and climate variability as well as the slope of the climate trend. Because SST is a basic factor that controls the TC intensity, we compared the time series of annual-mean SST averaged over tropical oceans between 30°S and 30°N from all models in the 20c3m runs with that from the National Oceanic and Atmospheric Administration (NOAA) Extended Reconstructed SST version 2 (ERSST V2; Smith and Reynolds 2004; see Fig. 1). The ensemble annual-mean SST based on the 18 models shows a cold bias in comparison with the ERSST. Several models simulate too-low SST during 1948–99 (Fig. 1b). The final column of Table 1 shows the root-mean-square error (RMSE) of the time series of annual-mean tropical-mean SST for each model relative to the ERSST observations. The second last column shows the same quantity, but for the global-mean SST. We chose to exclude the results of the Centre National de Recherches Météorologiques Coupled Global Climate Model, version 3 (CNRM-CM3), Commonwealth Scientific and Industrial Research Organisation Mark version 3.0 (CSIRO Mk3.0), and Goddard Institute for Space Studies Model E-R (GISS-ER) from our analyses. These three models display either the largest mean cold bias and RMSE values (CNRM-CM3, CSIRO Mk3.0)1 or a very weak tropical warming through 1pctto2× run (GISS-ER, not shown). We will focus on the ensemble mean of the remaining 15 model simulations in our analyses.

The TC THPI (defined in terms of the maximum near-surface wind speed) is calculated following Bister and Emanuel (2002):
i1520-0442-23-6-1354-e1
where Vm is the maximum near-surface wind speed, Ts is the SST, T0 is the outflow layer temperature, Ck is the exchange coefficient for enthalpy, and CD is the drag coefficient; here we take Ck/CD = 0.8, CAPE* is the convective available potential energy (CAPE) of air saturated at SST and lifted from sea level in reference to the environmental sounding, and CAPE is that of the boundary layer air at the radius of maximum wind. Additionally, the THPI can also be defined as follows (Bister and Emanuel 1998):
i1520-0442-23-6-1354-e2
where k*0 is the enthalpy of air in contact with the ocean and is assumed to be saturated enthalpy at SST and k is the enthalpy of air near the surface under the eyewall. Accordingly, the THPI is determined by SST, outflow layer temperature, and the degree of thermodynamic disequilibrium between the ocean and atmosphere. The thermodynamic efficiency is defined as (Bister and Emanuel 1998, 2002; Zeng et al. 2007, 2008)
i1520-0442-23-6-1354-e3
which is a function of SST and the outflow layer temperature. The latter is given as an output from the THPI calculation in Bister and Emanuel (2002). Because Ck/CD is considered as a constant, THPI can be decomposed into thermodynamic efficiency ε and the disequilibrium in enthalpy between ocean and atmosphere, k*0k. Because (1) or (2) includes only the thermodynamic factors, THPI is thus determined by thermodynamic control parameters only.
Based on observational analyses, Zeng et al. (2007, 2008) introduced a dynamical efficiency that takes into account the combined effect of vertical shear and translational speed
i1520-0442-23-6-1354-e4
where
i1520-0442-23-6-1354-e5
where Vshear is the vertical wind shear defined as the difference in wind speed between 200 and 850 hPa, and Vtrans is the translational speed of targeted TCs, which is estimated in this study by the mass-weighted wind speed between 850 and 300 hPa. Here, U0 is taken to be 60 m s−1. This dynamical efficiency can be considered as an empirical dynamical efficiency, which is smaller for larger UST, indicating the negative effects of both vertical shear and translational speed on TC intensity.
The modified PI, including the dynamical control, then is given by Zeng et al. (2008) as
i1520-0442-23-6-1354-e6
As a result, the modified PI is determined by both the thermodynamic control and dynamical efficiency. In general, it is closer to the actual TC intensity than the THPI (Zeng et al. 2008).

In the following analyses, we will focus on the major TC activity basins and their respective active TC seasons. The following six TC basins are defined (Table 2): North Atlantic (NATL; 8°–25°N, 40°–80°W), eastern Pacific (EPC; 10°–20°N, 100°–130°W), western North Pacific (NWP; 8°–30°N, 125°–170°E), north Indian Ocean (NIO; 8°–20°N, 60°–90°E), south Indian Ocean (SIO; 8°–20°S, 50°–115°E), and southwest Pacific (SWP; 8°–20°S, 155°–180°E). The active TC seasons are different among different basins. The active TC seasons defined in this study (Table 2) are from July to October (JASO) over the NWP, NATL, and EPC, from May–June and September–December (MJSOND) for the NIO, and January–March (JFM) for the SWP and SIO. Unless otherwise stated, changes in TC PI and its associated parameters are those for the TC basin averaged during its respective TC season from the 15-CGCM ensemble mean.

As mentioned already in the introduction, we are interested in the transient response of TC PI to the warming scenario with increasing CO2, and thus a projection that might be applicable to late-twenty-first-century conditions. Furthermore, although strong variability exists at interannual, decadal, and interdecadal time scales, we focus mainly on the linear trend of the TC PI in response to global warming resulting from increasing CO2 concentration. When we refer to the change resulting from the doubled CO2 concentration, we mean the total change in the 70 yr associated with the linear trend determined using a best-square fit to the first 70 yr of the 1pctto2x integrations.

3. Projected changes in thermodynamic control parameters

SSTs in the tropics/subtropics show a consistent increase with increasing CO2 concentration. The response in SST in the linear trend to the doubled CO2 concentration has a distinct spatial pattern with a larger increase in the equatorial and north Indian Ocean, equatorial eastern Pacific, and equatorial Atlantic (Fig. 2a). An east–west asymmetric SST response occurs in the equatorial Pacific with larger increase in the east, showing a spatial pattern that is similar to that in an SST anomaly in the mature phase of an El Niño event (Fig. 2b). With the doubled CO2 concentration, the annual-mean SST averaged in the tropics (30°S–30°N, 0°–360°) increases 1.54°C, with the Northern Hemisphere increasing more than the Southern Hemisphere (1.61° versus 1.48°C). The annual-mean tropical SST increases almost linearly with time in response to the increasing CO2 concentration (Fig. 3).

Table 2 gives the 70-yr changes associated with the linear trends in the annual-mean and the TC season–mean SSTs averaged over individual TC basins in response to an increasing CO2 concentration. The largest warming of 1.73°C in the annual-mean SST occurs in the NIO. The rest are 1.60°, 1.59°, 1.54°, 1.47°, and 1.44°C, respectively, in the SIO, EPC, NWP, NATL, and SWP. The changes in the TC season–mean SST averaged in individual basins are similar to those in the annual mean (Table 2). The TC season–mean SST increases by 1.69°, 1.61°, 1.67°, 1.55°, 1.52°, and 1.44°C in the NIO, SIO, EPC, NWP, NATL, and SWP, respectively. The SST increases more in the TC season than in the annual mean in the NWP, EPC, NATL, and SIO, but it is less or equal in the NIO and SWP. The above changes in SST are statistically significant at an over 99% confidence level. The linear regression variances contribute over 90% to the total variances in the six basins (not show).

The outflow layer temperature T0, which represents the upper-tropospheric temperature, is critical to the thermodynamic efficiency (3), and thus the TC THPI (Emanuel 1995; Bister and Emanuel 2002). The outflow temperature is not at a fixed pressure level but varies in space and time. In general, it is below −70°C in the major TC basins in their corresponding seasons, except for in the equatorial eastern Pacific where it is relatively high (contours in Fig. 4a). The outflow layer temperature increases 1.09°, 0.92°, and 0.84°C in the EPC, NATL, and NWP (Fig. 4a), and 1.63°, 0.98°, and 0.9°C in the SWP, SIO (Fig. 4b), and NIO (not shown) in their corresponding TC seasons in response to the 70-yr buildup of CO2 concentration. Note that the increase is larger in the outflow layer temperature than in SST in the SWP.

The 70-yr change associated with the linear trends in the thermodynamic efficiency in response to the increasing CO2 concentration displays a distinct spatial distribution (Fig. 5). The background thermodynamic efficiency shows the lowest averaged value (0.70) in the North Atlantic (Fig. 5a, Table 3) and the highest averaged value (0.74) in the southwest Pacific (Fig. 5b, Table 3). The thermodynamic efficiency increases in most of the TC basins but decreases over the eastern North Atlantic, South China Sea, far eastern Pacific, and southwest Pacific in response to an increasing CO2 concentration. The relative change (%) in the thermodynamic efficiency varies between −1.1% and +3.2% in the six TC basins. The basin-averaged relative changes are 0.13% (−0.81% to ∼1.23%), −0.02% (−0.83% to ∼3.2%) and 0.074% (−0.84% to ∼1.12%) in the NWP, EPC, and NATL in JASO (Fig. 5a); 0.15% (−1.1% to ∼1.4%) in the NIO during May–June and September–December (not shown); and 0.046% (−0.89% to ∼1.22%) and −0.52% (−1.0% to ∼0.068%) in the SIO and SWP in JFM (Fig. 5b). Therefore, the response of the thermodynamic efficiency to increasing CO2 is relatively small. This is mainly due to the fact that the increase in SST is largely offset by the increase in the outflow layer temperature.

The degree of thermodynamic disequilibrium in enthalpy between the ocean and atmosphere is a direct controlling factor to THPI in addition to the thermodynamic efficiency as shown in (2). Figure 6 shows the spatial distribution of changes in k*0k (%) over six TC basins. In general, the disequilibrium in enthalpy across the air–sea interface increases in six basins except for a small area across the south Caribbean Sea. The largest increase in k*0k occurs in the SIO with averaged increase of 3.48%, and the increase of 2.16%, 3.03%, and 1.21% occurs in the NWP, EPC, and NATL in JASO (Fig. 6a), respectively. Although the thermodynamic efficiency decreases in the SWP, k*0k increases by 2.05%, on average (Fig. 6b).

4. Projected changes in thermodynamic potential intensity

The THPI of TCs at each grid point in each month for each model was calculated using Eq. (1), following Bister and Emanuel (2002). The ensemble mean from the 15 CGCMs was then constructed for the TC seasons over the global tropics. The background THPI of TCs (Fig. 7a) shows the spatial distribution to be very similar to the background SST (Fig. 2a), indicating the dominant control of the THPI by the local SST. The background THPI is generally large in JASO in the major TC basins with the maximum 66.9 m s−1 averaged in the EPC (ranging from 39.8 to 80.2 m s−1), followed by 65.0 m s−1 in the NWP (46.6–77.4 m s−1), and the minimum of 61.4 m s−1 in the NATL (48.26–74.0 m s−1). The background THPI is, respectively, 67.1 (54.0–74.2 m s−1), 66.5 (43.4–78.8 m s−1), and 72.0 m s−1 (62.5–78.4 m s−1) averaged in the NIO, SIO, and SWP (Fig. 7b). Therefore, the maximum basin-averaged background THPI occurs in the SWP among the six active TC basins.

The THPI increases in all six TC basins in response to increasing CO2 concentration, but with different linear trends (Figs. 7a,b). The basin-averaged increases in THPI are 0.67 (−1.15 to ∼2.62 m s−1), 1.87 (−0.11 to ∼3.64 m s−1), and 1.3 m s−1 (−0.11 to ∼2.37 m s−1) in the NATL, EPC, and NWP (Fig. 7a), and 1.98 m s−1 in the NIO (not shown) and 2.06 (0.17 to ∼2.85 m s−1) and 0.93 m s−1 (0.46 to ∼1.66 m s−1) in the SIO and SWP (Fig. 7b). The increase in THPI is the smallest in the MDR over the NATL, with a distinct decrease in the central and eastern Atlantic (Fig. 7a). The corresponding relative changes in THPI are 1.0% (−2.35% to ∼4.45%), 2.84% (−0.18 to ∼7.33%), 2.04% (−0.22% to ∼3.86%), 2.98% (−0.14% to ∼5.15%), 3.09% (0.38% to ∼4.57%), and 1.32% (0.6% to ∼2.65%), respectively, in the MDR over the NATL, EPC, NWP, NIO, SIO, and SWP. Accordingly, the corresponding relative changes that are normalized by the local SST increase are then 0.63% (°C)−1, 1.7% (°C)−1, 1.38% (°C)−1, 1.76%/(°C), 1.91% (°C)−1, and 0.91% (°C)−1, respectively. The THPI has the minimum response in the MDR over the NATL with an only 1.0% increase with the 70-yr buildup of the doubled CO2 concentration. Even though the maximum response occurs in the SIO, the THPI increases only by 3.1%. Therefore, the response of the TC THPI to increasing CO2 concentration is generally not dramatic. Because the change in thermodynamic efficiency is smaller than that in the disequilibrium in enthalpy across the air–sea interface, the percentage in the THPI increase is mainly attributed to the increase in the thermodynamic disequilibrium, except for in the NATL. The THPI decrease in the eastern Atlantic and southern Caribbean Sea is due to the smaller increase in enthalpy disequilibrium than the increase in thermodynamic efficiency. Note that there is a considerable increase in THPI in the region off of the U.S. southeast coast in the Gulf of Mexico and NATL.

Figure 8 shows the time series of TC THPI deviations from the corresponding initial values and the associated linear trends in the individual TC basins. The increasing trend over all of the TC basins is modest in the transient response to the increasing CO2 concentration. There is no apparent further trend after the CO2 concentration is fixed at year 70, but the interannual fluctuations are clearly stronger in the period after year 70 than during the period of rising CO2 concentrations.2 The interannual variability is relatively large in the NIO and SWP, while is smaller in the NWP and NATL. The response of the TC THPI variability to the global warming is an interesting topic itself for a future study.

5. Projected changes in potential intensity, including dynamical control

It is well known that vertical shear of large-scale horizontal flow is an important dynamical control on TC development and intensity (Goldenberg et al. 2001; Emanuel et al. 2004; Wang et al. 2004) and is expected to reduce the THPI (Zeng et al. 2007, 2008). It is thus important to examine the response of the large-scale vertical shear to the increase in CO2 concentration. The initial background vertical shear is generally weak to modest in the major TC basins (Fig. 9). The averaged background vertical shears in their corresponding TC seasons are 10.7 m s−1 (varying from 6.8 to 15.5 m s−1) over the NATL, 10.7 m s−1 (7.6–18.3 m s−1) over the NWP, 8.9 m s−1 (7.3–13.1 m s−1) over the EPC (Fig. 9a), and 14.6 m s−1 (11.6–20.5 m s−1) over the NIO (not shown). Over the SIO and SWP (Fig. 9b), the background vertical shear averaged in JFM is 10.2 (6.3 to ∼15.2 m s−1) and 10.0 m s−1 (6.8 to ∼16.7 m s−1), respectively. The strong vertical shear over the NIO is mainly due to the effect of the Tibetan Plateau and the regional monsoon climate and is thus responsible for the low TC frequency and weak TCs in the basin.

The model-projected vertical shear in response to increasing CO2 concentration differs greatly among the six major TC basins (Figs. 9a,b). The vertical shear increases 0.97 m s−1 (or 8.9%) over the NATL, 0.38 m s−1 (4.5%) over the EPC, and 0.24 m s−1 (2.3%) over the SWP, while decreases 0.31 m s−1 (2.9%) over the NWP, 0.28 m s−1 (2.0%) over the NIO, and 0.31 m s−1 (3.1%) over the SIO in response to the 70-yr buildup of CO2 concentration (Table 3).

The translational speed is another dynamical factor affecting TC PI (Emanuel 2000; Wang and Wu 2004; Zeng et al. 2007, 2008). Figure 10 shows changes in the ensemble-mean translational speed (Vtrans) in JASO. The background Vtrans is about 4 m s−1 in TC active regions over the NWP and NATL basins; it increases by about 0.2 m s−1 in the southern NWP and 0.2 ∼ 0.4 m s−1 in the southern NATL and decreases by 0.2 ∼ 0.4 m s−1 in the northern NWP, Gulf of Mexico, Caribbean Sea, and central NATL. There is a significant decrease off the U.S. southeast coastal region. Figure 10b shows the changes in percentage and the background spatial distribution of |Vtrans − 5| because |Vtrans − 5| appears in Eq. (5). Here, |Vtrans − 5| decreases in a large portion of the NWP, with a 4% to ∼6% decrease in the western NWP, off of the southeast China coast, and the South China Sea. Also, |Vtrans − 5| varies over the NATL with a 9% to ∼12% increase in the southern NATL and a 4% to ∼6% decrease in the U.S. southeast coastal region. It increases by 6% to ∼9% over the EPC.

Comparing Fig. 10b with Fig. 9a, we can see that |Vtrans − 5| and vertical shear have similar spatial patterns, as do their percentage changes. However, vertical shear Vshear contributes much more to the factor Ust in Eq. (5) than |Vtrans − 5| because of the large difference in their background values. As we can see from Fig. 10b, the background |Vtrans − 5| is about 2 to ∼3 m s−1 over the NWP and the NATL and about 2 m s−1 over the EPC. These are much smaller than the background vertical shear Vshear, which is around 10 m s−1 (Fig. 9a). As a result, vertical shear dominates the factor Ust in Eq. (5), the dynamical efficiency in Eq. (4), and their changes in response to the increasing CO2 concentration.

The empirical dynamical efficiency is calculated based on Eq. (4) from Zeng et al. (2008). The background dynamical efficiency is over 0.8 over all TC basins (Fig. 11). It is relatively high (0.88) over the EPC, NWP, NATL (Fig. 11a), SIO and SWP (Fig. 11b), while relatively low in the NIO (0.84, not shown). The relative changes in the dynamical efficiency in response to the 70-yr buildup of CO2 concentration show decreases of 1.05% and 0.46% averaged over the NATL and EPC, respectively, with an increase of 0.33% averaged over the NWP (Fig. 11a). The changes over other three TC basins are quite small (generally less than 0.3%). As indicated above already, although the empirical dynamical efficiency of Zeng et al. (2007, 2008) includes effects of both vertical shear and translational speed, its change in response to the increasing CO2 concentration is dominantly determined by vertical shear. The relative changes in dynamical efficiency show an opposite trends to that in vertical shear (Figs. 9 and 11).

Figure 12 shows the spatial distributions of the changes in the modified PI, including the dynamical control in JASO (Fig. 12a) and JFM (Fig. 12c) and the differences between the percentage changes in the modified PI and that in the THPI in JASO (Fig. 12b) in response to the doubled CO2 concentration. Note that the changes in each grid box over the six TC basins are significant at the 99% confidence level for the 15-model ensemble or for each model. The largest increases in percentage change occur in the SIO and NIO with 3.39% and 3.26%, respectively (Figs. 12a,c). The intermediate increases occur in the EPC and NWP, with 2.41% and 2.33%, respectively. These are roughly consistent with the changes in the THPI. In contrast to an overall increase in the THPI (Fig. 7), there appears to be a coherent band with a significant reduction in the modified PI extending eastward from the Caribbean Sea to the MDR over the North Atlantic (Fig. 12a) as a result of the significant increase in vertical shear (Fig. 9a). However, consistent with the THPI, the modified PI also shows a significant increase near the U.S southeast coast, the Gulf of Mexico, areas near Bahamas, and Cuba. Regions with PI increase are collocated with those with the active TCs affecting Cuba and the United States. Therefore, this is an indication of potentially increased intensity of TCs that will affect either the United States or Cuba. Therefore, we calculated the modified PI averaged in a small box (Fig. 12a; 25° to ∼30°N, 90° to ∼50°W) to the north of the MDR over the NATL, which covers the U.S. southeast coastal region and the Gulf of Mexico, potentially a major region for TCs to landfall in the U.S. southeast coast. Although it decreases by 0.07% averaged in the MDR, the modified PI increases by 1.35 m s−1 (or 2.38%) over 70 yr averaged in the region slightly to the north of the MDR in the NATL.

The difference in the percentage changes between the modified PI and the THPI in response to the increasing CO2 concentration is generally negative over the NATL, ENP, and SWP, while it is positive over the WNP, NIO, and SIO (Figs. 12b and 13). Overall, the difference in the changes in the modified PI is larger than that in the THPI over basins with increasing dynamical efficiency, but smaller than that in the THPI over basins with decreasing dynamical efficiency (Table 3 and Fig. 13). According to Eq. (6), the percentage change in the modified PI would be the same as that in THPI if there is no change in dynamical efficiency. Therefore, the difference in the percentage changes between the modified PI and the THPI is mainly determined by the percentage changes in dynamical efficiency. Because there is a relatively large decrease in dynamical efficiency in the MDR over the NATL, the dynamical control is responsible for the 51.2% decrease in the percentage change in the modified PI compared to that in the THPI in the region. The contributions by the dynamical control to the difference in percentage changes between the modified PI and the THPI are −13.94%, +13.92, +8.31%, +9.12%, and −13.72% in the EPC, NWP, NIO, SIO, and SWP, respectively. For the west coastal region in the NATL and Caribbean Sea, the averaged contribution by dynamical control is about −15% because of the smaller increase in vertical shear and the larger increase in THPI than in the MDR/NATL.

In a dynamical downscaling study using a regional atmospheric model for a warmed climate with doubled CO2 concentration, Knutson et al. (2008) found a 2.9% increases in the mean maximum wind speed for tropical storms and hurricanes combined and a 1.7% increase for hurricanes alone over the NATL. This is roughly consistent with our PI changes when averaged over the NATL with the MDR excluded. Note that most TCs over the NATL do not necessarily reach their maximum intensity over the genesis and development region, but instead do so slightly to the northwest of the MDR (see Fig. 2 in Knutson et al. 2008, e.g.). Based on the latest TC reanalysis dataset during 1981 and 2006, Elsner et al. (2008) found significant upward trends for TC lifetime maximum wind speed quantiles above the 70th percentile, with trends as high as 0.3 ± 0.09 m s−1 yr−1 for the strongest cyclones. The largest increase at this quantile is found to occur over the North Atlantic, although not all basins show statistically significant increases. Note that the observed trends during a 26-yr period might be affected by strong natural interannual and decadal variabilities and could not be compared directly with the long-term projected relative changes in the TC PI discussed above for the difference in individual TC basins. Nevertheless, the increase in TC PI in response to global warming over the NATL in the global model ensemble mean documented in our study seems to be consistent with the observed increasing trend in TC intensity found by Elsner et al. (2008).

In addition to the six major active TC basins, although rare in the current climate, TCs may possibly form in other ocean basins, such as the South Atlantic. Indeed, Hurricane Catarina occurred in South Atlantic in March 2004 just off the coast of Santa Catarina and Rio Grande do Sul of Brazil in meteorology, a tropical cyclone is a storm system fueled by the heat released when moist air rises and condenses. From Fig. 12c, we can see that the modified PI increases by 1%–3% north of 15°S and 1% near the southeast coast of Brazil over the South Atlantic. These are closely related to the increase in THPI and the decrease in vertical shear to the north of 15°S, while mainly to the increase in THPI near the southeast coast of Brazil (Figs. 7b and 9b). Although the background PI is relatively low, our result thus suggests a potential increase in TC activity over the South Atlantic in a warmed climate.

The discussion thus far has focused on the multiple model ensembles. Figure 14 shows the percentage changes in the modified PI averaged over individual TC basins for each of the 15 models considered. There is a general increasing tendency of PI in all models, but with considerable variability among the different models and different basins. This is consistent with the results of Emanuel et al. (2008), who showed an overall tendency toward increased intensity of storms under IPCC emissions scenario A1B, but with large variability among the results obtained with large-scale trends taken from different global models. In our analysis all of the models show increasing trends of PI in the NIO; 14, 13, and 11 out of 15 models show increasing trends in the SIO, NWP, and EPC, respectively. In the NATL, however, six models show increasing trends, six models show decreasing trends, and three models show almost no trend in PI. We noticed that the trends in our calculated PI for the NATL are slightly different from the projected changes in Emanuel et al. (2008) from the same models. The difference is most likely due to the fact that in Fig. 14 we only show the result averaged in the MDR, while in Emanuel et al. (2008) TCs moving out of the MDR over the NATL were included. Another possible explanation for the difference is the fact that we only estimated the maximum potential intensity (the worst case), which is different from the mean intensity from all storms as given in Emanuel et al. (2008). In Emanuel et al. (2008), the intensity increase may be a result of the increase in the ratio of strong TCs, which is not necessarily due to the increase in the most intense storms, as also shown in Knutson et al. (2008).

6. Conclusions

In this study, we have analyzed the 70-yr changes associated with the linear trends in the thermodynamic and dynamical parameters that control TC PI in response to a transient global warming scenario resulting from the increase in CO2 concentration based on the ensemble of the projected simulations from 15 of the CGCMs that participated in the IPCC AR4. Two formulations of TC PI have been applied: one is the pure thermodynamic PI (THPI) and the other is a modified PI that includes the effect of the dynamical efficiency empirically determined based on observations by Zeng et al. (2008). Our major results are summarized below. Trends are expressed as the change over 70 yr of increasing CO2 concentration.

  1. The warming in SST is universal and robust in response to the increase in CO2 concentration. The SST over the tropical oceans (30°S to ∼30°N, 0° to ∼360°) on average increases 1.54°C. The increase is the smallest in the North Atlantic among the four TC basins in the Northern Hemisphere. In particular, the increase in SST in the equatorial eastern Pacific is larger than in the equatorial western Pacific, giving a pattern similar to that in the SST anomalies associated with an El Niño event.
  2. The linear trend of the outflow layer temperature is positive, indicating a warming in the upper troposphere resulting from the increasing CO2 concentration, although the trend is generally smaller than the increase in SST in the active TC basins. The thermodynamic efficiency that is critical to the TC PI shows a very small increasing trend of 0.046% to ∼0.13% averaged in most of the major TC basins, except for in the eastern Pacific and southwest Pacific where it shows a very small decreasing tendency of −0.02% and −0.52%, respectively.
  3. The change of the disequilibrium in enthalpy across the ocean–atmosphere interface varies from 1.21% to ∼3.48% averaged over different TC basins, which is much larger than that of the thermodynamic efficiency. As a result, the change in THPI is mainly due to the local thermodynamic disequilibrium between the ocean and the atmosphere, which is also a function of SST. The THPI increases in all of the TC basins, while the increase is the smallest in the MDR over the North Atlantic, with an average increase of 1%. The maximum increasing trend occurs in the south Indian Ocean with an average increase of 3.09%. The THPI increases by 2.84%, 2.04%, and 2.98% in the eastern Pacific, western North Pacific, and north Indian Ocean, respectively.
  4. The linear trend in vertical shear differs greatly among the six TC basins. Vertical shear increases in the North Atlantic, eastern Pacific, and southwest Pacific by 8.9%, 4.5%, and 2.3%, respectively, while it decreases in the other three TC basins. Although the percentage change in |Vtrans − 5| is generally larger than that in vertical shear, the background |Vtrans − 5| is much smaller than the background vertical shear. As a result, the spatial pattern and the linear trend in dynamical efficiency are both determined predominantly by vertical shear.
  5. The extent to which the dynamical control affects the PI of TCs depends strongly on changes in vertical shear. The dynamical control modifies the linear trend in THPI by −51.2%, −13.94%, and −13.72% averaged in the MDR over the North Atlantic, eastern Pacific, and southwest Pacific, and 13.92%, 8.31%, and 9.12% averaged in the western North Pacific, north Indian Ocean, and south Indian Ocean.
  6. The modified PI, including the empirical dynamical efficiency in response to the 70-yr buildup of CO2 concentration, increases by 2.41%, 2.33%, 3.26%, 3.39%, 1.03%, and 2.38% averaged over the eastern North Pacific, western North Pacific, north Indian Ocean, south Indian Ocean, southwest Pacific, and off the U.S. southeast coast, respectively, but it changes little in the MDR over the North Atlantic. Therefore, the most significant increase in the modified PI is in the north and south Indian Oceans, while there are only very small changes in the MDR over the North Atlantic and modest changes over the other TC basins. The percentage change off the U.S. southeast coast is consistent with the changes in simulated storm intensity found in the dynamical downscaling study of Knutson et al. (2008).

At least some earlier studies suggest that the changes in THPI are related to the actual changes seen in average storm intensity resulting from climate variability and change (Tonkin et al. 2000; Bister and Emanuel 2002). This study is the first to include the dynamical control on the quantitative TC PI estimation. An obvious limitation is that the dynamical efficiency formulation is empirically determined purely based on statistical analysis of observations for current climate. However, both vertical shear and translational speed are found to be dominant environmental dynamical factors limiting TC intensity from not only observations (DeMaria 1996; Paterson et al. 2005; Zeng et al. 2007, 2008), but also from high-resolution modeling of individual storms (Frank and Ritchie 2001; Wong and Chan 2004; Peng et al. 1999) and from simulations for climate change projections (Vecchi 2007). Therefore, the correction applied to the THPI using the empirical dynamical efficiency, taking into account the negative effects of vertical shear and translational speed, is physically plausible. We have shown that the inclusion of dynamical effects result in PI trend estimates that are most directly comparable to those recently obtained based on different dynamical downscaling approaches by Emanuel et al. (2008) and Knutson et al. (2008).

Finally, it is also found that the TC PI may experience larger interannual variability in the warmed climate than the present climate. A similar increase in interannual variability of the projected TC power dissipation index (PDI) over the western North Pacific has been reported in Stowasser et al. (2007), based on a dynamical downscaling approach using a regional climate model (see their Fig. 12). This could imply an overall increase in the interannual variability of TC activity and intensity in response to the global warming. This issue will be addressed in a future study.

Acknowledgments

We thank the two anonymous reviewers for helpful comments. This work is supported by Chinese National Science Foundation (40775060), Ministry of Science and Technology of China (GYHY200806009) Jiangsu Education Science Foundation (07KJB170065), Jiangsu Government Scholarship for Overseas Studies, and NSF grants ATM-0427128 and ATM-0754029 and ONR grant 000-14-06-10303 awarded to University of Hawaii. Additional support has been provided by the Japan Agency for Marine-Earth Science and Technology (JAMSTEC), NASA, and NOAA through their sponsorship of the International Pacific Research Center at the University of Hawaii.

REFERENCES

  • Anthes, R. A., , R. W. Corell, , G. Holland, , J. W. Hurrell, , M. C. MacCracken, , and K. E. Trenberth, 2005: Hurricanes and global warming—Potential linkages and consequences. Bull. Amer. Meteor. Soc., 88 , 623628.

    • Search Google Scholar
    • Export Citation
  • Barnett, T. P., , D. W. Pierce, , K. M. AchutaRao, , P. J. Gleckler, , B. D. Santer, , J. M. Gregory, , and W. M. Washington, 2005: Penetration of human-induced warming into the World’s Ocean. Science, 309 , 284287.

    • Search Google Scholar
    • Export Citation
  • Bister, M., , and K. A. Emanuel, 1998: Dissipative heating and hurricane intensity. Meteor. Atmos. Phys., 65 , 233240.

  • Bister, M., , and K. A. Emanuel, 2002: Low frequency variability of tropical cyclone potential intensity 1. Interannual to interdecadal variability. J. Geophys. Res., 107 , 4801. doi:10.1029/2001JD000776.

    • Search Google Scholar
    • Export Citation
  • Camp, J. P., , and M. T. Montgomery, 2001: Hurricane maximum intensity: Past and present. Mon. Wea. Rev., 129 , 17041717.

  • DeMaria, M., 1996: The effect of vertical shear on tropical cyclone intensity change. J. Atmos. Sci., 53 , 20762087.

  • DeMaria, M., , and J. Kaplan, 1994: Sea surface temperature and the maximum intensity of Atlantic tropical cyclones. J. Climate, 7 , 13251334.

    • Search Google Scholar
    • Export Citation
  • Elsner, J. B., , J. P. Kossin, , and T. H. Jagger, 2008: The increasing intensity of the strongest tropical cyclones. Nature, 455 , 9295.

  • Emanuel, K. A., 1986: An air–sea interaction theory for tropical cyclones. Part I: Steady-state maintenance. J. Atmos. Sci., 43 , 586604.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., 1987: The dependence of hurricane intensity on climate. Nature, 42 , 483485.

  • Emanuel, K. A., 1995: Sensitivity of tropical cyclone 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., 2000: A statistical analysis of tropical cyclone intensity. Mon. Wea. Rev., 128 , 11391152.

  • Emanuel, K. A., 2007: Environmental factors affecting tropical cyclone power dissipation. J. Climate, 20 , 54975509.

  • Emanuel, K. A., , C. DesAutels, , C. Holloway, , and R. Korty, 2004: Environmental control of tropical cyclone intensity. J. Atmos. Sci., 61 , 843858.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., , R. Sundararajan, , and J. Williams, 2008: Hurricane and global warming: Results from downscaling IPCC AR4 simulations. Bull. Amer. Meteor. Soc., 89 , 347367.

    • Search Google Scholar
    • Export Citation
  • Frank, W. M., , and E. A. Ritchie, 2001: Effects of vertical wind shear on hurricane intensity and structure. Mon. Wea. Rev., 129 , 22492269.

    • Search Google Scholar
    • Export Citation
  • Free, M., , M. Bister, , and K. A. Emanuel, 2004: Potential intensity of tropical cyclones: Comparison of results from radiosonde and reanalysis data. J. Climate, 17 , 17221727.

    • Search Google Scholar
    • Export Citation
  • Goldenberg, S. B., , and L. J. Shapiro, 1996: Physical mechanisms for the association of El Niño and West African rainfall with Atlantic major hurricane activity. J. Climate, 9 , 11691187.

    • Search Google Scholar
    • Export Citation
  • Goldenberg, S. B., , C. W. Landsea, , A. M. Mestas-Nuñez, , and W. M. Gray, 2001: The recent increase in Atlantic hurricane activity: Causes and implications. Science, 293 , 474479.

    • Search Google Scholar
    • Export Citation
  • Henderson-Sellers, A., and Coauthors, 1998: Tropical cyclones and global climate change: A post-IPCC assessment. Bull. Amer. Meteor. Soc., 79 , 1938.

    • Search Google Scholar
    • Export Citation
  • Holland, G. J., 1997: The maximum potential intensity of tropical cyclones. J. Atmos. Sci., 54 , 25192541.

  • Houghton, J. T., , Y-H. Ding, , D. J. Griggs, , M. Noguer, , P. J. van der Linden, , X. Dai, , K. Maskell, , and C. A. Johnson, Eds.,. 2001: Climate Change 2001: The Scientific Basis. Cambridge University Press, 881 pp.

    • Search Google Scholar
    • Export Citation
  • Knutson, T. R., , and R. E. Tuleya, 2004: Impact of CO2-induced warming on simulated hurricane intensity and precipitation: Sensitivity to the choice of climate model and convective parameterization. J. Climate, 17 , 34773495.

    • Search Google Scholar
    • Export Citation
  • Knutson, T. R., , R. E. Tuleya, , and Y. Kurihara, 1998: Simulated increase of hurricane intensities in a CO2-warmed climate. Science, 279 , 10181021.

    • Search Google Scholar
    • Export Citation
  • Knutson, T. R., , J. J. Sirutis, , S. T. Garner, , G. A. Vecchi, , and I. M. Held, 2008: Simulated reduction in Atlantic hurricane frequency under twenty-first-century warming conditions. Nat. Geosci., 1 , 359364. doi:10.1038/ngeo202.

    • Search Google Scholar
    • Export Citation
  • Landsea, C. W., 2005: Hurricanes and global warming. Nature, 438 , E11E13.

  • Landsea, C. W., , B. A. Harper, , K. Hoarau, , and J. A. Kanff, 2006: Can we detect trends in extreme tropical cyclones? Science, 313 , 452454.

    • Search Google Scholar
    • Export Citation
  • Latif, M., , N. Keenlyside, , and J. Bader, 2007: Tropical sea surface temperature, vertical shear, and hurricane development. Geophys. Res. Lett., 34 , L01710. doi:10.1029/2006GL027969.

    • Search Google Scholar
    • Export Citation
  • Malkus, B. S., , and H. Riehl, 1960: On the dynamics and energy transformation in steady-state hurricanes. Tellus, 12 , 120.

  • Meehl, G. A., , C. Covey, , T. Delworth, , M. Latif, , B. McAvaney, , J. F. B. Mitchell, , R. J. Stouffer, , and K. E. Taylor, 2007: THE WCRP CMIP3 multimodel dataset: A new era in climate change research. Bull. Amer. Meteor. Soc., 88 , 13831394.

    • Search Google Scholar
    • Export Citation
  • Nguyen, K. C., , and K. J. E. Walsh, 2001: Interannual, decadal, and transient greenhouse simulation of tropical cyclone–like vortices in a regional climate model of the South Pacific. J. Climate, 14 , 30433054.

    • Search Google Scholar
    • Export Citation
  • Oouchi, K., , J. Yoshimura, , H. Yoshimura, , R. Mizuta, , S. Kusunoki, , and A. Noda, 2006: Tropical cyclone climatology in a global-warming climate as simulated in a 20 km-mesh global atmospheric model: Frequency and wind intensity analyses. J. Meteor. Soc. Japan, 84 , 259276.

    • Search Google Scholar
    • Export Citation
  • Paterson, L. A., , B. N. Hanstrum, , and N. E. Davidson, 2005: Influence of environmental vertical wind shear on the intensity of hurricane-strength tropical cyclones in the Australian region. Mon. Wea. Rev., 133 , 36443660.

    • Search Google Scholar
    • Export Citation
  • Peng, M. S., , B-F. Jeng, , and R. T. Williams, 1999: A numerical study on tropical cyclone intensification. Part I: Beta effect and mean flow effect. J. Atmos. Sci., 56 , 14041423.

    • Search Google Scholar
    • Export Citation
  • Pielke Jr., R. A., , C. Landsea, , M. Mayfield, , J. Laver, , and R. Pasch, 2005: Hurricane and global warming. Bull. Amer. Meteor. Soc., 80 , 20272033.

    • Search Google Scholar
    • Export Citation
  • Rayner, N. A., , D. E. Parker, , E. B. Horton, , C. K. Folland, , L. V. Alexander, , D. P. Rowell, , E. C. Kent, , and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108 , 4407. doi:10.1029/2002JD002670.

    • Search Google Scholar
    • Export Citation
  • Rotunno, R., , and K. A. Emanuel, 1987: An air–sea interaction theory for tropical cyclones. Part II: Evolutionary study using a nonhydrostatic axisymmetric numerical model. J. Atmos. Sci., 44 , 542561.

    • Search Google Scholar
    • Export Citation
  • Santer, B. D., and Coauthors, 2006: Forced and unforced ocean temperature changes in Atlantic and Pacific tropical cyclogenesis regions. Proc. Natl. Acad. Sci. USA, 103 , 1390513910. doi:10.1073/PNAS.0602861103.

    • Search Google Scholar
    • Export Citation
  • Saunders, M. A., , and A. R. Harris, 1997: Statistical evidence links of exceptional 1995 Atlantic hurricane season to record sea warming. Geophys. Res. Lett., 24 , 12551258.

    • Search Google Scholar
    • Export Citation
  • Schade, L. R., 2000: Tropical cyclone intensity and sea surface temperature. J. Atmos. Sci., 57 , 31223130.

  • Schade, L. R., , and K. A. Emanuel, 1999: The ocean’s effect on the intensity of tropical cyclones: Results from a simple coupled atmosphere–ocean model. J. Atmos. Sci., 56 , 642651.

    • Search Google Scholar
    • Export Citation
  • Shen, W., , R. E. Tuleya, , and I. Ginis, 2000: A sensitivity study of thermodynamic environment on GFDL model hurricane intensity: Implications for global warming. J. Climate, 13 , 109121.

    • Search Google Scholar
    • Export Citation
  • Smith, T. M., , and R. W. Reynolds, 2004: Improved extended reconstruction of SST (1954–1997). J. Climate, 17 , 24662477.

  • Sobel, A. H., , I. M. Held, , and C. S. Bretherton, 2002: The ENSO signal in tropical tropospheric temperature. J. Climate, 15 , 27022706.

    • Search Google Scholar
    • Export Citation
  • Stowasser, M., , Y. Wang, , and K. Hamilton, 2007: Tropical cyclone changes in the western North Pacific in a global warming scenario. J. Climate, 20 , 23782396.

    • Search Google Scholar
    • Export Citation
  • Sugi, M., , A. Noda, , and N. Sato, 2002: Influence of the global warming on tropical cyclone climatology: An experiment with the JMA global model. J. Meteor. Soc. Japan, 80 , 249272.

    • Search Google Scholar
    • Export Citation
  • Tang, B. H., , and J. D. Neelin, 2004: ENSO influence on Atlantic hurricanes via tropospheric warming. Geophys. Res. Lett., 31 , L24204. doi:10.1029/2004GL021072.

    • Search Google Scholar
    • Export Citation
  • Tonkin, H., , G. J. Holland, , N. Holbrook, , and A. Henderson-Sellers, 2000: An evaluation of thermodynamic estimates of climatological maximum potential tropical cyclone intensity. Mon. Wea. Rev., 128 , 746762.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K., 2005: Uncertainty in hurricanes and global warming. Science, 308 , 17531754.

  • Vecchi, G. A., 2007: Effect of remote sea surface temperature change on potential tropical cyclone intensity. Nature, 450 , 10661070.

  • Vecchi, G. A., , and B. J. Soden, 2007: Increased tropical Atlantic wind shear in model projections of global warming. Geophys. Res. Lett., 34 , L08702. doi:10.1029/2006GL028905.

    • Search Google Scholar
    • Export Citation
  • Vecchi, G. A., , B. J. Soden, , A. T. Wittenberg, , I. M. Held, , A. Leetmaa, , and M. J. Harrison, 2006: Weakening of tropical Pacific atmospheric circulation due to anthropogenic forcing. Nature, 441 , 7376.

    • Search Google Scholar
    • Export Citation
  • Wang, Y., 2002: Vortex Rossby waves in a numerically simulated tropical cyclone. Part II: The role in tropical cyclone structure and intensity changes. J. Atmos. Sci., 59 , 12391262.

    • Search Google Scholar
    • Export Citation
  • Wang, Y., , and G. J. Holland, 1996: Tropical cyclone motion and evolution in vertical shear. J. Atmos. Sci., 53 , 33133332.

  • Wang, Y., , and C-C. Wu, 2004: Current understanding of tropical cyclone structure and intensity changes—A review. Meteor. Atmos. Phys., 87 , 257278.

    • Search Google Scholar
    • Export Citation
  • Wang, Y., , M. T. Montgomery, , and B. Wang, 2004: How much vertical shear can a tropical cyclone resist? Bull. Amer. Meteor. Soc., 85 , 661662.

    • Search Google Scholar
    • Export Citation
  • Wong, M. L. M., , and J. C. L. Chan, 2004: Tropical cyclone intensity in vertical wind shear. J. Atmos. Sci., 61 , 18591876.

  • Yang, B., , Y. Wang, , and B. Wang, 2007: The effect of internally generated inner core asymmetric structure on tropical cyclone intensity. J. Atmos. Sci., 64 , 11651188.

    • Search Google Scholar
    • Export Citation
  • Zeng, Z., , Y. Wang, , and C-C. Wu, 2007: Environmental dynamical control of tropical cyclone intensity—An observational study. Mon. Wea. Rev., 135 , 3859.

    • Search Google Scholar
    • Export Citation
  • Zeng, Z., , L-S. Chen, , and Y. Wang, 2008: An observational study of environmental dynamical control of tropical cyclone intensity in the North Atlantic. Mon. Wea. Rev., 136 , 33073322.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Time series of the annual-mean SST averaged over the topical oceans (30°S–30°N, 0°–360°) for the twentieth-century runs (20C3M) and the NOAA ERSST. Line 19: ERSST; line 20: all-model ensemble; line 21: ensemble with CNRM-CM3, CSIRO Mk3.0, and GISS-ER removed. The order from lines 1 to 18 is the same as that in Table 1. Ensemble mean from models (a) 25.31°C for 1900–99 and (b) 25.44°C for 1948–99 vs the NOAA ERSST mean 25.64° and 25.76°C (Table 1).

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Fig. 2.
Fig. 2.

(a) The 70-yr changes associated with the linear trends in the annual mean SST (°C) of the model ensemble mean in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario; and (b) the composite mean SST anomaly (°C) in June in an El Niño year. Contours in (a) show the ensemble-mean background (initial) annual-mean SST.

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Fig. 3.
Fig. 3.

Time series of CO2 concentration (dashed) used in the IPCC AR4 1pctto2× scenario and the annual-mean SST (solid) of the 15 CGCM ensembles averaged over the tropical oceans (30°S–30°N, 0°–360°).

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Fig. 4.
Fig. 4.

Spatial distribution of the 70-yr changes associated with the linear trends in the outflow layer temperature (°C) in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario from the 15-CGCM ensemble mean for the (a) JASO mean and (b) JFM season. Contours show the initial background outflow layer temperature (°C).

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Fig. 5.
Fig. 5.

The 70-yr relative changes (%) associated with the linear trends in thermodynamic efficiency in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario for the (a) JASO mean and (b) JFM mean based on the 15-CGCM ensemble mean. Contours show the corresponding initial background thermodynamic efficiency.

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Fig. 6.
Fig. 6.

The 70-yr relative changes (%) in the root of disequilibrium enthalpy between the ocean and atmosphere in response to the increasing CO2 concentration in the IPCC AR4 ipctto2× scenario for (a) JASO and (b) JFM from the 15-CGCM ensemble mean. Contours show the initial background fields (k*0k, m s−1).

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Fig. 7.
Fig. 7.

The 70-yr changes associated with the linear trends in the thermodynamic potential intensity of TCs (m s−1, THPI) in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario for the (a) JASO mean and (b) JFM mean based on the 15-CGCM ensemble. Contours in (a) and (b) show the corresponding initial background THPI (m s−1). Boxes indicate the major TC activity basins (see Table 1).

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Fig. 8.
Fig. 8.

Time series of THPI changes from the corresponding initial values: (a) lines 1 and 2 for the south Indian Ocean; lines 3 and 4 for the north Indian Ocean; line 5 and 6 for the eastern Pacific; lines 7 and 8 for the western North Pacific; (b) Lines 1 and 2 are the same as lines 7 and 8 in (a); lines 3 and 4 for the southwest Pacific; lines 5 and 6 for the North Atlantic. The THPI change averaged over all basins (solid curve) and the linear regression of THPI relative to the initial value (dot curve).

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Fig. 9.
Fig. 9.

The 70-yr changes (%) associated with the linear trends in the model ensemble-mean vertical shear in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario for the (a) JASO mean and (b) JFM mean. Contours show the initial background vertical shear (m s−1).

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Fig. 10.
Fig. 10.

The 70-yr changes associated with the linear trends in model ensemble mean translational speed (Vtrans) in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario for the (a) JASO mean and (b) corresponding changes in percentage of the absolute (Vtrans − 5). Contours show the (a) initial background Vtrans (m s−1) and (b) absolute (Vtrans − 5; m s−1).

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Fig. 11.
Fig. 11.

The 70-yr changes (%) associated with the linear trends in model ensemble mean dynamical efficiency in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario for the (a) JASO mean and (b) JFM mean. Contours show the initial background dynamical efficiency.

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Fig. 12.
Fig. 12.

The 70-yr changes (m s−1) associated with the linear trends in the TC PI with dynamical control in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario from the 15-CGCM ensemble for the (a) JASO mean and (c) JFM mean. (b) The difference between the changes in the modified PI and the THPI for JASO mean. Contours show the initial background modified PI (m s−1).

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Fig. 13.
Fig. 13.

The 70-yr changes (%) associated with the linear trends in various control parameters and PI of TCs averaged in six individual TC activity basins in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario from the 15-CGCM ensemble: the thermodynamic potential intensity (THPI, black), thermodynamic efficiency (green), dynamical efficiency (blue), and modified potential intensity (PI, red).

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Fig. 14.
Fig. 14.

The 70-yr change (%) in the PI modified by the empirical dynamical efficiency in six individual TC basins in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario from 15 CGCMs. The different color bars correspond to the different CGCMs as given in the legends. The values of the percentage changes averaged across all models are also given.

Citation: Journal of Climate 23, 6; 10.1175/2009JCLI2843.1

Table 1.

List of CGCMs that participated in the IPCC AR4 and the corresponding RMSE of the simulated global and tropical mean SST of each model for the twentieth century against those from the NOAA ERSST.

Table 1.
Table 2.

Definition of major TC basins and the corresponding TC activity seasons and the corresponding annual mean SST and TC seasonal mean SST increase as a result of the increasing CO2 concentration in the first 70 yr in the IPCC AR4 1pctto2× scenario from 15-CGCM ensemble mean.

Table 2.
Table 3.

The 70-yr changes associated with the linear trends in THPI, modified PI, and thermodynamic and dynamic control parameters of PI of TCs over individual basins in response to the increasing CO2 concentration in the IPCC AR4 1pctto2× scenario from the 15-CGCM ensemble mean.

Table 3.

1

Note that the RMSE of INM-CM3.0 is larger than CSIRO Mk3.0 during 1900–99 but is smaller than CSIRO Mk3.0 during 1948–99.

2

Note that the simulation length of CSIRO Mk3.5, MIROC3.2(hires), HadCM3, and HadGEM1 was only 80 yr, and thus only 11 models are available for the ensemble mean beyond 70 yr in Fig. 8.

Save