On a balmy spring evening in November 2016, a thunderstorm struck as Melburnians were on their way home from work.
The storm prompted ryegrass pollen swept in from agricultural areas on a north-westerly to burst, spreading tiny particles through the air and deep into the lungs of unsuspecting residents.
Within an hour, thousands of calls to emergency services began flooding in.
By the time it was over, seven men and three women – all aged between 18 and 57 – were dead.
However, just what made that storm – innocuous by any other standard – so deadly is still not fully understood.
Now, new in-depth research by some of Australia’s leading meteorological bodies including the CSIRO suggests that models to predict such events are based on a number of false assumptions about what causes the pollen particles to rupture.
This would mean that the combination of a high pollen count in the air and rainfall combined to create the 2016 event.
However, notable during the storm that night was the absence of significant rainfall: just 4mm fell to the west of Melbourne and 2mm to the east.
Humidity levels before the storm remained very low following the city’s first burst of summer heat that season.
Weather stations across Melbourne recording humidity levels of between just 18 and 19 per cent.
This didn’t rise to the level necessary to burst the pollen (80 per cent) until five hours after the storm had passed – yet the influx of calls to emergency services began within 30 minutes of the storm.
Moreover, if rainfall combined with a high pollen count were enough to create thunderstorm asthma, such deadly events would be far more common than they actually are.
As it is, thunderstorm asthma is more common in south-east Australia than in any other part of the world, yet still only occurs just once in every five years in Melbourne on average.
These contradictory findings prompted the researchers to investigate other mechanisms which could have caused the pollen to burst.
They began experiments to see if whole pollen could be burst by wind gusts, electrical activity or lightning.
It is the first time that the cause of thunderstorm asthma has been explored using atmospheric modelling studies.
The researchers tried blasting whole pollen particles against strong wind gusts to see if they would burst.
However, when mapped against the weather conditions leading up to the storm, the results suggested more pollen should have ruptured the day before, when the pollen count was actually higher – with a maximum hourly average of 177 grains per cubic metre, compared with 102 grains the following day.
They then turned to electrical activity as a potential driver, but reached similar conclusions.
It was only when they looked at lightning – which proliferated during the November 21 storm – that the timing of the wave of asthma attacks matched up with the weather conditions.
Questions still remain unanswered.
For one thing, the lightning struck not in Melbourne itself – where the asthma attacks were concentrated – but to the west and south.
The researchers say further studies are needed to determine the exact combination of weather conditions that produce thunderstorm asthma – even suggesting that the storm itself may be less critical than the combination of wind currents in the atmosphere.
Whatever the cause, they say it remains clear that the rain itself was not the main cause of the catastrophic loss of life in Melbourne that day.