• Adams, C. R., 1997: Impacts of temperature extremes. Proc., Workshop on the Social and Economic Impacts of Weather, Boulder, CO, U.S. Weather Research Program, 11–16.

    • Search Google Scholar
    • Export Citation
  • Alexander, L. V., P. Hope, D. Collins, B. Trewin, A. Lynch, and N. Nicholls, 2007: Trends in Australia’s climate means and extremes: A global context. Aust. Meteor. Mag., 56 , 118.

    • Search Google Scholar
    • Export Citation
  • Bureau of Meteorology, cited 2003: Climate activities in Australia. [Available online at http://www.bom.gov.au/inside/eiab/reports/caa03/chapter4/disease_human_health.shtml].

    • Search Google Scholar
    • Export Citation
  • Camilloni, I., and V. Barros, 1997: On the urban heat island effect dependence on temperature trends. Climatic Change, 37 , 665681.

  • Colle, B. A., and C. E. Mass, 1995: The structure and evolution of cold surges east of the Rocky Mountains. Mon. Wea. Rev., 123 , 25772610.

    • Search Google Scholar
    • Export Citation
  • Colucci, S. J., D. P. Baumhefner, and C. E. Konrad, 1999: Numerical prediction of a cold air outbreak: A case study with ensemble forecasts. Mon. Wea. Rev., 127 , 15381550.

    • Search Google Scholar
    • Export Citation
  • Curry, J. A., and P. J. Webster, 1999: Thermodynamics of Atmospheres and Oceans. International Geophysics Series, Vol. 65, Academic Press, 471 pp.

    • Search Google Scholar
    • Export Citation
  • Dare, R. A., and N. E. Davidson, 2004: Characteristics of tropical cyclones in the Australian region. Mon. Wea. Rev., 132 , 30493065.

  • Elliott, P. R., 1989: The Tasmanian snowstorm 24 and 25 July 1986. Meteorology Note 187, Bureau of Meteorology Australia, 34 pp.

  • Emergency Management Australia, cited 2008: EMA disaster database—Crop damage: Goulburn Valley, VIC. [Available online at http://www.ema.gov.au/ema/emadisasters.nsf/83edbd0553620d8cca256d09001fc8fd/32e4b08a3ec179dcca2572cf0023765a?OpenDocument].

    • Search Google Scholar
    • Export Citation
  • Foley, G. R., and B. N. Hanstrum, 1994: The capture of tropical cyclones by cold fronts off the west coast of Australia. Wea. Forecasting, 9 , 577592.

    • Search Google Scholar
    • Export Citation
  • Fortune, M. A., and V. E. Kousky, 1983: Two severe freezes in Brazil: Precursors and synoptic evolution. Mon. Wea. Rev., 111 , 181196.

    • Search Google Scholar
    • Export Citation
  • Garreaud, R. D., 2000: Cold air incursions over subtropical South America: Mean structure and dynamics. Mon. Wea. Rev., 128 , 25442559.

    • Search Google Scholar
    • Export Citation
  • Garreaud, R. D., 2001: Subtropical cold surges: Regional aspects and global distribution. Int. J. Climatol., 21 , 11811197.

  • Garreaud, R. D., and J. M. Wallace, 1998: Summertime incursions of midlatitude air into subtropical and tropical South America. Mon. Wea. Rev., 126 , 27132733.

    • Search Google Scholar
    • Export Citation
  • Hartjenstein, G., and R. Bleck, 1991: Factors affecting cold-air outbreaks east of the Rocky Mountains. Mon. Wea. Rev., 119 , 22802292.

    • Search Google Scholar
    • Export Citation
  • Hope, P. K., W. Drosdowsky, and N. Nicholls, 2006: Shifts in the synoptic systems influencing southwest Western Australia. Climate Dyn., 26 , 751764.

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

  • Kistler, R., and Coauthors, 2001: The NCEP–NCAR 50-Year Reanalysis: Monthly means CD-ROM and documentation. Bull. Amer. Meteor. Soc., 82 , 247267.

    • Search Google Scholar
    • Export Citation
  • Konrad, C. E., 1996: Relationships between the intensity of cold-air outbreaks and the evolution of synoptic and planetary scale features over North America. Mon. Wea. Rev., 124 , 10671083.

    • Search Google Scholar
    • Export Citation
  • Konrad, C. E., 1998: Persistent planetary scale circulation patterns and their relationship with cold air outbreak activity over the eastern United States. Int. J. Climatol., 18 , 12091221.

    • Search Google Scholar
    • Export Citation
  • Krishnamurti, T. N., M. Tewari, and D. R. Chakraborty, 1999: Downstream amplification: A possible precursor to major freeze events over southeastern Brazil. Wea. Forecasting, 14 , 242270.

    • 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
  • Lim, E-P., and I. Simmonds, 2007: Southern Hemisphere winter extratropical cyclone characteristics and vertical organization observed with the ERA-40 data in 1979–2001. J. Climate, 20 , 26752690.

    • Search Google Scholar
    • Export Citation
  • Marengo, J. A., and J. C. Rogers, 2001: Polar air outbreaks in the Americas: Assessments and impacts during modern and past climates. Interhemispheric Climate Linkages, V. Markgraf, Ed., Academic Press, 17–51.

    • Search Google Scholar
    • Export Citation
  • Marengo, J. A., A. Cornejo, P. Satyamurty, and C. Nobre, 1997: Cold surges in tropical and extratropical South America: The strong event in June 1994. Mon. Wea. Rev., 125 , 27592786.

    • Search Google Scholar
    • Export Citation
  • Marengo, J. A., T. Ambrizzi, G. Kiladis, and B. Liebmann, 2002: Upper-air wave trains over the Pacific Ocean and wintertime cold surges in tropical-subtropical South America leading the freezes in southern and southeastern Brazil. Theor. Appl. Climatol., 73 , 223242.

    • Search Google Scholar
    • Export Citation
  • Morley, K. T., 1957: Outbreaks of Antarctic air over southeast Australia. Aust. Meteor. Mag., 19 , 6263.

  • Morris, C. J. G., and I. Simmonds, 2000: Associations between varying magnitudes of the urban heat island and the synoptic climatology in Melbourne, Australia. Int. J. Climatol., 20 , 19311954.

    • Search Google Scholar
    • Export Citation
  • Müller, G. V., and G. J. Berri, 2007: Atmospheric circulation associated with persistent generalized frosts in central-southern South America. Mon. Wea. Rev., 135 , 12681289.

    • Search Google Scholar
    • Export Citation
  • Murray, R. J., and I. Simmonds, 1991a: A numerical scheme for tracking cyclone centres from digital data. Part 1: 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 2: Application to January and July general circulation model simulation. Aust. Meteor. Mag., 39 , 167180.

    • Search Google Scholar
    • Export Citation
  • Perrin, G., and I. Simmonds, 1995: The origin and characteristics of cold air outbreaks over Melbourne. Aust. Meteor. Mag., 44 , 4159.

    • Search Google Scholar
    • Export Citation
  • Pezza, A. B., and T. Ambrizzi, 2005: Dynamical conditions and synoptic tracks associated with different types of cold surge over tropical South America. Int. J. Climatol., 25 , 215241.

    • Search Google Scholar
    • Export Citation
  • Pezza, A. B., and I. Simmonds, 2008: Large scale factors in tropical and extratropical cyclone transition and extreme weather events. Ann. N. Y. Acad. Sci., 1146 , 189211.

    • Search Google Scholar
    • Export Citation
  • Pezza, A. B., I. Simmonds, and J. A. Renwick, 2007: Southern Hemisphere cyclones and anticyclones: Recent trends and links with decadal variability in the Pacific Ocean. Int. J. Climatol., 21 , 14031419.

    • Search Google Scholar
    • Export Citation
  • Rogers, J. C., and R. V. Rohli, 1991: Florida citrus freezes and polar anticylones in the Great Plains. J. Climate, 4 , 11031113.

  • Salinger, M. J., and G. M. Griffiths, 2001: Trends in New Zealand daily temperature and rainfall extremes. Int. J. Climatol., 21 , 14371452.

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

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

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., and T. Richter, 2000: Synoptic comparison of cold events in winter and summer in Melbourne and Perth. Theor. Appl. Climatol., 67 , 1932.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., and H. A. Rashid, 2001: An investigation of a dramatic cold outbreak over southeast Australia. Aust. Meteor. Mag., 50 , 249261.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., K. Keay, and E-P. Lim, 2003: Synoptic activity in the seas around Antarctica. Mon. Wea. Rev., 131 , 272288.

  • Trenberth, K. E., and Coauthors, 2007: Observations: Surface and atmospheric climate change. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 299–316.

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

  • Vavrus, S., J. E. Walsh, W. L. Chapman, and D. Portis, 2006: The behaviour of extreme cold air outbreaks under greenhouse warming. Int. J. Climatol., 26 , 11331147.

    • Search Google Scholar
    • Export Citation
  • Vera, C. S., and P. K. Vigliarolo, 2000: A diagnostic study of cold-air outbreaks over South America. Mon. Wea. Rev., 128 , 324.

  • Vincent, L. A., and Coauthors, 2005: Observed trends in indices of daily temperature extremes in South America 1960–2000. J. Climate, 18 , 50115023.

    • Search Google Scholar
    • Export Citation
  • Wiedenmann, J. M., A. R. Lupo, I. I. Mokhov, and E. A. Tikhonova, 2002: The climatology of blocking anticyclones for the Northern and Southern Hemispheres: Block intensity as a diagnostic. J. Climate, 15 , 34593473.

    • Search Google Scholar
    • Export Citation
  • View in gallery
    Fig. 1.

    Annual composite of MSLP (shaded) and MSLP anomalies (contoured) over (a) Melbourne and (b) Perth outbreak day, averaged from CEs occurring from 1958 to 2006. Areas of 95% significance are stippled. The solid boxes identify the 5° × 5° reference areas for the statistics discussed in the text, while the dashed boxes indicate the regions containing the majority of pressure couplets. Red asterisk indicates the approximate location of Melbourne in (a) and Perth in (b).

  • View in gallery
    Fig. 2.

    Anticyclone and cyclone tracks associated with cold events in Melbourne from 1958 to 2001 split into pre- and postevent components. The dots on the tracks give the location of the system center every 6 hours. (a) Dark blue shows the pre-event anticyclones and (b) light blue gives the respective postevent movement. (c) Red shows the pre-event cyclones and (d) orange the postevent cyclone progression. The average path is shown as a gray line. The green and yellow circles mark the average location of the anticyclonic and cyclonic pressure systems respectively on event day, and the squares give the median genesis point. The symbols at the bottom of each plot refer to the average speed, track duration, and distance traveled of CE-associated systems (colored) compared to the climatology of non-CE systems (black).

  • View in gallery
    Fig. 3.

    System density of anticyclones and cyclones associated with cold events in (a),(b) Melbourne and (c),(d) Perth from 1958 to 2001. Colored contours show the mean number of systems found in a 103 (° latitude)−2 area over the study period. Note that the scale size is different for each system type.

  • View in gallery
    Fig. 4.

    The average value and standard deviation of (a) TD, (b) DT, and (c) SP of pre- and post-CE track components for CE-associated systems and climatological, non-CE systems.

  • View in gallery
    Fig. 5.

    DP anomalies (hPa) of anticyclonic and cyclonic systems associated with CEs in (a),(b) Melbourne and (c),(d) Perth. The smoothing used is a Cressman-type with a 5.4° radius and a 0.09 pinching factor.

  • View in gallery
    Fig. 6.

    Composites of daily MSLP anomalies (contoured, hPa) with MSLP (shaded, hPa) for day −5 to day 0 for CEs in Melbourne. For the day 0 plot, areas of 95% significance are stippled.

  • View in gallery
    Fig. 7.

    Composites of daily geopotential height anomalies at 300 hPa (m) for day −5 to day 0 for CEs in Melbourne. For the day 0 plot, areas of 95% significance are stippled.

  • View in gallery
    Fig. 8.

    Composites of 300-hPa vector wind from day −2 to day 0 for CEs in Melbourne. Shading shows wind speed (m s−1). Wind direction is shown by streamlines.

  • View in gallery
    Fig. 9.

    Average surface skin temperature anomalies (°C) over the ocean for the seven days before a CE in (a) Melbourne summer and (b) Perth winter. Areas marked with a cross indicate regions of 90% significance.

  • View in gallery
    Fig. 10.

    Seasonal frequency analysis of CEs in (a) Melbourne and (b) Perth from 1958 to 2006 with annual linear trend overlaid.

  • View in gallery
    Fig. 11.

    Magnitude of difference between seasonal average maximum temperatures and event day maximum temperatures for each season in (a) Melbourne and (b) Perth from 1958 to 2006. For seasons exhibiting a significant or near-significant trend, the R2 value is given.

  • View in gallery
    Fig. 12.

    Annual running averages of the CEI and maximum daily temperature anomalies for (a) Melbourne and (b) Perth from 1958 to 2006. See text for further details.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 241 124 11
PDF Downloads 93 49 7

Cold Events over Southern Australia: Synoptic Climatology and Hemispheric Structure

Linden Claire AshcroftSchool of Earth Sciences, The University of Melbourne, Melbourne, Victoria, Australia

Search for other papers by Linden Claire Ashcroft in
Current site
Google Scholar
PubMed
Close
,
Alexandre Bernardes PezzaSchool of Earth Sciences, The University of Melbourne, Melbourne, Victoria, Australia

Search for other papers by Alexandre Bernardes Pezza in
Current site
Google Scholar
PubMed
Close
, and
Ian SimmondsSchool of Earth Sciences, The University of Melbourne, Melbourne, Victoria, Australia

Search for other papers by Ian Simmonds in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

Cold events (CEs) are an important feature of southern Australian weather. Unseasonably cold conditions can have a significant impact on Australia’s agricultural industry and other aspects of society. In this study the bottom 0.4% of maximum temperatures in Melbourne and Perth from the 1958–2006 period are defined as CEs, representing the large-scale patterns affecting most of extratropical Australia. Compiling 6-hourly progressions of the tracks of the cyclones and anticyclones that are geostrophically associated with CEs gives for the first time a detailed synoptic climatology over the area.

The anticyclone tracks display a “cloud” of high density across the Indian Ocean, which is linked, in the mean, to weak but significant negative SST anomalies in the region. The cyclone tracks display much variability, with system origins ranging from subpolar to tropical. Several CEs are found to involve tropical and extratropical interaction or extratropical transition of originally tropical cyclones (hurricanes). CE-associated systems travel farther and exhibit longer life spans than similar, non-CE systems. Upper-level analyses indicate the presence of a wave train originating more than 120° west of the CE. This pattern greatly intensifies over the affected area in conjunction with a merging of the subpolar and subtropical jets. The upper-level wave train is present up to five days before the CE. The absence of large orographic features in Australia highlights the importance of wave amplification in CE occurrence. No consistent trend in CE intensity over the period is found, but a significant negative trend in event frequency is identified for both Melbourne and Perth.

Corresponding author address: Alexandre Bernardes Pezza, School of Earth Sciences, The University of Melbourne, Melbourne, Victoria, 3010, Australia. Email: apezza@unimelb.edu.au

Abstract

Cold events (CEs) are an important feature of southern Australian weather. Unseasonably cold conditions can have a significant impact on Australia’s agricultural industry and other aspects of society. In this study the bottom 0.4% of maximum temperatures in Melbourne and Perth from the 1958–2006 period are defined as CEs, representing the large-scale patterns affecting most of extratropical Australia. Compiling 6-hourly progressions of the tracks of the cyclones and anticyclones that are geostrophically associated with CEs gives for the first time a detailed synoptic climatology over the area.

The anticyclone tracks display a “cloud” of high density across the Indian Ocean, which is linked, in the mean, to weak but significant negative SST anomalies in the region. The cyclone tracks display much variability, with system origins ranging from subpolar to tropical. Several CEs are found to involve tropical and extratropical interaction or extratropical transition of originally tropical cyclones (hurricanes). CE-associated systems travel farther and exhibit longer life spans than similar, non-CE systems. Upper-level analyses indicate the presence of a wave train originating more than 120° west of the CE. This pattern greatly intensifies over the affected area in conjunction with a merging of the subpolar and subtropical jets. The upper-level wave train is present up to five days before the CE. The absence of large orographic features in Australia highlights the importance of wave amplification in CE occurrence. No consistent trend in CE intensity over the period is found, but a significant negative trend in event frequency is identified for both Melbourne and Perth.

Corresponding author address: Alexandre Bernardes Pezza, School of Earth Sciences, The University of Melbourne, Melbourne, Victoria, 3010, Australia. Email: apezza@unimelb.edu.au

1. Introduction

Cold air events (CEs) occur in many regions around the world. Typified by unseasonably cold temperatures and associated meteorological phenomena, these extreme events can influence many facets of society. The most documented impact of CEs is economic damage to the agriculture sector. The Australian fruit industry is often adversely affected by CEs, with fruit crop damage worth up to $500 million occurring in recent years (Emergency Management Australia 2008). The large South American coffee production industry has also been greatly affected by CEs (Fortune and Kousky 1983; Marengo et al. 1997), and impacts on orange crops in North America have been widely recorded (Rogers and Rohli 1991; Colucci et al. 1999). Livestock loss, transportation disruptions, and increased energy demands are also associated with CEs in Australia and around the world, along with an increased health risk to elderly or vulnerable citizens (Adams 1997; Bureau of Meteorology 2003).

As it is wintertime CE occurrence that is more often associated with frosts, snow, and similar crop-damaging phenomena, many studies focus on this time of year. However, anomalous cold days can have an impact on society regardless of the season in which they occur. Many elements of Australian industry such as grazing and wool production can be adversely affected by a CE occurring during the warmer months (Simmonds and Richter 2000). Garreaud and Wallace (1998) also found that summertime CEs are the dominant mode of precipitation variability over much of South America.

The primary mechanism behind CEs is a surface anticyclone–cyclone couplet (or dipole) located over the affected region. The strong geostrophic wind anomaly located between the two systems forces cold advection and an equatorward incursion of high-latitude air (see Konrad 1998 for North America; Simmonds and Richter 2000 for Australia; Marengo and Rogers 2001 for South America). In the case of Australian CEs, this air generally originates from the high latitudes of the Southern Ocean (Perrin and Simmonds 1995). The couplet pattern is similar regardless of the season of CE occurrence (Garreaud 2000; Simmonds and Richter 2000), although few studies have examined the paths traveled by the systems with a view to identify their characteristic genesis locations, trajectories, and life spans in the lead up to a CE event. Pezza and Ambrizzi (2005) have examined the trajectories of systems associated with severe cold events over tropical South America during winter months; however, an Australian comparative study has not yet been conducted.

CEs over South and North America have been shown to be greatly intensified by the presence of the Andes and Rocky Mountains, respectively, with the mountains creating a channeling effect and forcing the cold air farther equatorward (Hartjenstein and Bleck 1991; Colle and Mass 1995; Garreaud 2000; Vera and Vigliarolo 2000; Garreaud 2001; Pezza and Ambrizzi 2005). It is important to note that despite the absence of a large mountain range in Australia, CEs have still influenced large areas of the continent, and can reach subtropical latitudes (Pezza and Simmonds 2008).

CEs are intrinsically associated with the amplification of baroclinic energy wave patterns. These can be due to pure baroclinic growth, topographic interactions, or teleconnection effects. Several studies have identified wave trains that develop from perturbations (convective or otherwise), generating gravity waves that will teleconnect to the extratropics and propagate long distances before ultimately forcing equatorward incursion (Krishnamurti et al. 1999; Marengo et al. 2002). Other research points to the interaction between, and superposition of, synoptic-scale waves in the midlatitudes as possible reasons for wave pattern amplification (Konrad 1996; Marengo et al. 1997; Krishnamurti et al. 1999; Müller and Berri 2007).

Compared to the number of papers published on CE occurrence in other areas of the world, the number of studies focusing on Australian CEs is quite small. Several earlier papers discussed cold incursions over southeast Australia (Morley 1957; Elliott 1989), and more recent studies have also focused on the impact of CEs over Melbourne (see Perrin and Simmonds 1995; Simmonds and Rashid 2001). The study by Simmonds and Richter (2000) examines cold events in Melbourne and Perth during summer and winter. Several different synoptic patterns associated with CEs were identified and, while the synoptic interactions did differ somewhat between them, the fundamental couplet structure was present for each type.

An important aspect of extreme events is their reaction to climatic variations, as it is these events that will have the greatest social impact in a changing climate (Trenberth et al. 2007). Several studies suggest that the frequency and intensity of extreme cold days is decreasing (Salinger and Griffiths 2001; Alexander et al. 2007). However, others find no significant decrease in the number or intensity of cold maximum temperatures or suggest that, while event frequencies will decrease in some regions by the end of this century, in others the frequency will remain the same or increase because of atmospheric circulation changes (Vincent et al. 2005; Vavrus et al. 2006).

This study examines the nature of CEs in several ways. Initially, a definition is required that will identify CEs in all seasons over the east and west of southern Australia, represented by Melbourne and Perth. From this, a hemispheric synoptic climatology of cyclones and anticyclones associated with these events is produced and dissected. Next, the average large-scale atmospheric situation and wave amplification mechanisms associated with the CEs are explored. The climate variability of these extreme events is also examined. Finally, a large-scale index is proposed that will place the conditions conducive to cold advection in climate context, irrespective of station data variability.

2. Method and data

This study employed a percentile-dependent temperature-based definition. The daily maximum temperatures in the bottom 0.4% among the entire series are classified as CE days, calculated on a monthly basis. If the temperature distribution of the areas of interest was normal, this would also equate to approximately three standard deviations below the mean temperature of each month. The days identified were divided evenly between the months, leading to the 6 coldest days in each month being classified. If a day reached an identical maximum temperature but was outside the 6 days (e.g., if the seventh coldest day had the same temperature as the sixth) it was also identified as a CE. This definition ensured that the events identified were extreme in nature, while returning a sample large enough to allow robust conclusions to be drawn from a statistical viewpoint.

Observed daily maximum temperatures for the period 1 January 1958 to 31 December 2006 were taken from the Australian Bureau of Meteorology high quality dataset for the Melbourne regional office (37.8°S, 144.8°E) and Perth airport (31.9°S, 116.0°E). The definition described above subsequently identified 80 CEs for Melbourne and 83 for Perth over the 48-year study period. Table 1 lists the number of CEs identified in each month for the two cities as well as the average maximum temperature and temperature anomalies (from mean monthly maximum) reached.

The Melbourne University Tracking Scheme (MUTS) was employed to plot CE-associated system tracks. MUTS uses mean sea level pressure (MSLP) data and works on the principle that the center of a low (high) pressure system is located within one grid space of the maximum (minimum) of the Laplacian of pressure (Murray and Simmonds 1991a,b; Simmonds and Keay 2000; Simmonds et al. 2003; Pezza and Ambrizzi 2005). The scheme also calculates valuable system statistics such as density (SD) and depth (DP). SD gives the number of cyclones or anticyclones present in a given area, while DP represents the pressure difference between the center and the edge of a system and gives an indication of the relative strength of a system (Lim and Simmonds 2007; Pezza et al. 2007).

Six-hourly MSLP reanalysis on a 2.5° × 2.5° grid from the 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) dataset (Uppala et al. 2005) for 1958–2001 was used for plotting system tracks with the MUTS. The National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) revised online reanalysis set (available at www.cdc.noaa.gov) (Kalnay et al. 1996; Kistler et al. 2001) was used for atmospheric circulation analysis and to identify the systems associated with the defined CEs. The results obtained using the NCEP–NCAR reanalysis agree well with the MUTS trajectories using ERA-40 data. The identification of each cyclone and anticyclone associated with the CEs was completed by diagnosing the systems geostrophically associated with the cold surge on a case-by-case basis as in Pezza and Ambrizzi (2005). Sea surface temperature (SST) information was adapted from the NCEP–NCAR surface skin temperature data.

3. Synoptic climatology

a. Anticyclone–cyclone couplet

Figure 1 shows the average MSLP and MSLP anomaly for Melbourne (Fig. 1a) and Perth (Fig. 1b) outbreak days for all CEs occurring from 1958 to 2001. From these figures the regional pressure couplet is clear, as is the associated northward geostrophic flow between the two systems. The pressure couplet is the dominant pattern, being statistically significant for the annual composite (84% of the cases in Melbourne and 94% in Perth). This is also true for all seasons (not shown) apart from Melbourne winter, which involved a significant number of blocking anticyclone type CEs (Simmonds and Richter 2000). While the MSLP anomalies display two closed pressure systems, the pressure shadings show the average cyclone component of the couplet as a trough extending to the east of the CE-affected region because of the high spatial variation associated with the cyclonic component of the couplet on event day. The solid and dashed boxes correspond to areas of maximum concentration of couplets when the overall variability taken into account. The larger dashed boxes are used to calculate the cold event index discussed in section 5, and the small solid boxes (5° × 5°) are used to derive statistical properties of the CE-related tracks [discussed in section 3b(2)].

b. System trajectories

Figure 2 shows the anticyclone and cyclone tracks associated with Melbourne CEs from 1958 to 2001. The corresponding pattern for Perth CEs (not shown) is similar when account is taken of the westward shift. The mean, or “center of gravity,” of the set of system tracks was evaluated for every 6 hours for as long as the number of systems present was greater than five. This mean path is plotted in gray. It is clear that the anticyclone tracks are confined to a smaller “cloud” than cyclone tracks, with the anticyclones following a clear arc within 30°–60°S. The CE-associated cyclones conversely vary greatly in the paths taken, with tracks ranging from polar to tropical origins for events in both cities. The “tropical interaction” type events have been previously identified by Simmonds and Richter (2000) and were more common for Perth CEs than Melbourne (not shown). The statistics properties indicated in the legend (speed, duration and distance traveled) are discussed in detail in section 3b(2).

1) System density

To examine the amount of variability of system tracks and locations on event day, Fig. 3 shows the SD of the anticyclones (Figs. 3a,c) and cyclones (Figs. 3b,d) tracked for Melbourne and Perth. Figures 3a,b show that the average location on event day for Melbourne CE-associated systems (marked in Figs. 2a,c) correlate well to the areas of maximum system density. This illustrates that there is much less spatial variation of the system locations on event day than there is before and after CE occurrence.

The remaining regions of anticyclone density coincide closely with the average track, with the majority of systems passing through the region from 90° to 170°E and 30° to 40°S. While there are longer system paths apparent in Fig. 2a, the SD plot does not identify them, indicating that they are infrequent. The cyclone SD pattern echoes the average system track post-CE; however, no systems are seen to the west of 130°E. This is in stark contrast to the many cyclones identified in Fig. 2c passing through this region before CE occurrence. This result shows quantitatively that the path taken by the cyclones pre-CE is much less uniform than that taken by the anticyclones.

The average location of the anticyclone associated with Perth CEs on event day also corresponds well to the area of highest density identified at around 35°S, 90°E (Fig. 3c). In general there was less deviation in the Perth anticyclone location than that of Melbourne. Figure 3d shows that the area of highest cyclone density for Perth is located farther north than the average event day location, with a second, smaller density peak corresponding with this point. This SD pattern shows the higher level of tropical interaction associated with Perth CEs and suggests that CE-associated cyclones originating from lower latitudes follow a consistent path. This plot also suggests that there are two distinct regions from which the systems originate; one in the extratropical region off the southwestern coast of Australia and the second from around the Western Australian coast at about 25°S.

2) Track characteristics

The mean and standard deviation of system track characteristics: duration in days (TD), distance traveled in km (DT), and speed in km day−1 (SP) are shown in Fig. 4. For Melbourne these properties are also graphically represented by the symbols in Fig. 2. To interpret these results relative to the climatological means, averages of these parameters were calculated for all systems occurring over the study period that passed through a 5° × 5° box located over the average system location on event day. The coordinates of each box are shown in Fig. 1.

The cyclones associated with CEs in both cities travel for much longer after CE occurrence than before with systems living over 3 days or 150% longer for Melbourne and 50% longer for Perth post-CE (Fig. 4a). This is presumably due to the interaction between the systems and the cold air mass as well as intensification associated with poleward progression.

From the climatological TD averages, Melbourne CE-associated anticyclones and cyclones exhibit a life span up to 50% greater than the average, non-CE-associated systems. The anomalous cyclonic track duration is likely due to the cold air interaction previously discussed, and the longer-than-average anticyclonic life span is believed to be due to large-scale interactions.

Examining the distance traveled by the systems in Fig. 4b yields a similar pattern as TD. The anticyclones travel on average over twice as far as the cyclonic systems pre-CE, indicating that the cyclones are more local in nature. The Melbourne cases display DT values that are on average 37% greater than the climatological averages, showing that they travel farther, as well as live longer, than the average, non-CE-associated system.

The SP plot in Fig. 4c shows that CE-associated anticyclones travel faster before CE interaction than afterward for both cities. Turning to the CE-associated cyclonic systems, a 14% faster-than-climatological-average speed is observed before CE interaction for both cities. This is the result of subpolar jet stream interactions, which are explored in section 4.

3) Depth

The average depth anomaly values of the CE-associated systems compared with the 1958–2001 climatology as calculated from the MUTS are plotted for the Melbourne cases (Figs. 5a,b) and those for Perth (Figs. 5c,d).

For Melbourne, an anomalous increase in DP around the average location of the cyclones on event day is observed, with a “tongue” of positive DP anomalies of up to 3 hPa present in the Tasman Sea (Fig. 5b). A small anomalous DP increase (around 1 hPa) in the anticyclones around their average location on event day is also apparent, as along with an area of greater DP anomaly over the southern Indian Ocean (Fig. 5a). The anticyclone figure also indicates that the systems that continue past Australia after a CE have their DP increased, remaining 1–2 hPa deeper than the climatological average. No anticyclone DP anomaly is present directly over Melbourne, however, indicating that event-associated anticyclones that do not continue to the Pacific decay as they cross the country because of the influence of the cyclone component of the couplet present in this region. This is a consequence of energy dispersion associated with the demise of the CE.

For Perth CEs it was found that the DP anomaly pattern of the cyclones (Fig. 5d) is comparable to those of Melbourne. The anticyclone DP distribution (Fig. 5c) shows a region of maximum depth around 10° south of the system location on event day and also suggests an intensification of the anticyclones that continue to travel southeastward after being associated with a cold incursion in Perth.

4. Large-scale structure and association links

a. Surface pressure

Figure 6 shows the composite MSLP anomaly and MSLP over the Southern Hemisphere for the five days (days −5 to −1) before each CE, as well as event day (day 0) for Melbourne. Perth cases are not shown as they have a similar pattern to Melbourne, shifted to the west. A 5-day period was chosen because it became apparent that this period can hold, on average, the fundamental signature associated with CEs in southern Australia. On an individual basis the CE signal can eventually be traced much further back in time. Areas that are statistically significant above 95% according to the Student’s t test are shown for the day-0 plot.

The main feature in these plots is the easterly progression of the positive and negative anomalies found over the event region on day 0, as seen locally in Fig. 1. A smaller positive anomaly to the east of the couplet is also apparent. These anomalies are significantly different from zero (as per Fig. 1) and are also the largest in the hemisphere, indicating the anomalous and extreme nature of the CEs being examined. For both Melbourne and Perth the positive anomaly associated with the couplet can be identified from day −5 approaching Australia from the west. What is also common between the cases for the two cities is the apparent weakening of the anticyclone anomaly and strengthening of the cyclone anomaly from day −1 to day 0. The anticyclonic weakening is most apparent in the Perth case (not shown).

The increase in cyclonic intensity when the wave is approaching the continent is partially a response to enhanced local baroclinicity due to the interaction of the southerly maritime flow with continental air and also with tropical–extratropical interaction features typically present in the Australian region (see section 4e). The latent heat release associated with the precipitation that can occur as a result of these interactions also helps to explain the rapid pressure decrease of cyclones as they approach the continent. Some of the systems intensify with such speed that they become “bombs”: cyclones exhibiting very rapid deepening and associated intensification (Sanders and Gyakum 1980; Lim and Simmonds 2002).

b. Upper-level pattern

Figure 7 presents the annual 300-hPa geopotential height anomaly plots for day −5 to day 0 for Melbourne CEs. For both Melbourne and Perth (Perth not shown), the anomalies over the CE-affected area are by far the largest for the hemisphere. Areas of 95% significance are shown by stippling for day 0.

A wave train over the southern Indian Ocean is present from about day −5, initializing over 120° from the CE-affected area. This train propagates slowly eastward, amplifying by event day with the large negative anomaly centered over the affected region. A two-ridge one-trough pattern is apparent for both cities also, with one positive anomaly center located to the west of the affected region, in the vicinity of the maximum MSLP anomaly, and a weaker ridge to the east. This pattern was consistently observed for each seasonal analysis, with autumn displaying the most organized pattern.

While the waves observed were similar in each city, the same wave patterns were not found to be responsible for CEs on the east and west side of Australia. The wave train patterns after CE occurrence (not shown) were found to decay around two days after inducing a CE. Wave trains associated with Perth CEs took about four days to reach the Melbourne region, by which time they were significantly weakened. This means that a CE in Perth is not normally followed by a CE in Melbourne. A similar pattern was observed for Melbourne CE-associated wave patterns, with the train greatly decaying before it reached the west coast of New Zealand. While a slight positive correlation between the annual frequencies of Melbourne and Perth CEs was found, it was not significant and no synoptic relationship was identified.

Comparing the geopotential patterns to the MSLP anomaly patterns, some conclusions can be drawn regarding the inclination of the vertical axis of the wave train. Both Melbourne and Perth CE-associated wave patterns exhibit a westward-tilted vertical axis. Toward event day, however, the positive pressure anomaly component of the Perth CE couplet alters to present a southward tilt, while the negative anomaly component remains westerly orientated in line with the climatology in that region (Lim and Simmonds 2007).

c. Jet stream involvement

The annual 300-hPa vector wind composites for day −2 to day 0 for Melbourne shown in Fig. 8 present a large wave amplification pattern in the subpolar jet over the CE-affected regions on day 0. Only two days before CE occurrence are shown, as a clear jet merging is only identifiable over this period. This equatorward “bulging” of the jet stream results in an increased strength of the subtropical jet to the north of the CE-affected area. The merging of the two jets over the CE-affected region is seen for both cities, although an increase in wind strength at the wave peak is seen for Perth as compared with Melbourne, as along with less clearly defined jet streams (Perth not shown). The wave amplification feature was present for all seasonal analysis. Because of seasonally varying latitudinal temperature gradients, however, the winter months displayed a much stronger subtropical jet than the summer months.

The jet merging discussed above is similar to that identified by Vera and Vigliarolo (2000), who found interaction with synoptic-scale subpolar and small-scale subtropical wave patterns to be associated with CEs that reached tropical regions in South America. The location of the strong subpolar jet stream over the Indian Ocean also supports the earlier hypothesis of the synoptic system speed being influenced by the jet.

These results are consistent with those where CE occurrence was found to be due to wave train amplification and large orographic forcing (Colle and Mass 1995; Vera and Vigliarolo 2000). The absence of a large mountain range in the Australian region indicates that a synoptic wave train is capable of causing an extreme CE on its own. While orography and other factors do influence the intensity of CEs, the wave pattern itself is the primary energy driver.

d. Role of sea surface temperature

Sea surface temperatures can have an influence on the strength and development of synoptic systems. A colder-than-average SST, in general, inhibits convection and thus promotes anticyclone development or intensification. Similarly, a warmer-than-average SST increases the instability in the lower atmosphere, leading to an intensification of cyclonic systems (Curry and Webster 1999). To explore the association of SSTs with CE-associated systems the SST anomalies for the week before event day were examined. These anomalies at each grid point were calculated by comparing SST data to a quadratic regression curve fitted to the data for a 90-day period centered over each event day, and then taking the average for the seven days before CE occurrence. This method ensured that global SST warming trends did not influence the results, nor did the time of year in which the CE occurred.

Figure 9 shows the associated SST anomalies for CEs during Melbourne summer (Fig. 9a) and Perth winter (Fig. 9b) over the relevant areas. The regions additionally marked with a cross are significant at the 90% level. For both cities, a wide, consistent region of negative SST anomalies is present over a similar area to the areas of high anticyclone density shown in Figs. 2 and 5. Areas of slightly positive SST anomalies are also apparent over the region occupied by the cyclones on and around event day, even in the subtropical regions frequented by Perth CE-associated systems. The anomalies are modest (less than 0.2°C) but are statistically significant over broad regions, not only in the seasonal plots but also annually. The presence of the negative anomalies may partly explain how the anticyclone systems are maintained over such a long distance, via reinforcing the subsidence for wave train intensification over the Indian Ocean. The positive anomalies can also assist in intensifying the cyclones as they approach CE interaction.

The summer and winter plots were chosen to show the presence of the negative anomaly in both the cold and warmer months. The other seasonal plots (not shown) also presented similar patterns with the exception of Melbourne winter and Perth autumn, which displayed more positive anomalies over the Indian Ocean. It is notable that Perth autumn events are not associated with negative anomalies in the Indian Ocean. Autumn has been associated with significant changes in precipitation and frontal system propagation affecting Western Australia (Hope et al. 2006) and there has also been a decrease in CE intensity in that season (to be discussed in section 5).

The location of the negative SST anomalies correspond well to the area of positive (anticyclonic) pressure anomalies identified five days before CE occurrence (see Fig. 6). In fact, the SST anomalies appear to coincide with the organization of the pressure wave pattern, which then continues to develop and amplify in conjunction with an upper-level synoptic wave. In the case of Melbourne, the major SST anomaly is located to the southwest of the anticyclonic anomaly on day −5. For Perth, the SST anomaly is slightly to the northwest of the anticyclone at day −5. A sharp increase in intensity is observed in this system, and also in the Perth CE pressure wave pattern from day −5 to day −4, when the anticyclonic anomaly passes over the region of slightly colder-than-average SST.

These results suggest a significant relationship between SST values and event-associated systems. These findings offer a new factor for consideration when examining the dynamics of CE occurrence, and also indicate the influence of climate and oceanic changes on CEs. This becomes increasingly important as SSTs are predicted to warm differentially over various regions in future years (Trenberth et al. 2007).

e. Extratropical transition and other association links

Figure 2c includes a solo tropical cyclone (or hurricane) track in the Melbourne track cloud. This track is that of Tropical Cyclone (TC) Nancy, which was associated with the coldest summer CE for the study period in February 1990. Examining this CE revealed that the anticyclone that made up the couplet with TC Nancy is that of the anomalously long track from South America, which is clearly visible within the anticyclone track “cloud” in Fig. 2a. Additionally, the movement of the system northward over South America is analogous to the path taken by systems associated with CEs in Argentina and Brazil (Pezza and Ambrizzi 2005), suggesting that a teleconnection between South America and Australia through wave amplification mechanisms cannot be discarded as a feature of Australian CEs. This is aligned with, and complementary to, earlier findings suggesting that gravity wave dispersion triggered by tropical convection in the western Pacific near Australia could also intensify cold events over South America (Marengo et al. 2002).

These insights were not identified in previous studies of this event (Simmonds and Richter 2000) and represent an interesting case of planetary-scale wave interaction. In particular, the presence of TC Nancy moving toward higher latitudes illustrates the importance and influence of extratropical transition in the Australian region (Dare and Davidson 2004; Pezza and Simmonds 2008). Extratropical transition (ET) occurs when a tropical cyclone moves toward higher, nontropical latitudes and undergoes thermodynamic and dynamic transformation toward a cold core. ETs and associated severe weather have been found to cause a lot of damage in many areas on both the east and west side of Australia (Foley and Hanstrum 1994). The identification of the interaction of ETs with CEs increases the importance of developing an accurate climatology of these systems in order to enhance CE prediction skills.

5. Trends in cold events

Figure 10 shows the frequency analysis of the 80 CEs identified from 1958 to 2006 for Melbourne (Fig. 10a) and the 83 identified for Perth (Fig. 10b). Both Melbourne and Perth plots exhibit significant (95%) negative trends over the period. Melbourne winter events in particular exhibited a stronger negative trend, as they were not identified after 1986.

Figure 11 gives, by seasons, the deviations of the maximum temperature on each event day from the monthly mean maximum temperature. These deviations are a good indication of relative event intensity. A small decrease in event intensity is present for autumn events in both cities, with the Perth autumn trend being significant. This is interesting to observe in light of the fact, as observed above, that autumn precipitation in Perth has also been decreasing (Hope et al. 2006). No other negative intensity trend is observed, however, and in fact weak positive trends are apparent for summer and winter in Melbourne and spring in Perth.

The negative frequency trends could be influenced by several factors. One possibility is the urban heat island (UHI) effect. The urban development of both cities from 1958 to 2006 may have produced an increasing UHI effect over this time (e.g., Morris and Simmonds 2000 for Melbourne), resulting in an artificial increase in environmental temperature, which, in the mean, made cold events warmer and thus harder to identify using the definition employed. However, the negative trend is present for both Melbourne and Perth, even though the Perth temperature data used were obtained from the weather station at Perth airport. This station is 9.8 km from the Perth metropolitan station and has been used in previous research as a source of rural data to examine the Perth UHI (Camilloni and Barros 1997). This suggests that an increasing UHI effect is not a major cause of the decrease in the number of CEs identified.

Another possible cause is an increase in climatological average temperatures. An examination of the seasonal average maximum temperatures at the two stations over the research period (not shown) shows a significant increase in average temperature for all seasons apart from summer. However, if the reason for the negative frequency trends was solely due to a linear warming of the background average temperature, one would expect to see a decrease in CE intensity over the study period. Figure 11 shows that this is not observed, suggesting that there are nonlinear processes behind the differential variability between CE frequency and intensity.

The influence of SST anomalies on CEs as discussed earlier means that SST changes may play a role in event frequency. The Indian Ocean has experienced a significant increase in SST over the study period that may be adversely impacting on the longevity of CE-associated anticyclones (Trenberth et al. 2007). The strength of this relationship is worthy of further research, as SSTs are predicted to increase in the future.

Another large-scale influence on CE occurrence could be variations in zonality over the Southern Hemisphere. Research in SH blocking indicates that there has been an increase in zonal flow and a decrease in 500-hPa geopotential wave amplitudes associated with the strengthening of the southern annular mode (Wiedenmann et al. 2002). Although this is not the scope of the present communication, a preliminary analysis suggests that no obvious relationship can be found between seasonal values of SAM and CE occurrence defined according to our criteria. The potential impacts, however, may happen indirectly as the zonality may influence on the ability for deep CE-associated troughs to develop. Further research into this relationship is needed.

a. Cold event index

Natural climate variability has a strong potential to modulate CE behavior in Australia. To test this hypothesis, the strength of the MSLP gradient was calculated over Melbourne and Perth for each day within the study period. The maximum and minimum daily MSLP anomalies were calculated within the dashed boxes shown in Fig. 1, as these boxes were found to contain the majority of MSLP anomalous maxima and minima during a CE. The difference between the two anomaly values was found for each day and termed the cold event index (CEI). The greater the CEI value, the stronger the pressure couplet.

Figure 12 shows the annual running averages of the CEI and maximum daily temperature anomalies for Melbourne (Fig. 12a) and Perth (Fig. 12b). No decreasing trend in CEI values is seen over the study period, although low-frequency variations are apparent. The CEI is an independent indicator of the circulation and is in agreement with the previous discussions that no apparent trends were observed in annual CE intensity over the study period. As expected there is an overall negative relationship between the two variables (R = −0.34), although examining individual day values yielded only a moderate correlation.

It appears that the wide range of variability of system locations and formations does not allow for this technique to be applied as an event identifier on its own. In other words, while a high CEI value indicates an environment conducive of CE occurrence, it does not necessarily occur. The technique, however, offers a new insight on the long-term dynamic indicators associated with large-scale conditions leading to CE occurrence in Australia.

6. Concluding remarks

This work has added to the knowledge of cold events over southern Australia, as well as the Southern Hemisphere. The fundamental mechanism of a CE was identified as a surface pressure couplet that sustains the cold advection through equatorward geostrophic flow. The synoptic systems comprising this couplet were found to be anomalously deep. On average the systems were also anomalously long in duration and track length. The anticyclones in particular were found to cover large distances before and after being associated with a CE. This was linked to small but significant negative SST anomalies over much of the genesis and propagation regions, which may have assisted the systems in their extended life spans via intensifying the subsidence over key areas of the Indian Ocean. These findings complement similar results from Pezza and Ambrizzi (2005) and others on South American CE-associated systems.

Wave trains amplifying over the CE-affected region on event day were identified as associated with the surface pressure couplet. The synoptic-scale waves associated with these Australian CEs can be identified several days before CE occurrence and represent clear planetary interaction. These wave patterns are similar to those identified in previous studies of CE occurrence. However, in those studies CEs were attributed to a combination of wave amplification and orographic forcing. As this forcing is present to a much lesser extent over the Australian region, the results suggest that wave train amplifications on their own are enough to induce CE occurrence. Another important feature observed for the Australian region was the occasional involvement of TCs (i.e., hurricanes). The identification of an interaction between tropical cyclones that undergo extratropical transition and CEs opens future venues of explorations, many of which would benefit from a climatology of cyclone transition in the SH. This area is recommended for future research, as very little is presently known about these important types of interaction. Other CE-associated cyclones may develop as rapidly intensifying “bombs,” another area in which future research is suggested. With additional knowledge regarding these systems a greater understanding of the links between them and cold event occurrence could be obtained, shedding more light on how CEs occur and how they are likely to change in frequency or intensity in the future.

The frequency of CEs indicates a slight negative trend in event occurrence in both Melbourne and Perth. This may be associated with a number of factors, including the inability of our CE definition to identify more recent events because of an increase in environmental temperatures. However, the relative intensity of the CEs identified was not found to be changing, as strong CEs are still identified for recent years. This is consistent with a new dynamic index developed to objectively quantify large-scale conditions favorable for CE occurrence, which has also not shown any trends. These results highlight the difficulty in developing a CE definition that is completely unbiased, and that at least from the viewpoint of our definition it is nonlinear climate variability processes that seem to be the dominant factor in CE behavior in southern Australia.

Acknowledgments

Parts of this research were made possible by funding from the Australian Research Council and the Antarctic Science Advisory Committee. The authors are indebted to Kevin Keay for his expertise and assistance in this work and to Luke Garde for his comments and advice. The authors are also very grateful for the insightful comments from three anonymous reviewers that greatly improved the work.

REFERENCES

  • Adams, C. R., 1997: Impacts of temperature extremes. Proc., Workshop on the Social and Economic Impacts of Weather, Boulder, CO, U.S. Weather Research Program, 11–16.

    • Search Google Scholar
    • Export Citation
  • Alexander, L. V., P. Hope, D. Collins, B. Trewin, A. Lynch, and N. Nicholls, 2007: Trends in Australia’s climate means and extremes: A global context. Aust. Meteor. Mag., 56 , 118.

    • Search Google Scholar
    • Export Citation
  • Bureau of Meteorology, cited 2003: Climate activities in Australia. [Available online at http://www.bom.gov.au/inside/eiab/reports/caa03/chapter4/disease_human_health.shtml].

    • Search Google Scholar
    • Export Citation
  • Camilloni, I., and V. Barros, 1997: On the urban heat island effect dependence on temperature trends. Climatic Change, 37 , 665681.

  • Colle, B. A., and C. E. Mass, 1995: The structure and evolution of cold surges east of the Rocky Mountains. Mon. Wea. Rev., 123 , 25772610.

    • Search Google Scholar
    • Export Citation
  • Colucci, S. J., D. P. Baumhefner, and C. E. Konrad, 1999: Numerical prediction of a cold air outbreak: A case study with ensemble forecasts. Mon. Wea. Rev., 127 , 15381550.

    • Search Google Scholar
    • Export Citation
  • Curry, J. A., and P. J. Webster, 1999: Thermodynamics of Atmospheres and Oceans. International Geophysics Series, Vol. 65, Academic Press, 471 pp.

    • Search Google Scholar
    • Export Citation
  • Dare, R. A., and N. E. Davidson, 2004: Characteristics of tropical cyclones in the Australian region. Mon. Wea. Rev., 132 , 30493065.

  • Elliott, P. R., 1989: The Tasmanian snowstorm 24 and 25 July 1986. Meteorology Note 187, Bureau of Meteorology Australia, 34 pp.

  • Emergency Management Australia, cited 2008: EMA disaster database—Crop damage: Goulburn Valley, VIC. [Available online at http://www.ema.gov.au/ema/emadisasters.nsf/83edbd0553620d8cca256d09001fc8fd/32e4b08a3ec179dcca2572cf0023765a?OpenDocument].

    • Search Google Scholar
    • Export Citation
  • Foley, G. R., and B. N. Hanstrum, 1994: The capture of tropical cyclones by cold fronts off the west coast of Australia. Wea. Forecasting, 9 , 577592.

    • Search Google Scholar
    • Export Citation
  • Fortune, M. A., and V. E. Kousky, 1983: Two severe freezes in Brazil: Precursors and synoptic evolution. Mon. Wea. Rev., 111 , 181196.

    • Search Google Scholar
    • Export Citation
  • Garreaud, R. D., 2000: Cold air incursions over subtropical South America: Mean structure and dynamics. Mon. Wea. Rev., 128 , 25442559.

    • Search Google Scholar
    • Export Citation
  • Garreaud, R. D., 2001: Subtropical cold surges: Regional aspects and global distribution. Int. J. Climatol., 21 , 11811197.

  • Garreaud, R. D., and J. M. Wallace, 1998: Summertime incursions of midlatitude air into subtropical and tropical South America. Mon. Wea. Rev., 126 , 27132733.

    • Search Google Scholar
    • Export Citation
  • Hartjenstein, G., and R. Bleck, 1991: Factors affecting cold-air outbreaks east of the Rocky Mountains. Mon. Wea. Rev., 119 , 22802292.

    • Search Google Scholar
    • Export Citation
  • Hope, P. K., W. Drosdowsky, and N. Nicholls, 2006: Shifts in the synoptic systems influencing southwest Western Australia. Climate Dyn., 26 , 751764.

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

  • Kistler, R., and Coauthors, 2001: The NCEP–NCAR 50-Year Reanalysis: Monthly means CD-ROM and documentation. Bull. Amer. Meteor. Soc., 82 , 247267.

    • Search Google Scholar
    • Export Citation
  • Konrad, C. E., 1996: Relationships between the intensity of cold-air outbreaks and the evolution of synoptic and planetary scale features over North America. Mon. Wea. Rev., 124 , 10671083.

    • Search Google Scholar
    • Export Citation
  • Konrad, C. E., 1998: Persistent planetary scale circulation patterns and their relationship with cold air outbreak activity over the eastern United States. Int. J. Climatol., 18 , 12091221.

    • Search Google Scholar
    • Export Citation
  • Krishnamurti, T. N., M. Tewari, and D. R. Chakraborty, 1999: Downstream amplification: A possible precursor to major freeze events over southeastern Brazil. Wea. Forecasting, 14 , 242270.

    • 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
  • Lim, E-P., and I. Simmonds, 2007: Southern Hemisphere winter extratropical cyclone characteristics and vertical organization observed with the ERA-40 data in 1979–2001. J. Climate, 20 , 26752690.

    • Search Google Scholar
    • Export Citation
  • Marengo, J. A., and J. C. Rogers, 2001: Polar air outbreaks in the Americas: Assessments and impacts during modern and past climates. Interhemispheric Climate Linkages, V. Markgraf, Ed., Academic Press, 17–51.

    • Search Google Scholar
    • Export Citation
  • Marengo, J. A., A. Cornejo, P. Satyamurty, and C. Nobre, 1997: Cold surges in tropical and extratropical South America: The strong event in June 1994. Mon. Wea. Rev., 125 , 27592786.

    • Search Google Scholar
    • Export Citation
  • Marengo, J. A., T. Ambrizzi, G. Kiladis, and B. Liebmann, 2002: Upper-air wave trains over the Pacific Ocean and wintertime cold surges in tropical-subtropical South America leading the freezes in southern and southeastern Brazil. Theor. Appl. Climatol., 73 , 223242.

    • Search Google Scholar
    • Export Citation
  • Morley, K. T., 1957: Outbreaks of Antarctic air over southeast Australia. Aust. Meteor. Mag., 19 , 6263.

  • Morris, C. J. G., and I. Simmonds, 2000: Associations between varying magnitudes of the urban heat island and the synoptic climatology in Melbourne, Australia. Int. J. Climatol., 20 , 19311954.

    • Search Google Scholar
    • Export Citation
  • Müller, G. V., and G. J. Berri, 2007: Atmospheric circulation associated with persistent generalized frosts in central-southern South America. Mon. Wea. Rev., 135 , 12681289.

    • Search Google Scholar
    • Export Citation
  • Murray, R. J., and I. Simmonds, 1991a: A numerical scheme for tracking cyclone centres from digital data. Part 1: 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 2: Application to January and July general circulation model simulation. Aust. Meteor. Mag., 39 , 167180.

    • Search Google Scholar
    • Export Citation
  • Perrin, G., and I. Simmonds, 1995: The origin and characteristics of cold air outbreaks over Melbourne. Aust. Meteor. Mag., 44 , 4159.

    • Search Google Scholar
    • Export Citation
  • Pezza, A. B., and T. Ambrizzi, 2005: Dynamical conditions and synoptic tracks associated with different types of cold surge over tropical South America. Int. J. Climatol., 25 , 215241.

    • Search Google Scholar
    • Export Citation
  • Pezza, A. B., and I. Simmonds, 2008: Large scale factors in tropical and extratropical cyclone transition and extreme weather events. Ann. N. Y. Acad. Sci., 1146 , 189211.

    • Search Google Scholar
    • Export Citation
  • Pezza, A. B., I. Simmonds, and J. A. Renwick, 2007: Southern Hemisphere cyclones and anticyclones: Recent trends and links with decadal variability in the Pacific Ocean. Int. J. Climatol., 21 , 14031419.

    • Search Google Scholar
    • Export Citation
  • Rogers, J. C., and R. V. Rohli, 1991: Florida citrus freezes and polar anticylones in the Great Plains. J. Climate, 4 , 11031113.

  • Salinger, M. J., and G. M. Griffiths, 2001: Trends in New Zealand daily temperature and rainfall extremes. Int. J. Climatol., 21 , 14371452.

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

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

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., and T. Richter, 2000: Synoptic comparison of cold events in winter and summer in Melbourne and Perth. Theor. Appl. Climatol., 67 , 1932.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., and H. A. Rashid, 2001: An investigation of a dramatic cold outbreak over southeast Australia. Aust. Meteor. Mag., 50 , 249261.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., K. Keay, and E-P. Lim, 2003: Synoptic activity in the seas around Antarctica. Mon. Wea. Rev., 131 , 272288.

  • Trenberth, K. E., and Coauthors, 2007: Observations: Surface and atmospheric climate change. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 299–316.

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

  • Vavrus, S., J. E. Walsh, W. L. Chapman, and D. Portis, 2006: The behaviour of extreme cold air outbreaks under greenhouse warming. Int. J. Climatol., 26 , 11331147.

    • Search Google Scholar
    • Export Citation
  • Vera, C. S., and P. K. Vigliarolo, 2000: A diagnostic study of cold-air outbreaks over South America. Mon. Wea. Rev., 128 , 324.

  • Vincent, L. A., and Coauthors, 2005: Observed trends in indices of daily temperature extremes in South America 1960–2000. J. Climate, 18 , 50115023.

    • Search Google Scholar
    • Export Citation
  • Wiedenmann, J. M., A. R. Lupo, I. I. Mokhov, and E. A. Tikhonova, 2002: The climatology of blocking anticyclones for the Northern and Southern Hemispheres: Block intensity as a diagnostic. J. Climate, 15 , 34593473.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Annual composite of MSLP (shaded) and MSLP anomalies (contoured) over (a) Melbourne and (b) Perth outbreak day, averaged from CEs occurring from 1958 to 2006. Areas of 95% significance are stippled. The solid boxes identify the 5° × 5° reference areas for the statistics discussed in the text, while the dashed boxes indicate the regions containing the majority of pressure couplets. Red asterisk indicates the approximate location of Melbourne in (a) and Perth in (b).

Citation: Journal of Climate 22, 24; 10.1175/2009JCLI2997.1

Fig. 2.
Fig. 2.

Anticyclone and cyclone tracks associated with cold events in Melbourne from 1958 to 2001 split into pre- and postevent components. The dots on the tracks give the location of the system center every 6 hours. (a) Dark blue shows the pre-event anticyclones and (b) light blue gives the respective postevent movement. (c) Red shows the pre-event cyclones and (d) orange the postevent cyclone progression. The average path is shown as a gray line. The green and yellow circles mark the average location of the anticyclonic and cyclonic pressure systems respectively on event day, and the squares give the median genesis point. The symbols at the bottom of each plot refer to the average speed, track duration, and distance traveled of CE-associated systems (colored) compared to the climatology of non-CE systems (black).

Citation: Journal of Climate 22, 24; 10.1175/2009JCLI2997.1

Fig. 3.
Fig. 3.

System density of anticyclones and cyclones associated with cold events in (a),(b) Melbourne and (c),(d) Perth from 1958 to 2001. Colored contours show the mean number of systems found in a 103 (° latitude)−2 area over the study period. Note that the scale size is different for each system type.

Citation: Journal of Climate 22, 24; 10.1175/2009JCLI2997.1

Fig. 4.
Fig. 4.

The average value and standard deviation of (a) TD, (b) DT, and (c) SP of pre- and post-CE track components for CE-associated systems and climatological, non-CE systems.

Citation: Journal of Climate 22, 24; 10.1175/2009JCLI2997.1

Fig. 5.
Fig. 5.

DP anomalies (hPa) of anticyclonic and cyclonic systems associated with CEs in (a),(b) Melbourne and (c),(d) Perth. The smoothing used is a Cressman-type with a 5.4° radius and a 0.09 pinching factor.

Citation: Journal of Climate 22, 24; 10.1175/2009JCLI2997.1

Fig. 6.
Fig. 6.

Composites of daily MSLP anomalies (contoured, hPa) with MSLP (shaded, hPa) for day −5 to day 0 for CEs in Melbourne. For the day 0 plot, areas of 95% significance are stippled.

Citation: Journal of Climate 22, 24; 10.1175/2009JCLI2997.1

Fig. 7.
Fig. 7.

Composites of daily geopotential height anomalies at 300 hPa (m) for day −5 to day 0 for CEs in Melbourne. For the day 0 plot, areas of 95% significance are stippled.

Citation: Journal of Climate 22, 24; 10.1175/2009JCLI2997.1

Fig. 8.
Fig. 8.

Composites of 300-hPa vector wind from day −2 to day 0 for CEs in Melbourne. Shading shows wind speed (m s−1). Wind direction is shown by streamlines.

Citation: Journal of Climate 22, 24; 10.1175/2009JCLI2997.1

Fig. 9.
Fig. 9.

Average surface skin temperature anomalies (°C) over the ocean for the seven days before a CE in (a) Melbourne summer and (b) Perth winter. Areas marked with a cross indicate regions of 90% significance.

Citation: Journal of Climate 22, 24; 10.1175/2009JCLI2997.1

Fig. 10.
Fig. 10.

Seasonal frequency analysis of CEs in (a) Melbourne and (b) Perth from 1958 to 2006 with annual linear trend overlaid.

Citation: Journal of Climate 22, 24; 10.1175/2009JCLI2997.1

Fig. 11.
Fig. 11.

Magnitude of difference between seasonal average maximum temperatures and event day maximum temperatures for each season in (a) Melbourne and (b) Perth from 1958 to 2006. For seasons exhibiting a significant or near-significant trend, the R2 value is given.

Citation: Journal of Climate 22, 24; 10.1175/2009JCLI2997.1

Fig. 12.
Fig. 12.

Annual running averages of the CEI and maximum daily temperature anomalies for (a) Melbourne and (b) Perth from 1958 to 2006. See text for further details.

Citation: Journal of Climate 22, 24; 10.1175/2009JCLI2997.1

Table 1.

The number of CEs per calendar month identified from 1958 to 2006 for Melbourne and Perth. The average maximum temperature (and anomalies) reached on the day of the CE are also shown.

Table 1.
Save