Sharp polar outbreak freezes Perth and Bunbury, dumps fresh snow on the Alps

A sharp polar/Antarctic outbreak affected southern Australia in early-to-mid August 2026, with Perth recording a notably cold morning.

Bureau of Meteorology daily observations for Perth (combining metro and airport sites) show a minimum of 2.5°C on Tuesday 11 August 2026, with a maximum of 16.1°C that day. This was well below typical early-August overnight lows for the city (commonly in the 4–9°C range). Some local reports noted it felt closer to 1°C in places due to wind chill or frost, with ice reported on cars in parts of the metro area and hills.

The cold snap followed a powerful Antarctic cold front that moved across southern Australia over the preceding weekend (around 8–10 August). It first impacted southwest Western Australia with rain, thunderstorms, small hail, strong/gusty winds and a large swell, then tracked east into South Australia, Victoria, Tasmania and southern New South Wales. Impacts included heavy rain (widespread 20–40 mm possible in the southeast), damaging winds, hail, thunderstorms and snow in alpine areas. An associated Antarctic low-pressure system helped draw the frigid air northward.

Perth and surrounding areas had already seen wet and wild conditions over the weekend (including hail and storms), with the clear, settled conditions under a following high-pressure system allowing temperatures to plummet overnight into Tuesday. Forecasts afterward pointed to a gradual warm-up later in the week under more settled, sunny weather.

Bunbury (southwest Western Australia) recorded a minimum of 0.8°C (33.4°F) on Tuesday 11 August 2026.

Official Bureau of Meteorology daily weather observations for the Bunbury site (Carey Park) list the overnight low as 0.8°C that morning (with a maximum of 16.5°C). This was the lowest temperature recorded at the site in the first half of August 2026 and stands out as an especially cold August morning for the location.

This occurred as the same polar/Antarctic air mass that brought Perth’s 2.5°C low settled over the southwest under clear skies after the cold front’s passage, allowing strong overnight radiative cooling.

The cold air then continued pushing eastward across southern Australia. It contributed to further alpine snowfall in the Victorian and New South Wales high country (including resorts in the Australian Alps). The broader system associated with the early-to-mid August polar outbreak delivered one of the more significant snowfalls of the 2026 season so far, with multiple rounds of snow reported or forecast through the period (helping improve what had been a relatively poor start to the ski season in some areas).

Overall, this fits the pattern of the sharp Antarctic outbreak that affected southern Australia around 8–13 August 2026, bringing cold mornings in the west followed by wintry conditions (including snow) farther east.

Australian Alps snowfall (primarily measured at Spencers Creek, 1830 m in the Snowy Mountains, NSW) — 2026 vs 2016 comparison (as of mid-August 2026).

Spencers Creek (Snowy Hydro records since 1954) is the standard long-term benchmark for natural snow depth in the Australian Alps. The long-term average peak depth is roughly 185–196 cm.

2026 Season (to date)

Very poor start: Effectively snowless into early July (near 0 cm at the start of July — only the second time in 72 years of records). Early July depths were among the worst on record.

Mid-July: Reached about 70–71 cm after some systems.

Early August (around 5 August): 87.1 cm — just above the all-time lowest seasonal peak of 85.1 cm set in 2006. This kept 2026 from becoming the outright worst season on record for peak depth.

Mid-August: Additional significant snow from the recent polar/Antarctic cold fronts (the same systems that brought the cold to Perth/Bunbury). Forecasts and reports indicated potential for 30–50+ cm (or more in higher terrain) from the systems around 9–13 August, providing a meaningful top-up. Resorts reported fresh snow and improving cover.

Overall character so far: Slow start comparable to other weak early seasons (e.g., 2015, 2017, 2020), but recovering toward more average territory for a slow-start year by mid-winter. Still well below the long-term average for this time of year. Peak is still weeks away (typically late August–early September).

2016 Season

Also a weaker-than-average year overall.

Featured a relatively slow/middling start and limited mid-season accumulation in some periods.

Notable late recovery: Significant snowfalls occurred later, including on the October long weekend and the following week (around 60 cm in one period), with powder reported as late as 23 October. September itself was quiet (only about 15 cm in one fall).

This late snow helped the season but arrived after the main winter period, limiting its impact on peak mid-season conditions and spring skiing base in some respects.

Exact peak depth is not as prominently cited in recent summaries as stronger years (e.g., 2018 peaked at ~225 cm), consistent with it being below the long-term average.

Direct Comparison

Aspect	2026 (to mid-Aug)	2016
Early season (Jun–early Jul)	Extremely poor (near 0 cm early Jul)	Weaker/middling
Mid-July depths	~70 cm	Lower than average for the period
Early–mid August	~87 cm + fresh additions from polar outbreaks	Building but not exceptional
Late season	Ongoing (potential further gains)	Strong late recovery (Oct snow)
Overall ranking	Tracking as a weak-to-average slow-start year so far; avoided record-low peak	Below-average year with late snow help
Context	Helped by recent Antarctic cold fronts	Benefited from late cold systems

2026 has had a significantly worse early-season start than 2016 (and most historical years), with depths that threatened record lows into early August.

However, the recent polar outbreak has delivered a notable mid-August boost similar to the kind of recovery systems that can salvage weaker seasons.

2016 was also not a strong year but finished with helpful late snow. Both are well below the best seasons (peaks often 200+ cm, with some historical years exceeding 300 cm). The final 2026 peak will depend on any remaining August–September systems.

Data is drawn primarily from Snowy Hydro Spencers Creek measurements and contemporary reports from Weatherzone, SnowBrains, and related sources.