The Bureau of Meteorology’s (BOM) first winter forecast of 2024 has confidently declared one of the warmest on record throughout Australia.

It has also slightly favoured above average rain across a wide swathe of the interior and the east.

The BOM’s modelling showed most of Australia will have both minimum and maximum temperatures of about 1-2 degrees Celsius above average.

And for most of Western Australia, Tasmania, Victoria and the eastern seaboard, there is more than 70 per cent chance this winter would be amongst the hottest 20 per cent of winters.

According to the BOM’s winter outlook there is a high probability this winter will be one of the warmest on record. (Supplied: BOM)

If the forecast proves correct, this year could challenge last year’s record warm winter, which brought the national temperature 1.53C above the 1961-1990 baseline average.

While the balmy winter forecast does not indicate the complete absence of cold spells, frost and snow, it does virtually ensure temperatures will be higher than normal when averaged across the entire season.

The bullish outlook is mainly due to the ongoing record high sea surface temperatures around the globe, which warm the overlying air — but is also influenced by Australia’s warming climate, which has seen the temperature rise 1.5C since 1910.

Warm ocean temperatures surrounding Australis should help to boost winter air temperatures. (ABC News)

Why the winter rain forecast may need an adjustment

While the temperature outlook is unambiguous, how much rain falls across Australia is less straightforward due to the far more variable impact climate change and warm oceans have on precipitation.

For most of Australia, the outlook suggested the odds of it being wetter or drier than normal is about 50-50, grading to a 70 per cent chance of above median rainfall across pockets of the interior.

It indicated a 70 per cent chance of below median rain for small parts of the Kimberley, Top End, and the South Australian coast. 

A weak wet signal is currently indicated across much of central and eastern Australia for this winter.(Supplied: BOM)

Is there really no strong indication of whether or not winter will be wet or dry?

As winter approaches and the seasonal forecasts are updated, there’s a distinct possibility the chances of a wet winter will increase due to a major spanner in the works – a possible La Niña.

Since La Niña is on the radar, why isn’t it already factored into the forecast?

The BOMs model used to produce our seasonal outlooks is somewhat of an outlier amongst global models – by holding the Pacific in a Neutral phase during the coming months, while several other forecasting agencies indicate a rapid cooling.

If Pacific Ocean water temperatures drop to La Niña levels this winter, the odds favouring a wetter season will shift upwards, especially across Australia’s central and eastern inland, a trend which would only intensify further into spring and summer.

If La Niña forms, then the chances of a wet winter and spring will increase across much of Australia. (Supplied: BOM)

The additional cloud cover associated with La Niña would also reduce the probability of above average daytime temperatures.

How the BOM can forecast the weather months in advance

Because day-to-day weather can’t be predicted with accuracy more than a week ahead, sometimes people assume seasonal forecasts must be unreliable.

Weather models work by ingesting millions of observations to simulate the atmosphere, then use the laws of physics to calculate how conditions will change into the future.

However, there are key differences that allow a seasonal forecast to have accuracy months in advance, and well beyond a standard weather forecast.

  1. 1.Seasonal forecasts cover a longer time period of up to three months and therefore smooth out the day-to-day noise of fluctuating weather. In other words, a seasonal forecast is just averaging the weather over an extended period compared to normal.
  2. 2.Seasonal forecasts are produced from model ensembles, which involve running a weather model dozens of times and averaging the output to offer guidance on the likelihood of any given scenario. For example, a model could be run 100 times, and if 70 of those runs or “members” show above average rain over a particular location, that would equate to a 70 per cent chance of above average rain.

Posted , updated