Australia: The Land Where Time Began

A biography of the Australian continent 

Stratospheric Polar Vortex Weakenings Induced Hot and Dry Extremes in Australia

There can be large impacts on human health, energy and water supplies, agriculture and wildfires that can result from extremes of hot and dry conditions in warm seasons. It has been known that hot and dry extremes are associated with the occurrence of El Niño and other variations of tropospheric circulation. In this study Lim et al. identified an additional driver: variability of the stratospheric Antarctic Polar Vortex. Based on statistical analyses using observational data that covered the past 40 years, they show that weakenings and warmings of the stratospheric polar vortex, which occurs episodically during the austral spring, increases the chances substantially of hot and dry extremes, as well as of associated weather across subtropical eastern Australia that is fire-conductive, from the austral spring to early summer. The downwards coupling of the weakened polar vortex to tropospheric levels, where it is linked to the low-index polarity of the Southern Annular Mode, a shift towards the equator of the mid-latitude jet steam and subsidence and warming in the subtropics, results in the promotion of these extremes of the Australian climate. The enhanced likelihood of hot and dry extremes and risks of wildfire across eastern Australia in the early summer may be predictable a season in advance during years of weakenings of the vortex, as a result of the long timescales of the variations of the polar vortex.

It is crucially important in Australia to understand, predict and anticipate extreme high temperatures and low rainfall in the warm seasons because of the risks and impacts on human health, agriculture, wildfires, utilities, infrastructure and water management (Managing Climate Variability R&D Program. Climate Kelpie). In particular, the impact of the reduced productivity of dairy and meat under conditions that are hot and dry can be felt beyond Australia as it is one of the world’s major exporters of these products (Australian Government Department of Agriculture Meat, wool and dairy. Australian Government). The warm phase of the El Niño-Southern Oscillation (ENSO) (Bjerknes, 2019), the positive phase of the Indian Ocean Dipole (IOD) (Saji et al., 1999) and the negative index polarity of the Southern Annular Mode (SAM) (Thompson & Wallace, 2000) are important drivers of extreme hot and dry conditions over various regions of Australia, mainly in spring, though summer as well (Hendon, Thompson & Wheeler, 2007; Min, Cai & Whetton, 2013; Perkins, Argüeso and White, 2015). In this paper Lim et al. highlight that a spring stratospheric polar vortex over Antarctica that is weaker than normal is a key driver and an important source of predictability of climatic conditions that are hot and dry over subtropical eastern Australia during its warm seasons. It is shown that in spring that anomalous weakening of the Southern Hemisphere polar stratospheric vortex subsequently promotes that negative phase of the Southern Annular Mode, which then promotes the high daytime temperature and low rainfall extremes, as well as the associated dangerous atmospheric conditions that are conducive to wildfires in Australia from the middle of spring to early summer.

Occurring in both hemispheres, the stratospheric polar vortex is characterised by maximum westerly winds in the upper stratosphere-lower mesosphere that circle the hemisphere. It is at its strongest in the winter half of the year disappearing during the summer, making its seasonal progression from mid-latitudes towards the poles from early winter to spring (Hirota, Hirooka & Shiotani, 1983; Kodera & Kuroda, 2002; Waugh, Sobel & Polavani, 2017). Also involved in this seasonal shift is a downwards progression of the maximum westerlies from the upper stratosphere in winter to the lower troposphere in spring, which results from vigorous interactions with waves on a planetary scale that are travelling upwards from the troposphere which act to decelerate the westerly vortex. In the Northern Hemisphere the polar vortex is mainly perturbed during boreal winter to early spring (Waugh, Sobel & Polvani, 2017; McIntyre, 1982), and the perturbation of the Southern Hemisphere vortex occurs during the austral winter to spring (Randel, 1988; Hio & Yoden, 2005). In the stratospheric polar vortex variability is produced by variability of the planetary scale waves which are propagating upwards which can couple downwards to the troposphere and promote low-frequency variations in circulation of the troposphere and the temperature, thereby serving as a source of extended-range predictability of surface weather and climate (Baldwin et al., 2003; Charltob et al., 2003; Byrne & Shepherd, 2018; Lim, Hendon & Thompson; 2018). E.g., downward coupling from the stratosphere to the troposphere in association with sudden stratospheric warmings in the Northern Hemisphere, which commonly occurs in the boreal winter and early spring, often results in sustained impacts on surface climate by promoting the negative polarity of the North Atlantic Oscillation/Northern Annular Mode (Baldwin et al., 2015; Sigmond et al., 2013; Baldwin & Dunkerton, 1999).

Variations of the stratospheric polar vortex occur episodically in the Southern Hemisphere in the austral spring to early summer (Shiotani & Hirota, 1985; Kuroda & Kodera, 1998), though they generally are not as spectacular as sudden stratospheric warming events in the Northern Hemisphere. During spring, anomalous Southern Hemisphere weakening or intensification of the polar vortex, which can be viewed as an earlier or later shift in the seasonal evolution of the vortex, respectively (Byrne & Shepherd, 2018; Shiotani, Shimods & Hirota, 1993), can lead to the sustained occurrence of the negative or positive polarity of the Southern Annular Mode (SAM denotes tropospheric SAM in this study, unless otherwise stated), respectively, during spring-summer (Baldwin et al., 2003; Byrne & Shepherd, 2018; Lim, Hendon & Thompson, 2018; Seviour et al., 2014; Thompson, Baldwin & Solomon, 2005). The Southern Annular Mode is well known to be a driver in surface climates in the Southern Hemisphere, including precipitation changes and surface temperatures (Hendon, Thompson & Wheeler, 2007; Thompson & Solomon, 2002; Gillett, Kell & Jones; Bandon et al., 2014), by the continental shifts in the extratropical storm track and edge of the Hadley circulation (Kang et al., 2011; Hendon, Lim & Nguyen). The Southern Annular Mode has impacts on the climate of Australia which are particularly pronounced, causing changes in maximum and minimum temperatures and rainfall across much of the southern and eastern parts of the continent, especially during spring and summer (Hendon, Thompson & Wheeler, 2007; Min, Cai & Whetton, 2013; Gillett, Kell & Jones, 2006; Lim & Hendon, 2015).

Identifying years when the Southern Hemisphere polar vortex weakens

This study builds on Lim et al. (Lim, Hendon & Thompson, 2018), who identified objectively years when the Southern Hemisphere polar vortex weakened significantly and broke down in spring earlier than normal, which consequently weakened the westerly winds propagating downwards from the stratosphere to the surface during spring to early summer. The opposite can also happen, with the breakdown of the vortex driving anomalously strong westerly winds. It was demonstrated by Lim et al. that there is a strong connection of this stratosphere-troposphere (S-T) coupling to the subsequent variation of the Southern Annular Mode during October-January, which drives the variation of surface climate throughout the Southern Hemisphere (Lim, Hendon & Thompson, 2018).

In this study Lim et al. selected 9 years when the Antarctic polar vortex was most significantly weaker than normal in spring during the period 1979-2016, as judged by the amplitude of the stratosphere-troposphere coupled mode index (18) ≥0.8 standard deviations (σ). The 9 years 1979, 1988, 2000, 2002, 2004, 2005, 2013 and 2016). They then examined the variation of the October-January mean maximum daily temperature (T_max) in Australia, daily minimum (T_min), potential for rainfall and fire as measured by the McArthur Forest Fire Danger Index (FFDI) (dowdy, 2018). They also monitored variations in the Southern Annular Mode, mean sea level pressure (MSLP), geopotential height and vertical velocity at the 500 hPa level (GPH500 and ω500, respectively) and total cloud cover. In the 9 years of significant weakening of the vortex they compared the means of these variables to the control group of the other 29 years when the stratosphere-troposphere coupled mode index <0.8σ). They included the years when the Southern Hemisphere vortex strengthened in the control group as their impact on the climate of Australia differs little from the normal conditions. In order to focus on the impact of the variation of the Southern Hemisphere polar vortex that is independent of ENSO from the variables examined in this study by the use of linear regression. They also detrended all the data for the period 1979-2016 in order to concentrate on the impacts of year-to-year variations of the polar vortex.

Impact of the weakening of the vortex on the Australian seasonal climate

The composite difference of the T_max, T_min and rainfall anomalies averages over October-January for the significant 9 years of weakening of the polar vortex (vortex weakening years) compared with the control group of all the other years (vortex non-weakening years). The strongest T_max anomalies during the years of vortex weakening occurred over subtropical eastern Australia – Queensland, New South Wales and northeastern South Australia (130-156^oE, 10-30^oS) where the 4 month mean anomaly is significantly higher, reaching up to 2^oC. Over most of eastern Australia coincident reduction of rainfall is found but especially over central Queensland. During the years of vortex weakening the significant increases of T_max and decreases of rainfall are statistically robust at the 5% level, assessed by a permutation resampling test (Wilks, 2006) (2-sided) using 4,000 random resamples. T_min appears, however, to be less sensitive, relatively, to the weakening of the polar vortex; therefore they excluded T_min in the subsequent analysis and discussion.

Significant weakening of the vortex especially affects the occurrences of extreme seasonal mean T_max and rainfall. They computed the ratio of the probability of occurrence of extreme events during the years of vortex weakening to that of the years when the vortex is not weakening using 2 definitions of an extreme event. They examined in the first case, the occurrence of a top-quintile hottest year, i.e., the top 20%, which equates to the 7 hottest events in the 38 year record, and a bottom-quintile driest year by use of the October-January mean T_max and rainfall, respectively. Consistent with the mean anomalies during the years of a weakening vortex, the probably that a year would be in the top quintile for T_max and the bottom quintile for rainfall increases by up to 4 to 8 times for large areas of Queensland and northern New South Wales.

As favourable conditions for wildfires are extreme high temperatures and low rainfall (Dowdy, 2018), the probability of dangerous fire weather occurring being in the top quintile (as indicated by the FFDI; Methods) is also increased significantly 4 to 8 times during the years of vortex weakening  over almost all of Queensland, northern New South Wales and northeastern South Australia. During the vortex weakening years the far west region of Western Australia is also vulnerable to increased danger to wildfire.

They also looked at the occurrence of individual months during October-January for having a T_max and rainfall in the top and bottom deciles (10%), respectively. There is an increase of greater than 4 times in the probability the occurrence of monthly mean T_max in the top decile and monthly mean rainfall being in the bottom decile over southern Queensland during the vortex weakening years. As a consequence of this the chance for a month during October to January having dangerous fire weather conditions increases substantially over southern Queensland and northeastern South Australia during the vortex weakening years. These frequency increases of extreme hot and dry events and associated danger of wildfire in October-January following the weakening and breakdown of the stratospheric polar vortex in spring are statistically significant at the 2.5% level, as assessed by a 1-sided permutation test.

Without first removing the trend and the linear relationship with ENSO, they have repeated these calculations. They found that the probability of occurrence of mean extreme high T_max, low rainfall and high risk of fire weather for October to January still increases by a factor of 4 in the same areas of eastern Australia during the vortex weakening years. During the vortex weakening years the probability of extreme T_max and the related fire danger still increases by more than 2 to 3 times. The role of the weakening of the stratospheric polar vortex in driving extreme hot, dry and fire-prone atmospheric conditions in subtropical eastern Australia during spring-summer is highlighted by these results.

How the Southern Hemisphere polar vortex is linked to Australian climate

During spring the weakening of the Antarctic Polar vortex promotes the negative polarity of the SAM, the anomalous hemispheric pattern of which is characterised by low pressure anomalies that are zonally symmetric in the mid-latitudes and high pressure anomalies in the high latitudes (Thompson & Wallace, 2000; Lin, Hendon & Thompson, 2018). During October – January when the polar vortex weakens and breaks down early, the SAM tends to be persistently negative. An important characteristic of the negative polarity of the SAM driven by weakening of the stratospheric vortex which appears to amplify the impact of the SAM on extreme climate conditions over Australia is this persistence. Through the warm season the negative polarity of the SAM that is not coupled to the weakening of the stratospheric polar vortex is less persistent, and appears to have less impact on extreme climate conditions than the negative SAM that is coupled to the weakening of the stratospheric polar vortex.

The negative polarity of the SAM is associated with surface pressure over Australia that is lower than normal and surface westerlies to the south of Australia that are stronger than normal (Hendon, Thompson & Wheeler, 2007). During October-January the pattern of surface pressure and wind differences in years of vortex weakening is consistent with this negative polarity SAM pattern. The difference in surface winds between the years of vortex weakening and non-weakening years is not statistically significant over subtropical eastern Australia.

Contrasting with the surface pressure that is lower than normal, mid-tropospheric geopotential height over Australia is significantly more positive during years when the vortex is weakening, and the downwards motion is increased, that is indicated by positive anomalies of ω500. There is also substantially reduced total cloud cover over eastern Australia during the vortex weakening years. The increased downwards motion and reduced cloud cover over Australia reflect a shift equatorwards of the downwards edge of the downwards branch of the Southern Hemisphere Hadley cell in the austral warm seasons, which has been shown previously to be a dynamical response to the negative polarity of the SAM (Kang et al., 2011; Hendon, Lim & Nguyen, 2014). The locations of reduced cloud cover and enhanced subsidence coincide with the regions of enhanced T_max and reduced rainfall over eastern Australia. These anomalies together support the notion that positive temperature anomalies in the eastern part of Australia are promoted primarily by adiabatic warming that result from enhanced subsidence and increased insolation that is associated with reduced cloud cover (35). As less soil moisture is available for evaporative cooling, reduced rainfall also contributes to higher temperature. The significant increases of the mid-tropospheric geopotential height and downwards motion imply that the negative surface pressure anomalies over Australia that are associated with the negative polarity of the SAM during the years of vortex weakening are likely to be thermally driven with a shallow vertical structure. According to Lim et al. warm and dry northwesterlies that are associated with the low pressure centre to the south of Australia could bring hot, dry air to the east and contribute to the extreme hot and dry climate in eastern Australia, as was observed in 2005 and 2016.

Prospect for skillful prediction of extreme climate in the Southern Hemisphere

Lim et al have highlighted the connection between anomalous weakening of the Southern Hemisphere stratospheric polar vortex and the significantly enhanced occurrences of extreme hot and dry conditions over eastern Australia during the warm season. The weakening of the stratospheric polar vortex promotes the negative polarity of the SAM, which induces the anomalous sinking of air and a clearer sky over eastern Australia and appears to serve as a pathway for the stratospheric polar vortex signal to affect subtropical eastern Australia climate conditions. It has been demonstrated by previous studies that predictability of the SAM during spring and summer arises due to its connection to ENSO (L’Heureux & Thompson, 2006; Zhou & Yu, 2004; Lim, Hendon & Rashid, 2013), though it is suggested by the results of the current study that the variability of SAM that is induced by stratospheric variations during October to January is independent of ENSO and has a greater impact on the climate of subtropical eastern Australia than does ENSO. It was reported by Lim et al. (Lim, Hendon & Thompson, 2018) that stratosphere-troposphere coupling during the months of spring and summer is often preconditioned by anomalies in upwards-propagating planetary wave activity and a meridional shift of the vortex as high as the stratosphere and as early as June. Lim et al. suggest that it might be anticipated, therefore, increased chance of extreme hot and dry conditions in late spring to early summer at least a season in advance when the Antarctic polar vortex shows an anomalous weakening signal as early as winter Byrne & Shepherd, 2018; Lim, Hendon & Thompson, 2018).

It was reported by previous work (Seviour et al., 2014) that in spring the variability of the Antarctic polar vortex and the variation of the associated tropospheric SAM can be skillfully predicted in a dynamical seasonal climate forecast system when initialised at the beginning of August. It is implied by this capability that the potential for long-lead predictive skill of the occurrence of an extreme Australian climate as well as other regions of the Southern Hemisphere that are affected by variation of the SAM, which includes Antarctic sea ice (Lim, Hendon & Thompson, 2018; Bandoro et al., 2014; Wang et al., 2019) in the spring and summer. This predictability that arises from variations of the polar stratosphere would be in addition to predictability that arises from ENSO, which is traditionally seen as the main source of long lead predictability for global climate. According to Lim et al. research efforts in the future will focus on exploring the predictability of extreme climate conditions that are provided by polar stratospheric variations using the Australian Bureau of Meteorology’s new subseasonal to seasonal climate forecast system, CCESS-S1 (Hudson et al., 2017), which well resolves the stratosphere but runs with prescribed climatological ozone. In the lower stratospheric polar vortex ozone varies with the strength of the vortex, and as incoming ultraviolet radiation is absorbed efficiently by ozone, ozone can act to anomalously warm or cool the vortex, thereby acting to extend the impact of the vortex variation of surface climate during summer (Bandoro et al., 2014; Shaw et al., 2011; Gillett et al., 2019). A key area for the development of improved climate prediction systems, especially in light of the expected recovery of the Antarctic ozone hole over coming decades (Solomon et al., 2016), is the implementation of prognostic ozone.

Sources & Further reading

Lim, E.-P., et al. (2019). "Australian hot and dry extremes induced by weakenings of the stratospheric polar vortex." Nature Geoscience 12(11): 896-901.