• Akima, H., 1970: A new method of interpolation and smooth curve fitting based on local procedures. J. Assoc. Comput. Mach., 17, 589600.

    • Search Google Scholar
    • Export Citation
  • Akperov, M. G., , and I. I. Mokhov, 2010: A comparative analysis of the method of extratropical cyclone identification. Atmos. Oceanic Phys., 46, 574590.

    • Search Google Scholar
    • Export Citation
  • Allen, J. T., , A. B. Pezza, , and M. T. Black, 2010: Explosive cyclogenesis: A global climatology comparing multiple reanalyses. J. Climate, 23, 64686484.

    • Search Google Scholar
    • Export Citation
  • Bengtsson, L., , S. Hagemann, , and K. I. Hodges, 2004: Can climate trends be calculated from reanalysis data? J. Geophys. Res., 109, D11111, doi:10.1029/2004JD004536.

    • Search Google Scholar
    • Export Citation
  • Blender, R., , and M. Schubert, 2000: Cyclone tracking in different spatial and temporal resolutions. Mon. Wea. Rev., 128, 377384.

  • Bromwich, D. H., , R. L. Fogt, , K. I. Hodges, , and J. E. Walsh, 2007: A tropospheric assessment of the ERA-40, NCEP, and JRA-25 global reanalyses in the polar regions. J. Geophys. Res., 112, D10111, doi:10.1029/2006JD007859.

    • Search Google Scholar
    • Export Citation
  • Brunke, M. A., , Z. Wang, , X. Zeng, , M. Bosilovich, , and C.-L. Shie, 2011: An assessment of the uncertainties in ocean surface turbulent fluxes in 11 reanalysis, satellite-derived, and combined global datasets. J. Climate, 24, 54695493.

    • Search Google Scholar
    • Export Citation
  • Chen, J., , A. D. Del Genio, , B. E. Carlson, , and M. G. Bosilovich, 2008: The spatiotemporal structure of twentieth-century climate variations in observations and reanalyses. Part I: Long-term trend. J. Climate, 21, 26112633.

    • Search Google Scholar
    • Export Citation
  • Cullather, R. I., , and M. G. Bosilovich, 2011: The moisture budget of the polar atmosphere in MERRA. J. Climate, 24, 28612879.

  • Cullather, R. I., , and M. G. Bosilovich, 2012: The energy budget of the polar atmosphere in MERRA. J. Climate, 25, 524.

  • Dacre, H. F., , M. K. Hawcroft, , M. A. Stringer, , and K. I. Hodges, 2012: An extratropical cyclone atlas: A tool for illustrating cyclone structure and evolution characteristics. Bull. Amer. Meteor. Soc., 93, 1497–1502.

    • Search Google Scholar
    • Export Citation
  • Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553597.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K., 2005: Genesis and maintenance of “Mediterranean hurricanes.” Adv. Geosci., 2, 217220.

  • Fink, A. H., , T. Brücher, , E. Ermert, , A. Krüger, , and J. G. Pinto, 2009: The European Storm Kyrill in January 2007: Synoptic evolution and considerations with respect to climate change. Nat. Hazards Earth Syst. Sci., 9, 405423.

    • Search Google Scholar
    • Export Citation
  • Gulev, S. K., , O. Zolina, , and S. Grigoriev, 2001: Extratropical cyclone variability in the Northern Hemisphere winter from the NCEP/NCAR-Reanalysis data. Climate Dyn., 17, 795809.

    • Search Google Scholar
    • Export Citation
  • Hodges, K. I., , B. J. Hoskins, , J. Boyle, , and C. Thorncroft, 2003: A comparison of recent reanalysis datasets using objective feature tracking: Storm tracks and tropical easterly waves. Mon. Wea. Rev., 131, 20122037.

    • Search Google Scholar
    • Export Citation
  • Hodges, K. I., , R. W. Lee, , and L. Bengtsson, 2011: A comparison of extratropical cyclones in recent reanalyses ERA-Interim, NASA MERRA, NCEP CFSR, and JRA-25. J. Climate, 24, 48884906.

    • Search Google Scholar
    • Export Citation
  • Hoskins, B., , and K. Hodges, 2002: New perspectives on the Northern Hemisphere winter storm tracks. J. Atmos. Sci., 59, 10411061.

  • Jung, T., , S. K. Gulev, , I. Rudeva, , and V. Soloviov, 2006: Sensitivity of extratropical cyclone characteristics to horizontal resolution in the ECMWF model. Quart. J. Roy. Meteor. Soc., 132, 18391857.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471.

  • Kanamitsu, M., , W. Ebisuzaki, , J. Woollen, , S.-K. Yang, , J. J. Hnilo, , M. Fiorion, , and J. Potter, 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 16311643.

    • Search Google Scholar
    • Export Citation
  • König, W. R., , R. Sausen, , and F. Sielmann, 1993: Objective identification of cyclones in GCM simulations. J. Climate, 6, 22172231.

  • Liberato, M. R., , J. G. Pinto, , I. F. Trigo, , and R. M. Trigo, 2011: Klaus—An exceptional winter storm over northern Iberia and southern France. Weather, 66, 330334.

    • Search Google Scholar
    • Export Citation
  • Lim, E.-P., , and I. Simmonds, 2002: Explosive cyclone development in the Southern Hemisphere and a comparison with Northern Hemisphere events. Mon. Wea. Rev., 130, 21882209.

    • Search Google Scholar
    • Export Citation
  • Lionello, P., , and F. Giorgi, 2007: Winter precipitation and cyclones in the Mediterranean region: Future climate scenarios in a regional simulation. Adv. Geosci., 12, 153158.

    • Search Google Scholar
    • Export Citation
  • Livezey, R. E., , and W. Y. Chen, 1983: Statistical field significance and its determination by Monte Carlo techniques. Mon. Wea. Rev., 111, 4659.

    • Search Google Scholar
    • Export Citation
  • Löptien, U., , O. Zolina, , S. K. Gulev, , M. Latif, , and V. Soloviov, 2008: Cyclone life cycle characteristics over the Northern Hemisphere in coupled GCMs. Climate Dyn., 31, 507532.

    • Search Google Scholar
    • Export Citation
  • Mailier, P. J., , D. B. Stephenson, , C. A. T. Ferro, , and K. I. Hodges, 2006: Serial clustering of extratropical cyclones. Mon. Wea. Rev., 134, 22242240.

    • Search Google Scholar
    • Export Citation
  • Murray, R. J., , and I. Simmonds, 1991a: A numerical scheme for tracking cyclone centres from digital data. Part I: Development and operation of the scheme. Aust. Meteor. Mag., 39, 155166.

    • Search Google Scholar
    • Export Citation
  • Murray, R. J., , and I. Simmonds, 1991b: A numerical scheme for tracking cyclone centres from digital data. Part II: Application to January and July general circulation model simulations. Aust. Meteor. Mag., 39, 167180.

    • Search Google Scholar
    • Export Citation
  • Neu, U., and Coauthors, 2013: IMILAST—A community effort to intercompare extratropical cyclone detection and tracking algorithms. Bull. Amer. Meteor. Soc.,94, 529–547.

  • Onogi, K., and Coauthors, 2007: The JRA-25 reanalysis. J. Meteor. Soc. Japan, 85, 369432.

  • Orlanski, I., 1998: Poleward deflection of storm tracks. J. Atmos. Sci., 55, 25772602.

  • Pinto, J. G., , U. Ulbrich, , G. C. Leckebusch, , T. Spangehl, , M. Reyers, , and S. Zacharias, 2007: Changes in storm track and cyclone activity in three SRES ensemble experiments with the ECHAM5/MPI-OM1 GCM. Climate Dyn., 29, 195210.

    • Search Google Scholar
    • Export Citation
  • Pinto, J. G., , S. Zacharias, , A. H. Fink, , G. C. Leckebusch, , and U. Ulbrich, 2009: Factors contributing to the development of extreme North Atlantic cyclones and their relationship with the NAO. Climate Dyn., 32, 711737.

    • Search Google Scholar
    • Export Citation
  • Raible, C. C., 2007: On the relation between extremes of midlatitude cyclones and the atmospheric circulation using ERA-40. Geophys. Res. Lett., 34, L07703, doi:10.1029/2006GL029084.

    • Search Google Scholar
    • Export Citation
  • Raible, C. C., , and R. Blender, 2004: Northern Hemisphere midlatitude cyclone variability in GCM simulations with different ocean representations. Climate Dyn., 22, 239248.

    • Search Google Scholar
    • Export Citation
  • Raible, C. C., , P. Della-Marta, , C. Schwierz, , H. Wernli, , and R. Blender, 2008: Northern Hemisphere extratropical cyclones: A comparison of detection and tracking methods and different reanalyses. Mon. Wea. Rev., 136, 880897.

    • Search Google Scholar
    • Export Citation
  • Raible, C. C., , H. Saaroni, , B. Ziv, , and M. Wild, 2010: Winter synoptic-scale variability over the Mediterranean Basin under future climate conditions as simulated by the ECHAM5. Climate Dyn., 35, 473488.

    • Search Google Scholar
    • Export Citation
  • Rančić, M., , J. C. Derber, , D. Parrish, , R. Treadon, , and D. T. Kleist, 2008: The development of the first-order time extrapolation to the observation (FOTO) method and its application in the NCEP global data assimilation system. Proc. 12th Symp. on Integrated Observing and Assimilation Systems for the Atmosphere, Oceans, and Land Surface (IOAS-AOLS), New Orleans, LA, Amer. Meteor. Soc., J6.1. [Available online at http://ams.confex.com/ams/pdfpapers/131816.pdf.]

  • Rienecker, M. M., and Coauthors, 2011: MERRA: NASA's Modern-Era Retrospective Analysis for Research and Applications. J. Climate, 24, 36243648.

    • Search Google Scholar
    • Export Citation
  • Rogers, E., , and L. F. Bosart, 1986: An investigation of explosively deepening oceanic cyclones. Mon. Wea. Rev., 114, 702718.

  • Rudeva, I., , and S. K. Gulev, 2007: Climatology of cyclone size characteristics and their changes during the cyclone life cycle. Mon. Wea. Rev., 135, 25682587.

    • Search Google Scholar
    • Export Citation
  • Rudeva, I., , and S. K. Gulev, 2011: Composite analysis of North Atlantic extratropical cyclones in NCEP–NCAR reanalysis data. Mon. Wea. Rev., 139, 14191446.

    • Search Google Scholar
    • Export Citation
  • Saha, S., and Coauthors, 2010: The NCEP Climate Forecast System. J. Climate, 19, 34833517.

  • Sanders, F., , and J. R. Gyakum, 1980: Synoptic-dynamic climatology of the “bomb.” Mon. Wea. Rev., 108, 15891606.

  • Screen, J. A., , I. Simmonds, , and K. Keay, 2011: Dramatic interannual changes of perennial Arctic sea ice linked to abnormal summer storm activity. J. Geophys. Res., 116, D15105, doi:10.1029/2011JD015847.

    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., , and A. P. Barrett, 2008: The summer cyclone maximum over the central Arctic Ocean. J. Climate, 21, 10481065.

  • Simmonds, I., , and K. Keay, 2000a: Mean Southern Hemisphere extratropical cyclone behavior in the 40-year NCEP–NCAR reanalysis. J. Climate, 13, 873885.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., , and K. Keay, 2000b: Variability of Southern Hemisphere extratropical cyclone behavior, 1958–97. J. Climate, 13, 550561.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., , and K. Keay, 2009: Extraordinary September Arctic sea ice reductions and their relationships with storm behavior over 1979–2008. Geophys. Res. Lett., 36, L19715, doi:10.1029/2009GL039810.

    • Search Google Scholar
    • Export Citation
  • Simmons, A. J., , S. Uppala, , D. Dee, , and S. Kobayashi, 2007: ERA-Interim: New ECMWF reanalysis products from 1989 onwards. ECMWF Newsletter, No. 110, ECMWF, Reading, United Kingdom, 25–35.

  • Simmons, A. J., , K. M. Willett, , P. D. Jones, , P. W. Thorne, , and D. P. Dee, 2010: Low-frequency variations in surface atmospheric humidity, temperature, and precipitation: Inferences from reanalyses and monthly gridded observational data sets. J. Geophys. Res., 115, D01110, doi:10.1029/2009JD012442.

    • Search Google Scholar
    • Export Citation
  • Sinclair, M. R., 1997: Objective identification of cyclones and their circulation, intensity, and climatology. Wea. Forecasting, 12, 595612.

    • Search Google Scholar
    • Export Citation
  • Sinclair, M. R., , and I. G. Watterson, 1999: Objective assessment of extratropical weather systems in simulated climates. J. Climate, 12, 34673485.

    • Search Google Scholar
    • Export Citation
  • Sterl, A., 2004: On the (in)homogeneity of reanalysis products. J. Climate, 17, 38663873.

  • Stewart, R. E., , and N. R. Donaldson, 1989: On the nature of rapidly deepening Canadian East Coast winter storms. Atmos.–Ocean, 27, 87107.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., , J. T. Fasullo, , and J. Mackaro, 2011: Atmospheric moisture transports from ocean to land and global energy flows in reanalyses. J. Climate, 24, 49074924.

    • Search Google Scholar
    • Export Citation
  • Trigo, I. F., 2006: Climatology and interannual variability of storm-tracks in the Euro-Atlantic sector: A comparison between ERA-40 and NCEP/NCAR reanalyses. Climate Dyn., 26, 127143.

    • Search Google Scholar
    • Export Citation
  • Ulbrich, U., , and M. Christoph, 1999: A shift of the NAO and increasing storm track activity over Europe due to anthropogenic greenhouse gas forcing. Climate Dyn., 15, 551559.

    • Search Google Scholar
    • Export Citation
  • Ulbrich, U., , T. Brücher, , A. H. Fink, , G. C. Leckebusch, , A. Krüger, , and J. G. Pinto, 2003: The central European floods in August 2002: Part I—Rainfall periods and flood development. Weather, 58, 371376.

    • Search Google Scholar
    • Export Citation
  • Ulbrich, U., , J. G. Pinto, , H. Kupfer, , G. C. Leckebusch, , T. Spangehl, , and M. Reyers, 2008: Changing Northern Hemisphere storm tracks in an ensemble of IPCC climate change simulations. J. Climate, 21, 16691679.

    • Search Google Scholar
    • Export Citation
  • Uppala, S. M., and Coauthors, 2005: The ERA-40 Re-Analysis. Quart. J. Roy. Meteor. Soc., 131, 29613012.

  • Wang, X. L., , H. Wan, , and V. R. Swail, 2006: Observed changes in cyclone activity in Canada and their relationships to major circulation regimes. J. Climate, 19, 896915.

    • Search Google Scholar
    • Export Citation
  • White, G., 2000: Long-term trends in the NCEP/NCAR reanalysis. Proc. Second Int. Conf. on Reanalyses, Reading, United Kingdom, WMO, 54–57.

  • Woollings, T., 2008: Vertical structure of anthropogenic zonal-mean atmospheric circulation change. Geophys. Res. Lett., 35, L19702, doi:10.1029/2008GL034883.

    • Search Google Scholar
    • Export Citation
  • Woollings, T., , and M. Blackburn, 2012: The North Atlantic jet stream under climate change and its relation to the NAO and EA patterns. J. Climate, 25, 886902.

    • Search Google Scholar
    • Export Citation
  • Yau, M. K., , and M. Jean, 1989: Synoptic aspects and physical processes in the rapidly intensifying cyclone of 6–8 March 1986. Atmos.–Ocean, 27, 5986.

    • Search Google Scholar
    • Export Citation
  • Yin, J. H., 2005: A consistent poleward shift of the storm tracks in simulations of 21st century climate. Geophys. Res. Lett., 32, L18701, doi:10.1029/2005GL023684.

    • Search Google Scholar
    • Export Citation
  • Zolina, O., , and S. K. Gulev, 2002: Improving the accuracy of mapping cyclone numbers and frequencies. Mon. Wea. Rev., 130, 748759.

  • Zolina, O., , and S. K. Gulev, 2003: Synoptic variability of ocean–atmosphere turbulent fluxes associated with atmospheric cyclones. J. Climate, 16, 30233041.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Annual climatologies (1979–2010) of cyclone numbers in NASA-MERRA (a) before and (b) after filtering of cyclones over elevated terrain, and differences in the annual cyclone numbers between NASA-MERRA and (c) NCEP–DOE, (d) JRA-25, (e) ERA-Interim, and (f) NCEP-CFSR. Units are tracks per year per circle with a radius of 2° latitude (equivalent to approximately 155 000 km2).

  • View in gallery

    Absolute variance in the number of cyclones computed using Eq. (1) for the total number of cyclones in the (a) DJF and (b) JJA during 1979–2010. Units are as in Fig. 1.

  • View in gallery

    Total annual numbers of cyclones (season−1) of different intensity identified in five reanalyses over the (a),(b) continents and (c),(d) oceans for (left) DJF and (right) JJA during 1979–2010, plotted as a function of spectral resolution of reanalysis. Colors define minimum central pressure of cyclone. Spectral resolution is shown at the x axis for four spectral reanalyses; the location of NASA-MERRA (nonspectral model) is given tentatively. Error bars correspond to interannual std dev of the number of cyclones of different types in different reanalyses.

  • View in gallery

    Statistical significance of differences in number of cyclones between different products estimated according to a Student's t test during 1979–2010. Estimates are performed for (top) ocean areas and (bottom) continents for: (a),(d) deep (<980 hPa); (b),(e) moderate (980–1000 hPa); and (c),(f) shallow (>1000 hPa) cyclones. Estimates below and above the main diagonals in each of the six panels correspond to DJF and JJA, respectively. Differences significant at 90%, 95%, and 99% levels are marked by different tones of blue, with statistically insignificant differences remaining uncolored.

  • View in gallery

    (left) DJF and (right) JJA climatological (1979–2010) occurrence histograms for the cyclone (a),(b) lifetime and (c),(d) propagation speed in five reanalyses.

  • View in gallery

    Time series of (a) the total annual number of cyclones over the NH and (b),(c) the number of very deep (<960 hPa) cyclones over the North Atlantic and North Pacific, respectively. Thin lines correspond to interannual values and thick lines show 5-yr running means.

  • View in gallery

    Changes in the number of cyclones of different intensities during the 32-yr period (1979–2010) for (a) DJF and (b) JJA in different reanalyses.

  • View in gallery

    (a),(b) Estimates of the linear trends of the number of cyclones per 155 000 km2 (decade−1) during 1979–2010 in NASA-MERRA for DJF and JJA, respectively. Black dots denote trends that are statistically significant at the 90% level according to a Student's t test. (c),(d) The number of reanalyses simultaneously indicating statistically significant (85% level) linear trends of the same sign during 1979–2010 for DJF and JJA, respectively. In (c) and (d), the hatched areas outside of the major storm tracks indicate where the climatological number of cyclones is small (<2 season−1 over 155 000 km2 per season).

  • View in gallery

    (top) Annual normalized occurrence anomalies of cyclone central pressure in NASA-MERRA computed using Eq. (2) and (bottom) the number of reanalyses simultaneously indicating anomalies of the same sign for a given year and central pressure for the same areas for: (a),(e) the Atlantic; (b),(f) the Arctic; (c),(g) the Pacific; and (d),(h) Europe.

  • View in gallery

    Estimates of pair correlations between the number of cyclones of different intensities for all possible pairs of datasets for the winter (blue circles) and summer (red circles) for NH (a) oceans and (b) continents. Gray dotted lines denote the confidence limit for correlation coefficients at the 99% level. Text captions denote the pairs showing the highest and lowest correlations.

  • View in gallery

    Spatial distribution of the point correlations between the total number of cyclones per year identified for: (a),(e) NASA-MERRA and NCEP–DOE; (b),(f) NASA-MERRA and JRA-25; (c),(g) NASA-MERRA and ERA-Interim; and (d),(h) NASA-MERRA and NCEP-CFSR, for (top) DJF and (bottom) JJA during 1979–2010. Only correlations higher than 0.5 are shown. The 99% significance level of the correlation is 0.49.

  • View in gallery

    Winter time series of the anomalies of the total number of rapidly intensifying cyclones [core pressure drops of more than 24 hPa (24 h)−1] in the (a) North Atlantic and (b) North Pacific in different reanalyses.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 79 79 16
PDF Downloads 49 49 10

Comparing Cyclone Life Cycle Characteristics and Their Interannual Variability in Different Reanalyses

View More View Less
  • 1 P. P. Shirshov Institute of Oceanology, and Lomonosov Moscow State University, Moscow, Russia
  • | 2 P. P. Shirshov Institute of Oceanology, Moscow, Russia, and University of Melbourne, Melbourne, Australia
  • | 3 Lomonosov Moscow State University, Moscow, Russia
© Get Permissions
Full access

Abstract

Characteristics of Northern Hemisphere extratropical cyclone activity were compared for five concurrent reanalyses: the NCEP–U.S. Department of Energy (DOE) reanalysis (herein NCEP–DOE), the Japanese 25-year Reanalysis Project (JRA-25), the ECMWF Interim Re-Analysis (ERA-Interim), the National Aeronautics and Space Administration's Modern-Era Retrospective Analysis for Research and Applications (NASA-MERRA), and the NCEP Climate Forecast System Reanalysis (NCEP-CFSR), for the period 1979–2010 using a single cyclone tracking algorithm. The total number of cyclones, ranging from 1400 to more than 1800 yr−1, was found to depend strongly on the spatial resolution of the respective reanalysis. The largest cyclone population was identified using NASA-MERRA data, which also showed the highest occurrence of very deep cyclones. Of the reanalyses, two (NCEP–DOE and ERA-Interim) are associated with statistically significant positive trends in the total number of cyclones from 1% to 2% decade−1. These trends result from moderate and shallow cyclones contributing to approximately 90% of the total cyclone count on average. The number of very deep cyclones (<960 hPa) in the North Atlantic increased in most reanalyses until 1990 and then declined during the last decade. In the North Pacific, the number of these events reached a peak in 2000 and then decreased during the last decade. The winter pattern is characterized by robust trends in cyclone numbers, with an enhancement of the North Atlantic storm track and a weakening of the North Pacific subtropical storm track. In the summer, there is a robust intensification of the Mediterranean storm track and a decrease in counts over the North Atlantic. Interannual variability and decadal-scale variations of the cyclone counts are highly correlated among the reanalyses, with the greatest agreement in moderate and deep cyclones.

Corresponding author address: Natalia Tilinina, P. P. Shirshov Institute of Oceanology, RAS, 36 Nakhimovsky Ave., 117997 Moscow, Russia. E-mail: tilinina@sail.msk.ru

Abstract

Characteristics of Northern Hemisphere extratropical cyclone activity were compared for five concurrent reanalyses: the NCEP–U.S. Department of Energy (DOE) reanalysis (herein NCEP–DOE), the Japanese 25-year Reanalysis Project (JRA-25), the ECMWF Interim Re-Analysis (ERA-Interim), the National Aeronautics and Space Administration's Modern-Era Retrospective Analysis for Research and Applications (NASA-MERRA), and the NCEP Climate Forecast System Reanalysis (NCEP-CFSR), for the period 1979–2010 using a single cyclone tracking algorithm. The total number of cyclones, ranging from 1400 to more than 1800 yr−1, was found to depend strongly on the spatial resolution of the respective reanalysis. The largest cyclone population was identified using NASA-MERRA data, which also showed the highest occurrence of very deep cyclones. Of the reanalyses, two (NCEP–DOE and ERA-Interim) are associated with statistically significant positive trends in the total number of cyclones from 1% to 2% decade−1. These trends result from moderate and shallow cyclones contributing to approximately 90% of the total cyclone count on average. The number of very deep cyclones (<960 hPa) in the North Atlantic increased in most reanalyses until 1990 and then declined during the last decade. In the North Pacific, the number of these events reached a peak in 2000 and then decreased during the last decade. The winter pattern is characterized by robust trends in cyclone numbers, with an enhancement of the North Atlantic storm track and a weakening of the North Pacific subtropical storm track. In the summer, there is a robust intensification of the Mediterranean storm track and a decrease in counts over the North Atlantic. Interannual variability and decadal-scale variations of the cyclone counts are highly correlated among the reanalyses, with the greatest agreement in moderate and deep cyclones.

Corresponding author address: Natalia Tilinina, P. P. Shirshov Institute of Oceanology, RAS, 36 Nakhimovsky Ave., 117997 Moscow, Russia. E-mail: tilinina@sail.msk.ru

1. Introduction

Cyclone activity plays an important role in atmospheric circulation and the advection of heat and moisture in midlatitudes. Extratropical cyclones are also responsible for extreme weather conditions that lead to natural hazards, such as wind storms and flooding (Ulbrich et al. 2003; Fink et al. 2009; Pinto et al. 2009). During the past decade, leading meteorological centers have produced dynamically consistent continuous datasets of atmospheric state variables known as reanalyses. These products cover time periods of more than three decades and provide an important data source for the analysis of the observed climate variability. One application of these data is the analysis of cyclone activity quantified through numerical cyclone tracking.

Uncertainties in the characteristics of cyclone activity result from the application of different tracking methods and the use of different data. One source of uncertainty relates to the differences in the cyclone detection and tracking methods (e.g., Sinclair 1997; Sinclair and Watterson 1999; Simmonds and Keay 2000a,b; Hodges et al. 2003; Rudeva and Gulev 2007; Raible et al. 2008). This particular aspect was comprehensively quantified in the Intercomparison of Mid Latitude Storm Diagnostics (IMILAST) project (Neu et al. 2013), which analyzed the results of applying different detection and tracking methods to a single reanalysis. The largest disagreement was found to be between methods in the detection of weak cyclones and in the capturing of the earliest and the latest stages of the cyclone life cycle. Another source of uncertainty relates to the use of reanalyses that differ in model formulation, resolution, and data assimilation methods (Hodges et al. 2003, 2011; Bromwich et al. 2007; Raible et al. 2008; Allen et al. 2010), as these may influence the ability of the whole reanalysis system to accurately simulate the atmospheric circulation conditions associated with extratropical cyclones. Explorations into the impact of resolution have been conducted by Blender and Schubert (2000) and Jung et al. (2006); however, with reanalysis data now available on spatial scales of less than 75 km (T255), further examination into this problem is necessary. Care should nevertheless be taken to consider multiple reanalyses when analyzing climatological variability to ensure that problems within the respective datasets do not negatively influence the results.

Early intercomparisons of cyclone activity in reanalyses were mostly focused on the National Centers for Environmental Prediction (NCEP)–National Center for Atmospheric Research (NCAR) reanalysis (herein NCEP–NCAR) and 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) (e.g., Akperov and Mokhov 2010; Hodges et al. 2003; Trigo 2006; Wang et al. 2006; Bromwich et al. 2007; Löptien et al. 2008; Raible et al. 2008). These comparisons demonstrated generally higher numbers of cyclones in ERA-40 relative to NCEP–NCAR, primarily in the summer season, as well as a stronger deepening of cyclones in ERA-40. Interannual trends (1958–2002) were found to be consistent in the Atlantic–European sector in NCEP–NCAR and in ERA-40 (e.g., Trigo 2006). Löptien et al. (2008) reported significant differences across different reanalyses in cyclone life cycle characteristics, such as cyclone intensity, deepening rates, and propagation speed. Recently, Hodges et al. (2011) intercompared four modern-era reanalyses—the ECMWF Interim Re-Analysis (ERA-Interim), the Japanese 25-year Reanalysis Project (JRA-25), the NCEP Climate Forecast System Reanalysis (NCEP-CFSR), and the National Aeronautics and Space Administration's Modern-Era Retrospective Analysis for Research and Applications (NASA-MERRA)—and concluded that spatial climatologies of the cyclone counts were relatively consistent in all reanalyses. The smallest number of cyclones was found in JRA-25, and the highest counts were identified in ERA-Interim and NASA-MERRA. Furthermore, in contrast to the other reanalyses, NASA-MERRA demonstrated a remarkably higher occurrence of extremely deep cyclones of less than 920-hPa central pressure. However, these cyclones should be considered with caveats regarding the nonspectral resolution of NASA-MERRA and the potential systematic biases in NASA-MERRA surface pressure. Allen et al. (2010) analyzed the rapidly intensifying cyclones (RICs) in the four reanalyses [the NCEP–U.S. Department of Energy (DOE) reanalysis (herein NCEP–DOE), ERA-40, ERA-Interim, and JRA-25] and found that spatial pattern and interannual variations of RICs are highly consistent in these products in the Northern Hemisphere (NH), with resolution playing an important role.

Despite the large and varied scope of explorations to date, some critical questions about the robustness of the characteristics of cyclone activity in different reanalyses have yet to be addressed. For example, differences in the regional characteristics of the cyclone life cycle, including cyclone depth, lifetime, deepening rate, and propagation speed, are not well known. Comparing characteristics of interannual variability in cyclone activity may help to quantify the extent to which these characteristics are representative in different reanalyses, and thus identify the most and least robust parameters of the cyclone life cycle. A comparative assessment of cyclone activity in different reanalyses with an alternative method [with respect to, e.g., Hodges et al. (2011)] is justified by the limitations of different individual tracking methods. For instance, Hodges et al. (2011) reported some noticeably different conclusions about the reanalysis-to-reanalysis differences in cyclone activity drawn from the tracking based on sea level pressure (SLP) and relative vorticity.

In this research, we focus on the intercomparison of cyclone activity in five different reanalyses (NCEP–DOE, JRA-25, ERA-Interim, NCEP-CFSR, and NASA-MERRA) and apply a single numerical tracking algorithm extensively tested in different applications (Zolina and Gulev 2003; Jung et al. 2006; Löptien et al. 2008; Rudeva and Gulev 2007, 2011). All of these reanalyses cover the same period from 1979 to 2010 but differ in spatial and spectral resolution (Table 1) as well as in model configuration and assimilation methods. The paper begins with a description of the datasets, the tracking algorithm, and our strategy of intercomparison (section 2). Section 3 illustrates the comparisons of cyclone climatologies and characteristics of cyclone life cycles in different reanalyses. Comparisons of the pattern of decadal-scale and interannual variability are presented in section 4. In section 5, we discuss the comparability of cyclone characteristics in different reanalyses and suggest further directions for this research approach.

Table 1.

Basic characteristics of the reanalysis datasets used in this study (available from 1979 to present and at a time resolution of 6 h).

Table 1.

2. Data and methods

a. Datasets

Given that earlier work has previously compared the NCEP–NCAR and ERA-40 (Kalnay et al. 1996; Uppala et al. 2005), we analyze five reanalyses: NCEP–DOE (Kanamitsu et al. 2002), JRA-25 (Onogi et al. 2007), ERA-Interim (Simmons et al. 2007; Dee et al. 2011), NCEP-CFSR (Saha et al. 2010), and NASA-MERRA (Rienecker et al. 2011). In addition to the increasing spectral resolution from T62 in NCEP–DOE to T382 in NCEP-CFSR, these products differ in terms of model configuration, data assimilation system, and data assimilation input. Most reanalyses are based on three-dimensional variational data assimilation (3DVAR), and NCEP-CFSR also includes first-order time interpolation to the observation (FOTO) (Rančić et al. 2008). Four-dimensional variational data assimilation (4DVAR) is implemented in ERA-Interim. The primary novelty of the most recent NCEP-CFSR (Saha et al. 2010) is a coupling to the ocean during the generation of the 6-h guess field. Detailed descriptions of individual reanalysis systems are given in the background references in the beginning of this section and in other comparative assessments (e.g., Raible et al. 2008; Hodges et al. 2011; Trenberth et al. 2011). Table 1 also shows the basic characteristics of different reanalyses and the spatial and spectral/vertical resolutions of the output. General climatological characteristics of all five reanalyses were compared for the 32-yr period from 1979 to 2010 covered by all products. Given that some key characteristics of cyclone activity (e.g., the total number of cyclones and cyclone center pressure) in ERA-Interim, JRA-25, NCEP-CFSR, and NASA-MERRA were extensively covered for the period of 1989–2009 by Hodges et al. (2011), in this comparison we concentrate on the parameters of cyclone life cycle, regional differences, and interannual variability.

b. Cyclone identification and tracking

Cyclone tracking in all reanalyses was performed using the numerical algorithm of Zolina and Gulev (2002, 2003) and Rudeva and Gulev (2007), which has been effective in representing both climatological features and variability patterns of cyclone activity (e.g., Neu et al. 2013). The tracking is performed on a polar orthographic projection of 181 × 181 points (Zolina and Gulev 2002), allowing for effective cyclone identification north of 25°N. Interpolation of the original reanalysis grids onto the polar orthographic grid is carried out using the modified method of local procedures (Akima 1970), which yields better accuracy and does not result in unrealistic local extrema compared with the alternative procedures (e.g., spline functions). Interpolated 181 × 181 fields used for the tracking provide an actual resolution virtually equivalent to 0.5°–1° in the polar regions and to less than 2° in midlatitudes. Given the spatial resolution of the output from 0.5° in NCEP-CFSR to 2.5° in a number of other reanalyses, the fields used did not exhibit any significant spatial smoothing of the original resolutions but also did not generate any artificial refinement as a result of the abovementioned skills of the Akima (1970) method.

The tracking begins with the dynamical interpolation of 6-h SLP fields onto 1-h time steps discriminating between the step-by-step cyclone migrations and the distances between the neighboring cyclone centers (Jung et al. 2006; Rudeva and Gulev 2007). In the next step, the local SLP minima (<1015 hPa) are analyzed to determine the cyclone centers. This step involves several iterations of the analysis of SLP derivatives computed from the 17 neighboring points, the determination of the impact area, and the analysis of SLP characteristics within this impact area. To build the trajectories from a first guess of the cyclone migration we used the so-called method of the nearest neighbors, as in Murray and Simmonds (1991a,b), König et al. (1993), Blender and Schubert (2000), and Hoskins and Hodges (2002). Further identification of the tracks involves the three-pass analysis of cyclone propagation velocities, sorting of the crossing trajectories, and the separate analysis of the stationary cyclones.

The output of tracking (coordinates, central pressure, and time) was subjected to preprocessing, which included the truncation of transients with very small lifetimes or short migration distances. In this study, unless stated otherwise, we applied truncation thresholds of 2 days for the lifetime and 1000 km for the migration distances to all tracks. These values were used to make the results easily comparable to Hodges et al. (2011). To avoid the impact of potential errors in the pressure adjustment to mean sea level (which are different in different reanalyses), in areas with orography higher than 1500 m we filtered out all cyclones reaching a minimum SLP over those areas. This approach is different to that used by Neu et al. (2013), who only removed parts of the tracks over elevated orography. This led to a reduction of the cyclone life time in the vicinity of mountains, but had very little effect on other cyclone characteristics and on the overall number of cyclones. Our approach does not cause a shortening of cyclone tracks but may lead to a reduction of the total number of cyclones. For the time being, it is difficult to prove which approach is superior. When mapping cyclone characteristics, we used a circular grid with circular cells of a 2° latitude radius (equivalent to approximately 155 000 km2).

This grid has a higher resolution than that used in Hodges et al. (2011), allowing for a more accurate quantification of the regional differences. Considering here hemispheric climatologies and trends, we primarily focused on the analysis of cyclone numbers (i.e., the number of tracks within each cell with multiple entries to be eliminated), although cyclone frequencies (estimated as the number of 6-hourly occurrences of a cyclone center in the grid cell) were also computed. Cyclone numbers and cyclone frequencies are analogous to track and cyclone density (Hodges et al. 2003). To effectively map cyclone numbers and frequencies, 6-hourly trajectories were interpolated linearly onto 10-min time steps. This process eliminates underestimation of the number of cyclones and random errors in cyclone frequencies that can occur when this procedure is not applied (Zolina and Gulev 2002).

Other parameters of the cyclone life cycle included minimum SLP during the cyclone lifetime, mean and maximum deepening/filling rates (δP and δPmax, respectively), cyclone propagation speed, and cyclone lifetime (Gulev et al. 2001; Rudeva and Gulev 2007, 2011). All results of the storm tracking were derived for the winter [December–February (DJF)], spring [March–May (MAM)], summer [June–August (JJA)], and autumn [September–November (SON)] seasons.

3. Intercomparison of the climatological characteristics of cyclone life cycle

An important aspect of the analysis of cyclone activity is the mean climatology (Figs. 1a,b) identifying the major storm tracks. We use the term “storm track” here for identification of the areas of the highest storm track density, while this term is also frequently attributed to the areas of maximum standard deviations of the band-passed height of 500-hPa surface in the synoptic range (e.g., 2.5–6 days). Filtering of cyclones reaching minimum central pressure over the regions with elevated orography decreases cyclone numbers (e.g., over the Rocky Mountains, Greenland, and the Tibetan plateau by 70%–80%). In the area of the Mediterranean storm track, however, the effect is not as strong (approximately 10%). The annual climatology shown for NASA-MERRA in Figs. 1a and 1b identifies the major North Atlantic and North Pacific storm tracks, whose positioning is very consistent in all five reanalyses. Over both the North Atlantic and North Pacific, NASA-MERRA systematically shows a higher number of cyclones than in NCEP–DOE and JRA-25 by 15%–20% and 10%–15%, respectively, with the larger differences (up to 30%) over the North Pacific (Figs. 1c,d). Over the continental storm tracks, the differences amount to 50%. Compared with the higher-resolution NCEP-CFSR and ERA-Interim (Figs. 1e,f), NASA-MERRA shows very similar cyclone counts over the oceanic storm tracks, with somewhat smaller values over the Gulf Stream and higher counts over the Kuroshio. Over the continents, the differences in cyclone counts between NASA-MERRA on the one hand and the other four reanalyses on the other hand are comparable, with the NASA-MERRA–NCEP-CFSR differences being the largest. Furthermore, NASA-MERRA systematically shows fewer cyclones over the Labrador Sea, the Irminger Sea, and mid-Asia. We can hypothesize that this can be attributed to the differences in the representation of orography (particularly for Greenland) and surface pressure adjustment to sea level in the nonspectral NASA-MERRA model and in spectral models used in the other reanalyses. The results of Figs. 1c–f are based on the number of cyclone tracks. This metric has an uncertainty associated with the artificial splitting of some tracks in the tracking algorithms. We performed similar comparisons for cyclone frequencies, quantifying only the cyclone centers with no attribution to the trajectories. This comparison (not shown) demonstrated high consistency with the results shown in Fig. 1.

Fig. 1.
Fig. 1.

Annual climatologies (1979–2010) of cyclone numbers in NASA-MERRA (a) before and (b) after filtering of cyclones over elevated terrain, and differences in the annual cyclone numbers between NASA-MERRA and (c) NCEP–DOE, (d) JRA-25, (e) ERA-Interim, and (f) NCEP-CFSR. Units are tracks per year per circle with a radius of 2° latitude (equivalent to approximately 155 000 km2).

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00777.1

Seasonal changes in total cyclone counts are qualitatively consistent in all reanalyses and are associated with the enhancement of the oceanic storm tracks in the winter and an increasing number of cyclones over the continental storm tracks in the summer. These weak summer continental cyclones are not, however, an artifact of the tracking method, as proven by the analysis of the probability density functions (PDFs) of the cyclone core pressure and lifetime (not shown).

The spread in seasonal cyclone numbers was quantified by absolute and relative variances (Fig. 2). This approach allows for the estimation of dataset-to-dataset variations when the truth (i.e., reference) is unknown. We denote the absolute variation between the number of cyclones in a grid cell in datasets “1” and “2” as D1,2 = |x1 − x2|, and δ1,2 = D1,2/|x2| gives the relative variation of results of dataset 1 with respect to dataset 2. For n datasets,
e1
where n = 5 denotes the number of reanalyses considered and the binomial coefficient in the denominator equals 10 (number of pairs in a set of five), with self-comparisons (i = j) being eliminated. Over oceans, northern Europe, and eastern North America in both the winter and summer, the absolute variance is considerably less than 1 cyclone season−1 (Figs. 2a,b), with a relative variance of 2%–7%, implying that the reanalyses in these regions are very consistent. Locally, high absolute variance (equivalent to 10% and 30% of relative variance in the winter and summer, respectively) is observed over the Mediterranean storm track, implying that different reanalyses are less robust in replicating Mediterranean cyclone statistics, likely as a result of resolution, as noted by Jung et al. (2006). One can argue that the large spread of cyclone counts over the Mediterranean can primarily be imposed by the 2000-km threshold on the cyclone traveling distance, potentially truncating slow-moving and stationary cyclones, such as medicanes (Emanuel 2005), whose role may increase in the future climate (Lionello and Giorgi 2007; Raible et al. 2010). Our tests (Table 2) show that counting all cyclones, including those traveling less than 2000 km, changes the absolute variance in accordance with the increase in the total cyclone count, but does not considerably change the relative variance in this region.
Fig. 2.
Fig. 2.

Absolute variance in the number of cyclones computed using Eq. (1) for the total number of cyclones in the (a) DJF and (b) JJA during 1979–2010. Units are as in Fig. 1.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00777.1

Table 2.

Cyclone counts as well as absolute and relative variance in DJF and JJA (season−1) among the five reanalyses 1979–2010 over the Mediterranean for two thresholds (“>2000 km” and “All cyclones”).

Table 2.

Hodges et al. (2011) demonstrated the consistency of the seasonal cyclone counts over the NH, with NASA-MERRA, NCEP-CFSR, and ERA-Interim typically showing a slightly higher number of cyclones than JRA-25 for all seasons. Our results (Table 3) are in a qualitative agreement with Hodges et al. (2011) and confirm the highest annual and seasonal cyclone counts in NASA-MERRA. Relative to NASA-MERRA, the smallest number of cyclones is found in NCEP–DOE (23% less on average), and similar counts are identified in NCEP-CFSR and ERA-Interim (5% less). These differences are largest in the summer and smallest in the winter (27% and 19%, respectively, for NCEP–DOE).

Table 3.

Mean seasonal (DJF, MAM, JJA, and SON; season−1) and annual (yr−1) numbers of cyclones identified in the five reanalyses during the period 1979–2010 over the NH.

Table 3.

Model resolution is likely the major factor affecting cyclone counts in different reanalyses. Increasing cyclone numbers with resolution are evident for both the winter and summer seasons and for all cyclone intensities over oceans and continents (Fig. 3). Cyclones were attributed to the continental or oceanic domain by the location of the minimum central pressure during the cyclone life cycle. The largest increase in the number of cyclones in relation to resolution (approximately 25% over oceans and up to 40% over continents) is observed in the range T62–T255 for deep and moderate cyclones, while with higher spectral resolutions the dependence is not that clear, which is consistent with the results of Jung et al. (2006). Thus, the differences between ERA-Interim and NCEP-CFSR are actually minor. NASA-MERRA may fall out of this general dependence, particularly over continents, which may be partly explained by the nonspectral model used in NASA-MERRA. Hodges et al. (2011) found a small fraction of very deep cyclones (below 930–920 hPa) in NASA-MERRA, contrary to the other reanalyses. Our results obtained with a different method confirm this finding, showing 1–3 cyclones yr−1 that are very deep in NASA-MERRA, primarily over the North Pacific storm track.

Fig. 3.
Fig. 3.

Total annual numbers of cyclones (season−1) of different intensity identified in five reanalyses over the (a),(b) continents and (c),(d) oceans for (left) DJF and (right) JJA during 1979–2010, plotted as a function of spectral resolution of reanalysis. Colors define minimum central pressure of cyclone. Spectral resolution is shown at the x axis for four spectral reanalyses; the location of NASA-MERRA (nonspectral model) is given tentatively. Error bars correspond to interannual std dev of the number of cyclones of different types in different reanalyses.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00777.1

Estimation of the statistical significance of differences in the cyclone counts according to the Student's t test (Fig. 4) implies significant differences primarily among NASA-MERRA, NCEP-CFSR, and ERA-Interim on the one hand and JRA-25 and NCEP–DOE on the other. Over the oceans, this holds for all cyclone intensities during the summer and also for moderately deep cyclones in the winter. Over the continents, differences in cyclone numbers are statistically significant only for the moderately deep cyclones in both seasons and shallow cyclones in the winter. Importantly, differences in continental cyclone counts in the winter are also statistically significant for the pairs of modern reanalyses (e.g., ERA-Interim–NASA-MERRA and ERA-Interim–NCEP-CFSR).

Fig. 4.
Fig. 4.

Statistical significance of differences in number of cyclones between different products estimated according to a Student's t test during 1979–2010. Estimates are performed for (top) ocean areas and (bottom) continents for: (a),(d) deep (<980 hPa); (b),(e) moderate (980–1000 hPa); and (c),(f) shallow (>1000 hPa) cyclones. Estimates below and above the main diagonals in each of the six panels correspond to DJF and JJA, respectively. Differences significant at 90%, 95%, and 99% levels are marked by different tones of blue, with statistically insignificant differences remaining uncolored.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00777.1

Major disagreements among the reanalyses with respect to the moderately deep and shallow cyclones suggest potentially different cyclone life cycle parameters in different datasets. Because weak cyclones have generally a shorter lifetime than deep ones (e.g., Gulev et al. 2001; Rudeva and Gulev 2007), we can anticipate differences in the distributions of cyclone lifetime. Although winter histograms of lifetime (Fig. 5a) show slightly higher occurrences of short-lived cyclones in NASA-MERRA relative to the other reanalyses, the summer probability distributions of the cyclone lifetime (Fig. 5b) confirm a higher occurrence of short-lived cyclones in NASA-MERRA. Applying chi-squared (χ2) and Kolmogorov–Smirnov (K–S) tests reveals a statistical significance at the 95% level. Probability distributions of the cyclone propagation speed (Figs. 5c,d) are highly consistent in the winter, while in the summer ERA-Interim and NCEP-CFSR show a higher number of slow-moving cyclones than NCEP–DOE, JRA-25, and NASA-MERRA confirmed by χ2 and K–S tests at the 95% significance level. Our analysis also identified somewhat stronger cyclone deepening rates in NASA-MERRA, particularly over oceanic storm tracks (not shown).

Fig. 5.
Fig. 5.

(left) DJF and (right) JJA climatological (1979–2010) occurrence histograms for the cyclone (a),(b) lifetime and (c),(d) propagation speed in five reanalyses.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00777.1

The potential of reanalyses to accurately replicate the occurrence of RICs is of special interest because these systems are associated with extreme weather events. First described by Sanders and Gyakum (1980) and Rogers and Bosart (1986), RICs represent a class of oceanic cyclones developing faster than 1 Bergeron [24 hPa (24 h)−1]. Although the definition of rapid deepening based on 24-h pressure declines is widely accepted (e.g., Stewart and Donaldson 1989; Yau and Jean 1989), in some studies rapid deepening was also quantified by 6-h pressure declines (e.g., Gulev et al. 2001). Some studies have used the extended definition of RICs based on the consideration of tendencies in relative central pressure after removing the background field (Lim and Simmonds 2002; Allen et al. 2010). We quantified RICs according to the criterion
e2
where δP is the central pressure fall (within 6 or 24 h), τ is the time step (6 or 24 h), and ϕ is the latitude (Sanders and Gyakum 1980; Allen et al. 2010).

The total annual number of RICs (Table 4) in our study is 20%–30% higher than that of Allen et al. (2010), who used the University of Melbourne (Murray and Simmonds 1991a,b) method, which reported generally smaller total cyclone counts relative to our method (Neu et al. 2013). In all reanalyses, approximately 50% of the RICs are observed in the winter, with the other half nearly equally split between the spring and autumn seasons. According to both estimates (based on 24- or 6-h pressure falls), the highest winter number of RICs is found in NASA-MERRA, showing a 5%–12% higher rate than NCEP-CFSR and ERA-Interim and an approximately 20%–25% higher rate than NCEP–DOE and JRA-25 over both the North Atlantic and North Pacific. However, the relative fraction of RICs in the total number of transients over the oceans (Table 4) in NASA-MERRA (16.5% and 6.5% for 6- and for 24-h pressure falls, respectively, for annual numbers) is similar to reanalyses. This conclusion holds for all seasons, implying that the fraction of RICs in the total count is quite stable in different products.

Table 4.

Mean seasonal [DJF, MAM, and SON; 24 hPa (24 h)−1] and annual [6 hPa (6 h)−1 and 24 hPa (24 h)−1] numbers of RICs using different definitions for the five reanalyses during the period 1979–2010 over the NH. The RICs in the total number of cyclones over the NH (%) are given in parentheses. The number of RICs during JJA is not shown, but typically amounts to <4 season−1.

Table 4.

4. Interannual- to decadal-scale variability in cyclone characteristics

Consistency in mean climatological cyclone characteristics may not necessarily imply a similar pattern of interannual- to decadal-scale changes and vice versa. Trends in reanalysis-derived cyclone characteristics, and similarly in the other variables, can be influenced by technical flaws, such as the lack of homogeneity in data assimilation input. This feature is particularly apparent in the case of the sharp increase in assimilated information because of increasingly available satellite data, resulting in strong temporal inhomogeneities, particularly in the Southern Hemisphere (White 2000; Sterl 2004). Bengtsson et al. (2004) found that changes in the data assimilation input can also affect trends in global quantities, such as global mean temperature, integrated water vapor, and kinetic energy. Neu et al. (2013) demonstrated that trend patterns in cyclone activity over the NH in ERA-Interim are highly consistent when derived with different tracking methods. Thus, a comparison of trends in different reanalyses may shed more light on the reliability of trend signals in cyclone activity during the past several decades. Statistically significant (Student's t test) positive trends in the total number of cyclones over the NH during the period of 1979–2010 (Fig. 6) were found only in NCEP–DOE (29.7 decade−1) and ERA-Interim (18.8 decade−1). These trends are largely influenced by the period of 1989–2010, during which NCEP–DOE and ERA-Interim report an increase of 160 (12%) and 50 (3%) cyclones, respectively. The other three reanalyses, NASA-MERRA, NCEP-CFSR, and JRA-25, do not indicate significant linear trends during this period. The trends in total cyclone count are largely explained by the change in the number of shallow and moderate cyclones (>990 hPa), demonstrating an increase of 6%–11% decade−1 in both the winter and summer seasons in all products except NCEP-CFSR (Fig. 7). In contrast to the increasing number of shallow cyclones, the number of deep cyclones consistently decreases during 1989–2010 for both seasons (Figs. 7a,b). The number of intense events (<960 hPa, 2%–3% of the total number of cyclones) in the North Atlantic consistently increased in all reanalyses during the 1980s and decreased afterward (Fig. 6b). In the North Pacific, the number of intense events alternatively declined during the 1980s and increased in the 1990s with a subsequent decline in the recent period, with a robust signal in all products (Fig. 6c).

Fig. 6.
Fig. 6.

Time series of (a) the total annual number of cyclones over the NH and (b),(c) the number of very deep (<960 hPa) cyclones over the North Atlantic and North Pacific, respectively. Thin lines correspond to interannual values and thick lines show 5-yr running means.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00777.1

Fig. 7.
Fig. 7.

Changes in the number of cyclones of different intensities during the 32-yr period (1979–2010) for (a) DJF and (b) JJA in different reanalyses.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00777.1

The winter spatial trend pattern (shown for NASA-MERRA in Fig. 8a) demonstrates positive trends of more than 2 cyclones decade−1 over the major North American–North Atlantic storm track and negative trends of a similar magnitude in the eastern midlatitudinal Atlantic and over the Mediterranean storm track. Over the North Pacific in the winter, no statistically significant trends are observed in the Kuroshio formation region. Significantly positive trends are observed over the subpolar storm track, with a decrease in the number of cyclones in the subtropics. The summer trend pattern (Fig. 8b) is characterized by the weakening of cyclone activity in the western North Atlantic and increasing cyclone counts for the central North Atlantic, the Atlantic U.S. coast south of 45°N, the eastern Mediterranean, and the eastern Arctic. The latter pattern has been reported by Serreze and Barrett (2008) and is associated with the declining Arctic Sea ice cover.

Fig. 8.
Fig. 8.

(a),(b) Estimates of the linear trends of the number of cyclones per 155 000 km2 (decade−1) during 1979–2010 in NASA-MERRA for DJF and JJA, respectively. Black dots denote trends that are statistically significant at the 90% level according to a Student's t test. (c),(d) The number of reanalyses simultaneously indicating statistically significant (85% level) linear trends of the same sign during 1979–2010 for DJF and JJA, respectively. In (c) and (d), the hatched areas outside of the major storm tracks indicate where the climatological number of cyclones is small (<2 season−1 over 155 000 km2 per season).

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00777.1

To quantify the extent to which spatial distributions of the linear trends are consistent with each other, we estimated the number of reanalyses reporting significant (at the 85% level) trends of the same sign for each grid cell (Fig. 8). For this level of individual significance (85%), a binominal distribution implies a 95% group significance for four and a 90% group significance for three consistent estimates (e.g., Livezey and Chen 1983). In the winter, all five reanalyses confirm the enhancement of the North American–North Atlantic storm track and the decrease in the number of cyclones over the eastern Atlantic and Mediterranean. Similarly, in the North Pacific, most reanalyses indicate strengthening of the subpolar storm track with simultaneous weakening of the subtropical storm track. In the summer (Fig. 8d), all reanalyses consistently show a decreasing number of cyclones over North America, along with a positive tendency in the cyclone count over northern Siberia. Summer enhancement of the eastern Mediterranean storm track is confirmed by at least four of the five reanalyses. Over the entire NH, there is no indication of the existence of the collocated positive and negative trends, reported by at least two reanalyses for each sign (not shown).

The pattern of significantly positive trends over midlatitudinal and subpolar North Atlantic regions is paired with negative trends in the number of cyclones in the eastern midlatitudinal Atlantic and Mediterranean (Figs. 8a,c). This finding hints at the tendency of the poleward deflection of the storm tracks in the Atlantic–European region (Orlanski 1998) that was evident in the model simulation of the warming climate (Ulbrich and Christoph 1999; Pinto et al. 2007; Löptien et al. 2008; Woollings and Blackburn 2012). In the eastern Pacific, all reanalyses also demonstrate a consistently growing number of cyclones over the subpolar storm track, with the strongest tendency in ERA-Interim of 2 decade−1, and declining cyclone counts in the subtropics, indicating a poleward shift of the cyclone trajectories in the North Pacific (Löptien et al. 2008). We also analyzed trend pattern in cyclone frequencies (not shown) that were consistent with those for cyclone numbers in individual reanalyses. They also demonstrated characteristics of the consistency across different reanalyses which were close to those in Figs. 8c and 8d.

The temporal evolution of the number of cyclones (Fig. 6) is largely influenced by decadal-scale variability, which is most pronounced in NASA-MERRA. To quantify the consistency of decadal changes in the number of cyclones of different intensities, we analyzed the temporal evolution of the occurrence anomalies of the cyclone central pressure in different regions (Fig. 9). Occurrence anomalies were computed from the seasonal histograms of cyclone intensity as follows:
e3
where x is the cyclone central pressure class, P(x) is the probability of occurrence of cyclones in this class for an individual year, σ[P(x)] is the standard deviation of the probability distribution for this class, and the overbar is the time-averaging operator.
Fig. 9.
Fig. 9.

(top) Annual normalized occurrence anomalies of cyclone central pressure in NASA-MERRA computed using Eq. (2) and (bottom) the number of reanalyses simultaneously indicating anomalies of the same sign for a given year and central pressure for the same areas for: (a),(e) the Atlantic; (b),(f) the Arctic; (c),(g) the Pacific; and (d),(h) Europe.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00777.1

Occurrence anomalies of cyclones of different intensities in NASA-MERRA in the North Atlantic (Figs. 9a,e) show a general decrease in the occurrence of very deep cyclones in the 2000s (consistent with Fig. 6b), superimposed with a positive anomaly of the number of moderately deep cyclones. The negative anomaly of the number of deep cyclones in the 1980s occurs together with the positive anomaly of the number of moderate cyclones. This pattern is reversed in the late 1980s and 1990s, which were characterized by the higher occurrence of deep cyclones that is consistent with the findings of Gulev et al. (2001). In the North Pacific (Figs. 9c,g), the tripole of the negative anomaly for very deep and shallow cyclones and the positive anomaly for moderately deep cyclones in the 1980s and early 1990s reverses in the mid-1990s before returning again to its initial state in the late 2000s. The Arctic (Figs. 9b,f) pattern is characterized by a dipole between deep and shallow cyclones, with negative and positive anomalies of deep cyclones in the 1980s and 1990s, respectively. Over Europe (Figs. 9d,h), there is a clear pattern of decreasing numbers of deep cyclones starting in 1995, along with an increasing frequency in shallow cyclones.

These patterns are generally well replicated by most products, as demonstrated in Fig. 9, which shows the number of reanalyses indicating either positive or negative occurrence anomalies in the same manner as in Figs. 8c and 8d. Most regional patterns of decadal-scale variability are confirmed by at least three of the five reanalyses considered; the patterns are most evident in NASA-MERRA, ERA-Interim, and NCEP-CFSR and, to a lesser extent, in NCEP–DOE and JRA-25, particularly in the Pacific.

The consistency of interannual variability in cyclone activity was analyzed by estimating correlations between cyclone counts after detrending the time series. Hemispheric total counts over both oceans and continents are not as highly correlated as one would expect, with the highest correlation being slightly higher than 0.60 between NASA-MERRA and JRA-25 in the winter. An analysis of pair correlations for the cyclones of different intensities (Fig. 10) shows that in both the winter and summer seasons, the highest correlation is observed between the numbers of deep cyclones over the oceans, with the correlation for most pairs ranging from 0.7 to 0.9. For moderate and shallow oceanic cyclones, pair correlations decrease relative to the deep cyclones and may drop below the significance level, particularly for shallow cyclones. Over the continents (Fig. 10b), the largest spread of the correlations is observed in the summer, when the largest correlation (up to 0.8) is identified for the pair ERA-Interim and NCEP-CFSR for all three classes of cyclone intensity. The lowest correlation may drop to close to zero values for the pairs NCEP–DOE and JRA-25 or NCEP–DOE and NASA-MERRA for shallow cyclones. In the winter, it is difficult to identify individual reanalyses that systematically show the highest or lowest correlation with each other. For the summer, the lowest correlations are typically identified for the pairs that include NCEP–DOE.

Fig. 10.
Fig. 10.

Estimates of pair correlations between the number of cyclones of different intensities for all possible pairs of datasets for the winter (blue circles) and summer (red circles) for NH (a) oceans and (b) continents. Gray dotted lines denote the confidence limit for correlation coefficients at the 99% level. Text captions denote the pairs showing the highest and lowest correlations.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00777.1

For further insight into the relatively low level of correlations among the aggregated cyclone counts (Fig. 10), we present the maps of correlations between cyclone numbers in NASA-MERRA and in the four other reanalyses in Fig. 11. In the winter, the highest correlations of more than 0.8 are observed over northern Europe, European Russia, and eastern Siberia, with the strongest consistency (where the correlation approaches 0.9) is among NASA-MERRA, ERA-Interim, and NCEP-CFSR. Over the eastern North Pacific and North Atlantic storm tracks, correlations among different reanalyses range from 0.7 to 0.8, again showing a somewhat higher consistency among NASA-MERRA, ERA-Interim, and NCEP-CFSR. Winter cyclone numbers in the Mediterranean storm track are highly correlated in different products, with the closest covariability between NASA-MERRA and NCEP-CFSR (r > 0.8). The lowest winter correlation is observed over North America and eastern Eurasia, where the interannual variability of the cyclone counts in different products is largely uncorrelated, with correlation coefficients being lower than 0.3 or even negative. The summer correlation among cyclone counts in different reanalyses drops by 15%–20% relative to the winter season. The highest correlation of more than 0.8 is found over the eastern parts of the oceanic storm tracks and over northern Europe, with the strongest similarity between NASA-MERRA and ERA-Interim. Summer cyclone counts over the continents and Mediterranean storm track are largely uncorrelated in different products.

Fig. 11.
Fig. 11.

Spatial distribution of the point correlations between the total number of cyclones per year identified for: (a),(e) NASA-MERRA and NCEP–DOE; (b),(f) NASA-MERRA and JRA-25; (c),(g) NASA-MERRA and ERA-Interim; and (d),(h) NASA-MERRA and NCEP-CFSR, for (top) DJF and (bottom) JJA during 1979–2010. Only correlations higher than 0.5 are shown. The 99% significance level of the correlation is 0.49.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00777.1

Our analysis of interannual variability in the number of RICs generally confirms the results of Allen et al. (2010), who did not find significant trend in the number of RICs over the NH in NCEP–DOE, JRA-25, or ERA-Interim. Time series of the winter anomalies in the number of RICs (identified using 24-h pressure declines; Table 4) for the North Atlantic and North Pacific regions (Fig. 12) show a pronounced variability from 2 to 4 yr, with the strongest magnitude for ERA-Interim in the North Atlantic and for NCEP–DOE in the North Pacific. Statistically significant (90% level) linear trends in the number of RICs are only observed for NASA-MERRA in the North Atlantic (1.7 decade−1) and for NCEP–DOE in the North Pacific (2.8 decade−1), implying an increase of 8%–10% in the number of RICs during the 32-yr period. Correlation between the interannual variability of the number of RICs in different reanalyses in the North Atlantic varies from 0.6 to more than 0.7 for different pairs, with the largest correlation of 0.73 between ERA-Interim and NCEP-CFSR. In the North Pacific, where all reanalyses effectively capture the minimum in 1990, correlation coefficients range from 0.53 (NCEP–DOE and NASA-MERRA) to 0.84 (ERA-Interim and NASA-MERRA). As is the case for cyclones of different intensity (Fig. 10), it is difficult to identify particular pairs of systematically high or low correlations for RICs. A similar analysis performed for RICs quantified with 6-h pressure falls does not qualitatively change the results, confirming the small magnitude of long-term trends and showing a slightly higher correlation among different reanalyses at interannual time scales.

Fig. 12.
Fig. 12.

Winter time series of the anomalies of the total number of rapidly intensifying cyclones [core pressure drops of more than 24 hPa (24 h)−1] in the (a) North Atlantic and (b) North Pacific in different reanalyses.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00777.1

5. Summary and discussion

We compared characteristics of cyclone activity over the NH for five reanalyses using a single cyclone identification and tracking algorithm. The total number of cyclones increases with increasing resolution of the reanalysis, with less than 1400 yr−1 in NCEP–DOE, approximately 1500 yr−1 in JRA-25, and more than 1700 yr−1 in ERA-Interim, NCEP-CFSR, and NASA-MERRA. This is consistent with earlier analyses by Blender and Schubert (2000) and Jung et al. (2006). Among all of the reanalyses investigated, NASA-MERRA shows the highest total number of 1800 yr−1. This increase in numbers with resolution holds for both oceanic and continental cyclones during the summer and winter. The differences among the reanalyses are almost entirely provided by moderate and shallow cyclones, which contribute more than 95% to the total count. The smallest set-to-set absolute and relative variances of cyclone numbers are observed over the oceanic storm tracks, and the largest spread in cyclone counts in different datasets is identified over the continents, particularly over the Mediterranean storm track in the summer. Characteristics of the cyclone life cycle in different reanalyses are quite consistent with each other, while NASA-MERRA shows a significantly higher number of very deep cyclones, as shown by Hodges et al. (2011). In the summer, we found a higher occurrence of short-lived cyclones in NASA-MERRA and a higher occurrence of long-lived systems in ERA-Interim and NCEP-CFSR. These two reanalyses also show a significantly higher occurrence for slow-moving cyclones in the summer. Although the largest absolute number of RICs over both the North Atlantic and North Pacific was identified in NASA-MERRA, the relative fraction of RICs in all reanalyses is remarkably stable.

Significant positive linear trends in the total number of cyclones ranging from 1% to 2% decade−1 could only be identified in NCEP–DOE and ERA-Interim. These trends are most apparent in the period 1989–2010 and are primarily associated with moderate and shallow cyclones. The number of intense cyclones (<960 hPa) in the North Atlantic consistently increases in most reanalyses until 1990 and subsequently declines until the present. In the North Pacific, the number of intense cyclones approaches its maximum near 2000 and subsequently decreases. The most robust trends both in sign and magnitude in the winter are observed over the North American–North Atlantic storm track and subpolar North Pacific storm track (positive trends), as well as over the eastern North Pacific storm track, eastern Europe, and the Arctic (negative trends). In the summer, positive trends in most reanalyses were identified over the Mediterranean storm track, while the North Atlantic storm track shows a consistent pattern of negative trends. Statistically significant trends in the number of RICs (approximately 8%–10% during a 30-yr period) were found only in NASA-MERRA in the North Atlantic and NCEP–DOE in the North Pacific. The highest correlation between the interannual variability in the total number of cyclones was observed over the oceanic storm tracks and northern Eurasia. This correlation is higher for moderate and deep cyclone counts and during the winter season. The strongest correlation was observed in the NASA-MERRA, ERA-Interim, and NCEP-CFSR.

Notable differences in the mean total cyclone counts between different products mainly result from different resolutions of the reanalysis models. Nevertheless, for the modern-era reanalyses, the model formulation and data assimilation algorithms are also likely responsible for significantly higher numbers of cyclones and their higher intensity (as in NASA-MERRA) relative to the other products. In this respect, our results are consistent with those of Hodges et al. (2011); however, we also found that NASA-MERRA shows stronger deepening rates and shorter lifetimes than the other reanalyses. Given the conceptually different tracking schemes and the different periods used in Hodges et al. (2011) and in our study (+1989 and +1979, respectively), the conclusion regarding the consistency of cyclone climatologies in modern-era reanalyses and the higher counts of extreme cyclones in NASA-MERRA appears to be viable. Neu et al. (2013), in comparing 15 different schemes, applied these to a single reanalysis dataset and did not find that the tracking scheme (vorticity versus central pressure) had an impact on the total number of cyclones. However, Neu et al. (2013) were focused on the comparison of the total number of cyclones, paying less attention to the RICs. In this context, some differences in the number of RICs found in our study and in the work of Allen et al. (2010)—who used a somewhat different definition of the deepening rate and the pressure Laplacian-based (analogous to vorticity) numerical algorithm of Murray and Simmonds (1991a,b)—suggest that the number of RICs may depend on the tracking algorithm. Given that schemes based on vorticity (or Laplacian) tend to identify cyclones at the earlier stage of their life cycle (relative to those based on the central pressure), it is likely that these schemes may capture more RICs, many of which are known to exhibit rapid deepening at the earlier stage of the development. This notion along with some differences in definitions and chosen criteria may explain the differences in the means and variability patterns in the number of RICs between our estimates and those of Allen et al. (2010).

In all reanalyses, we observed an enhancement of the winter North American–North Atlantic midlatitudinal storm track and the simultaneous weakening of the Mediterranean storm track during the period of 1979–2010. A similar pattern that is also notable across different reanalyses was found in the eastern North Pacific. This observation corresponds with the tendency of a poleward deflection of the storm tracks in the NH midlatitudes identified in model simulations of anthropogenically forced climate (Yin 2005; Löptien et al. 2008; Ulbrich et al. 2008; Woollings and Blackburn 2012). From our results, it is difficult to conclude whether this trend has a steady tendency associated with a warming signal or a pattern indicative of shorter-period (decadal to annual) natural variability. In the future, it will be important to evaluate the mechanisms responsible for this phenomenon in different reanalyses, looking particularly at the zonal inhomogeneity of global temperature trends (Chen et al. 2008) and the SST signals reflecting changes in the Atlantic meridional circulation (Woollings 2008). We note that Raible and Blender (2004) found the poleward shift of the storm tracks in the North Pacific in response to the realistic representation of SST in a coupled climate model.

In the following, studies of the differences in cyclone activity in different reanalyses should take into account the uncertainties of numerical algorithms applied for storm identification and tracking. Neu et al. (2013) demonstrated that uncertainties associated with the use of different tracking algorithms are generally higher than the differences between different reanalyses found by Hodges et al. (2011) and in our study. In this respect, an application of different algorithms for comparative assessments may help to discriminate between differences that are most evident and those that are not consistently replicated by different tracking schemes.

Further quantitative estimation and understanding of the differences in cyclone characteristics among different products requires advanced methods of diagnosing cyclone activity. In this respect, an investigation into the role of midlatitudinal cyclones in forming atmospheric heat and moisture transports in different reanalyses will be of special interest. Trenberth et al. (2011) demonstrated that the differences in the total moisture content and ocean-to-continent moisture transports between different reanalyses can be quite large, particularly for stronger midlatitudinal transports in the NH in NASA-MERRA relative to ERA-Interim. Cullather and Bosilovich (2011, 2012) demonstrated significant differences from 40% to 50% in the moisture flux components and energy budgets in the Arctic NASA-MERRA, NCEP-CFSR, and ERA-Interim. Simmons et al. (2010) found significant discrepancies in moisture characteristics in reanalyses when compared with observations. Considerable differences in air–sea turbulent fluxes are shown in different reanalyses, particularly in the North Atlantic and North Pacific midlatitudes, as recently demonstrated by Brunke et al. (2011). Brunke et al. found that NASA-MERRA is the least biased, relative to the other products, with respect to direct observations. It is important to understand how these differences are mirrored in the structure of cyclones, their heat and moisture balances, and their role in the advection of moisture. For this purpose, the composites of cyclone characteristics built in Hodges et al. (2011) should be complemented with a comparative analysis of the regional composites of cyclone moisture and energy budgets (Rudeva and Gulev 2011; Dacre et al. 2012). Another important diagnostic might be assessing the clustering of cyclones responsible for the enhancement of the moisture transports in the North Pacific and North Atlantic (e.g., Mailier et al. 2006). Along with the comparative case studies for selected high-impact cyclones, such as storms Kyrill and Klaus (Fink et al. 2009; Liberato et al. 2011), these analyses will shed more light on the differences in mechanisms driving midlatitudinal cyclone activity in different products. Further efforts in comparing cyclone characteristics in reanalyses may help researchers to better associate cyclone activity with large-scale circulation modes, such as the North Atlantic Oscillation (NAO) (Gulev et al. 2001; Raible 2007) and the Pacific decadal oscillation (PDO), as well as with oceanic signals and sea ice characteristics (Serreze and Barrett 2008; Simmonds and Keay 2009; Screen et al. 2011).

Acknowledgments

This work was supported by the Russian Ministry of Education and Science under the Federal “World Ocean” Programme (Contracts 16.420.12.0001 and 16.420.12.0006), Contract 14.515.11.0008, Agreement 8333, and by a special Grant 11.G.34.31.0007 for establishing excellence at Russian universities. We highly appreciate the feedback from the three anonymous reviewers who critiqued the first and revised versions of the manuscript. We thank Olga Zolina of LGGE (Grenoble), Kevin Hodges of the University of Reading, and Siegfried Schubert of NASA Goddard Space Flight Center, Greenbelt, for useful discussions on different aspects of this work.

REFERENCES

  • Akima, H., 1970: A new method of interpolation and smooth curve fitting based on local procedures. J. Assoc. Comput. Mach., 17, 589600.

    • Search Google Scholar
    • Export Citation
  • Akperov, M. G., , and I. I. Mokhov, 2010: A comparative analysis of the method of extratropical cyclone identification. Atmos. Oceanic Phys., 46, 574590.

    • Search Google Scholar
    • Export Citation
  • Allen, J. T., , A. B. Pezza, , and M. T. Black, 2010: Explosive cyclogenesis: A global climatology comparing multiple reanalyses. J. Climate, 23, 64686484.

    • Search Google Scholar
    • Export Citation
  • Bengtsson, L., , S. Hagemann, , and K. I. Hodges, 2004: Can climate trends be calculated from reanalysis data? J. Geophys. Res., 109, D11111, doi:10.1029/2004JD004536.

    • Search Google Scholar
    • Export Citation
  • Blender, R., , and M. Schubert, 2000: Cyclone tracking in different spatial and temporal resolutions. Mon. Wea. Rev., 128, 377384.

  • Bromwich, D. H., , R. L. Fogt, , K. I. Hodges, , and J. E. Walsh, 2007: A tropospheric assessment of the ERA-40, NCEP, and JRA-25 global reanalyses in the polar regions. J. Geophys. Res., 112, D10111, doi:10.1029/2006JD007859.

    • Search Google Scholar
    • Export Citation
  • Brunke, M. A., , Z. Wang, , X. Zeng, , M. Bosilovich, , and C.-L. Shie, 2011: An assessment of the uncertainties in ocean surface turbulent fluxes in 11 reanalysis, satellite-derived, and combined global datasets. J. Climate, 24, 54695493.

    • Search Google Scholar
    • Export Citation
  • Chen, J., , A. D. Del Genio, , B. E. Carlson, , and M. G. Bosilovich, 2008: The spatiotemporal structure of twentieth-century climate variations in observations and reanalyses. Part I: Long-term trend. J. Climate, 21, 26112633.

    • Search Google Scholar
    • Export Citation
  • Cullather, R. I., , and M. G. Bosilovich, 2011: The moisture budget of the polar atmosphere in MERRA. J. Climate, 24, 28612879.

  • Cullather, R. I., , and M. G. Bosilovich, 2012: The energy budget of the polar atmosphere in MERRA. J. Climate, 25, 524.

  • Dacre, H. F., , M. K. Hawcroft, , M. A. Stringer, , and K. I. Hodges, 2012: An extratropical cyclone atlas: A tool for illustrating cyclone structure and evolution characteristics. Bull. Amer. Meteor. Soc., 93, 1497–1502.

    • Search Google Scholar
    • Export Citation
  • Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553597.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K., 2005: Genesis and maintenance of “Mediterranean hurricanes.” Adv. Geosci., 2, 217220.

  • Fink, A. H., , T. Brücher, , E. Ermert, , A. Krüger, , and J. G. Pinto, 2009: The European Storm Kyrill in January 2007: Synoptic evolution and considerations with respect to climate change. Nat. Hazards Earth Syst. Sci., 9, 405423.

    • Search Google Scholar
    • Export Citation
  • Gulev, S. K., , O. Zolina, , and S. Grigoriev, 2001: Extratropical cyclone variability in the Northern Hemisphere winter from the NCEP/NCAR-Reanalysis data. Climate Dyn., 17, 795809.

    • Search Google Scholar
    • Export Citation
  • Hodges, K. I., , B. J. Hoskins, , J. Boyle, , and C. Thorncroft, 2003: A comparison of recent reanalysis datasets using objective feature tracking: Storm tracks and tropical easterly waves. Mon. Wea. Rev., 131, 20122037.

    • Search Google Scholar
    • Export Citation
  • Hodges, K. I., , R. W. Lee, , and L. Bengtsson, 2011: A comparison of extratropical cyclones in recent reanalyses ERA-Interim, NASA MERRA, NCEP CFSR, and JRA-25. J. Climate, 24, 48884906.

    • Search Google Scholar
    • Export Citation
  • Hoskins, B., , and K. Hodges, 2002: New perspectives on the Northern Hemisphere winter storm tracks. J. Atmos. Sci., 59, 10411061.

  • Jung, T., , S. K. Gulev, , I. Rudeva, , and V. Soloviov, 2006: Sensitivity of extratropical cyclone characteristics to horizontal resolution in the ECMWF model. Quart. J. Roy. Meteor. Soc., 132, 18391857.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471.

  • Kanamitsu, M., , W. Ebisuzaki, , J. Woollen, , S.-K. Yang, , J. J. Hnilo, , M. Fiorion, , and J. Potter, 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 16311643.

    • Search Google Scholar
    • Export Citation
  • König, W. R., , R. Sausen, , and F. Sielmann, 1993: Objective identification of cyclones in GCM simulations. J. Climate, 6, 22172231.

  • Liberato, M. R., , J. G. Pinto, , I. F. Trigo, , and R. M. Trigo, 2011: Klaus—An exceptional winter storm over northern Iberia and southern France. Weather, 66, 330334.

    • Search Google Scholar
    • Export Citation
  • Lim, E.-P., , and I. Simmonds, 2002: Explosive cyclone development in the Southern Hemisphere and a comparison with Northern Hemisphere events. Mon. Wea. Rev., 130, 21882209.

    • Search Google Scholar
    • Export Citation
  • Lionello, P., , and F. Giorgi, 2007: Winter precipitation and cyclones in the Mediterranean region: Future climate scenarios in a regional simulation. Adv. Geosci., 12, 153158.

    • Search Google Scholar
    • Export Citation
  • Livezey, R. E., , and W. Y. Chen, 1983: Statistical field significance and its determination by Monte Carlo techniques. Mon. Wea. Rev., 111, 4659.

    • Search Google Scholar
    • Export Citation
  • Löptien, U., , O. Zolina, , S. K. Gulev, , M. Latif, , and V. Soloviov, 2008: Cyclone life cycle characteristics over the Northern Hemisphere in coupled GCMs. Climate Dyn., 31, 507532.

    • Search Google Scholar
    • Export Citation
  • Mailier, P. J., , D. B. Stephenson, , C. A. T. Ferro, , and K. I. Hodges, 2006: Serial clustering of extratropical cyclones. Mon. Wea. Rev., 134, 22242240.

    • Search Google Scholar
    • Export Citation
  • Murray, R. J., , and I. Simmonds, 1991a: A numerical scheme for tracking cyclone centres from digital data. Part I: Development and operation of the scheme. Aust. Meteor. Mag., 39, 155166.

    • Search Google Scholar
    • Export Citation
  • Murray, R. J., , and I. Simmonds, 1991b: A numerical scheme for tracking cyclone centres from digital data. Part II: Application to January and July general circulation model simulations. Aust. Meteor. Mag., 39, 167180.

    • Search Google Scholar
    • Export Citation
  • Neu, U., and Coauthors, 2013: IMILAST—A community effort to intercompare extratropical cyclone detection and tracking algorithms. Bull. Amer. Meteor. Soc.,94, 529–547.

  • Onogi, K., and Coauthors, 2007: The JRA-25 reanalysis. J. Meteor. Soc. Japan, 85, 369432.

  • Orlanski, I., 1998: Poleward deflection of storm tracks. J. Atmos. Sci., 55, 25772602.

  • Pinto, J. G., , U. Ulbrich, , G. C. Leckebusch, , T. Spangehl, , M. Reyers, , and S. Zacharias, 2007: Changes in storm track and cyclone activity in three SRES ensemble experiments with the ECHAM5/MPI-OM1 GCM. Climate Dyn., 29, 195210.

    • Search Google Scholar
    • Export Citation
  • Pinto, J. G., , S. Zacharias, , A. H. Fink, , G. C. Leckebusch, , and U. Ulbrich, 2009: Factors contributing to the development of extreme North Atlantic cyclones and their relationship with the NAO. Climate Dyn., 32, 711737.

    • Search Google Scholar
    • Export Citation
  • Raible, C. C., 2007: On the relation between extremes of midlatitude cyclones and the atmospheric circulation using ERA-40. Geophys. Res. Lett., 34, L07703, doi:10.1029/2006GL029084.

    • Search Google Scholar
    • Export Citation
  • Raible, C. C., , and R. Blender, 2004: Northern Hemisphere midlatitude cyclone variability in GCM simulations with different ocean representations. Climate Dyn., 22, 239248.

    • Search Google Scholar
    • Export Citation
  • Raible, C. C., , P. Della-Marta, , C. Schwierz, , H. Wernli, , and R. Blender, 2008: Northern Hemisphere extratropical cyclones: A comparison of detection and tracking methods and different reanalyses. Mon. Wea. Rev., 136, 880897.

    • Search Google Scholar
    • Export Citation
  • Raible, C. C., , H. Saaroni, , B. Ziv, , and M. Wild, 2010: Winter synoptic-scale variability over the Mediterranean Basin under future climate conditions as simulated by the ECHAM5. Climate Dyn., 35, 473488.

    • Search Google Scholar
    • Export Citation
  • Rančić, M., , J. C. Derber, , D. Parrish, , R. Treadon, , and D. T. Kleist, 2008: The development of the first-order time extrapolation to the observation (FOTO) method and its application in the NCEP global data assimilation system. Proc. 12th Symp. on Integrated Observing and Assimilation Systems for the Atmosphere, Oceans, and Land Surface (IOAS-AOLS), New Orleans, LA, Amer. Meteor. Soc., J6.1. [Available online at http://ams.confex.com/ams/pdfpapers/131816.pdf.]

  • Rienecker, M. M., and Coauthors, 2011: MERRA: NASA's Modern-Era Retrospective Analysis for Research and Applications. J. Climate, 24, 36243648.

    • Search Google Scholar
    • Export Citation
  • Rogers, E., , and L. F. Bosart, 1986: An investigation of explosively deepening oceanic cyclones. Mon. Wea. Rev., 114, 702718.

  • Rudeva, I., , and S. K. Gulev, 2007: Climatology of cyclone size characteristics and their changes during the cyclone life cycle. Mon. Wea. Rev., 135, 25682587.

    • Search Google Scholar
    • Export Citation
  • Rudeva, I., , and S. K. Gulev, 2011: Composite analysis of North Atlantic extratropical cyclones in NCEP–NCAR reanalysis data. Mon. Wea. Rev., 139, 14191446.

    • Search Google Scholar
    • Export Citation
  • Saha, S., and Coauthors, 2010: The NCEP Climate Forecast System. J. Climate, 19, 34833517.

  • Sanders, F., , and J. R. Gyakum, 1980: Synoptic-dynamic climatology of the “bomb.” Mon. Wea. Rev., 108, 15891606.

  • Screen, J. A., , I. Simmonds, , and K. Keay, 2011: Dramatic interannual changes of perennial Arctic sea ice linked to abnormal summer storm activity. J. Geophys. Res., 116, D15105, doi:10.1029/2011JD015847.

    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., , and A. P. Barrett, 2008: The summer cyclone maximum over the central Arctic Ocean. J. Climate, 21, 10481065.

  • Simmonds, I., , and K. Keay, 2000a: Mean Southern Hemisphere extratropical cyclone behavior in the 40-year NCEP–NCAR reanalysis. J. Climate, 13, 873885.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., , and K. Keay, 2000b: Variability of Southern Hemisphere extratropical cyclone behavior, 1958–97. J. Climate, 13, 550561.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., , and K. Keay, 2009: Extraordinary September Arctic sea ice reductions and their relationships with storm behavior over 1979–2008. Geophys. Res. Lett., 36, L19715, doi:10.1029/2009GL039810.

    • Search Google Scholar
    • Export Citation
  • Simmons, A. J., , S. Uppala, , D. Dee, , and S. Kobayashi, 2007: ERA-Interim: New ECMWF reanalysis products from 1989 onwards. ECMWF Newsletter, No. 110, ECMWF, Reading, United Kingdom, 25–35.

  • Simmons, A. J., , K. M. Willett, , P. D. Jones, , P. W. Thorne, , and D. P. Dee, 2010: Low-frequency variations in surface atmospheric humidity, temperature, and precipitation: Inferences from reanalyses and monthly gridded observational data sets. J. Geophys. Res., 115, D01110, doi:10.1029/2009JD012442.

    • Search Google Scholar
    • Export Citation
  • Sinclair, M. R., 1997: Objective identification of cyclones and their circulation, intensity, and climatology. Wea. Forecasting, 12, 595612.

    • Search Google Scholar
    • Export Citation
  • Sinclair, M. R., , and I. G. Watterson, 1999: Objective assessment of extratropical weather systems in simulated climates. J. Climate, 12, 34673485.

    • Search Google Scholar
    • Export Citation
  • Sterl, A., 2004: On the (in)homogeneity of reanalysis products. J. Climate, 17, 38663873.

  • Stewart, R. E., , and N. R. Donaldson, 1989: On the nature of rapidly deepening Canadian East Coast winter storms. Atmos.–Ocean, 27, 87107.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., , J. T. Fasullo, , and J. Mackaro, 2011: Atmospheric moisture transports from ocean to land and global energy flows in reanalyses. J. Climate, 24, 49074924.

    • Search Google Scholar
    • Export Citation
  • Trigo, I. F., 2006: Climatology and interannual variability of storm-tracks in the Euro-Atlantic sector: A comparison between ERA-40 and NCEP/NCAR reanalyses. Climate Dyn., 26, 127143.

    • Search Google Scholar
    • Export Citation
  • Ulbrich, U., , and M. Christoph, 1999: A shift of the NAO and increasing storm track activity over Europe due to anthropogenic greenhouse gas forcing. Climate Dyn., 15, 551559.

    • Search Google Scholar
    • Export Citation
  • Ulbrich, U., , T. Brücher, , A. H. Fink, , G. C. Leckebusch, , A. Krüger, , and J. G. Pinto, 2003: The central European floods in August 2002: Part I—Rainfall periods and flood development. Weather, 58, 371376.

    • Search Google Scholar
    • Export Citation
  • Ulbrich, U., , J. G. Pinto, , H. Kupfer, , G. C. Leckebusch, , T. Spangehl, , and M. Reyers, 2008: Changing Northern Hemisphere storm tracks in an ensemble of IPCC climate change simulations. J. Climate, 21, 16691679.

    • Search Google Scholar
    • Export Citation
  • Uppala, S. M., and Coauthors, 2005: The ERA-40 Re-Analysis. Quart. J. Roy. Meteor. Soc., 131, 29613012.

  • Wang, X. L., , H. Wan, , and V. R. Swail, 2006: Observed changes in cyclone activity in Canada and their relationships to major circulation regimes. J. Climate, 19, 896915.

    • Search Google Scholar
    • Export Citation
  • White, G., 2000: Long-term trends in the NCEP/NCAR reanalysis. Proc. Second Int. Conf. on Reanalyses, Reading, United Kingdom, WMO, 54–57.

  • Woollings, T., 2008: Vertical structure of anthropogenic zonal-mean atmospheric circulation change. Geophys. Res. Lett., 35, L19702, doi:10.1029/2008GL034883.

    • Search Google Scholar
    • Export Citation
  • Woollings, T., , and M. Blackburn, 2012: The North Atlantic jet stream under climate change and its relation to the NAO and EA patterns. J. Climate, 25, 886902.

    • Search Google Scholar
    • Export Citation
  • Yau, M. K., , and M. Jean, 1989: Synoptic aspects and physical processes in the rapidly intensifying cyclone of 6–8 March 1986. Atmos.–Ocean, 27, 5986.

    • Search Google Scholar
    • Export Citation
  • Yin, J. H., 2005: A consistent poleward shift of the storm tracks in simulations of 21st century climate. Geophys. Res. Lett., 32, L18701, doi:10.1029/2005GL023684.

    • Search Google Scholar
    • Export Citation
  • Zolina, O., , and S. K. Gulev, 2002: Improving the accuracy of mapping cyclone numbers and frequencies. Mon. Wea. Rev., 130, 748759.

  • Zolina, O., , and S. K. Gulev, 2003: Synoptic variability of ocean–atmosphere turbulent fluxes associated with atmospheric cyclones. J. Climate, 16, 30233041.

    • Search Google Scholar
    • Export Citation
Save