How science won the World Rally Championship

Prudent motorists always check the weather before they set off. But the Estonian rally champions Ott Tänak and Martin Järveoja have something better: predictions about the state of every stretch of road, courtesy of meteorologist Hannes Tõnisson.

Professional rally teams typically use detailed weather forecasts, but Tõnisson’s speciality is predicting how the weather will affect each part of a given rally stage, where road terrain can vary from tarmac and gravel to mud, snow and ice.

“The drivers don’t want to know what the weather is, they want to know the conditions on the road,” says Tõnisson, a senior researcher at Talinn University’s Institute of Ecology exploring how climate change is set to impact Estonia. “I always go to the rally many days before, so we have the possibility to drive through all the stages where the drivers are going to drive, so I have quite a good idea of what the roads look like and what’s going to happen with them.”

Drawing on his research experience of monitoring the on-the-ground effects of extreme weather, Tõnisson is using computer models and mapping software to figure out exactly what a weather forecast will mean for the race ahead.  Last year, Tõnisson’s expertise helped Tänak and co-driver Järveoja win the World Rally Championship. Tänak is the first ever champion from Estonia, and the first non-Frenchman to win the title for 15 years.

A long-time rally enthusiast, Tõnisson began making his predictions as a spectator, just for fun, before the value of what he was doing was spotted by Tänak’s team.

Answering rally’s most important questions

If it’s going to rain, Tõnisson can estimate how quickly the road will dry afterwards, or where pools of water are likely to form and how deep they will be. By predicting air temperature and humidity, Tõnisson helps teams optimise the car’s engine cooling. If there’s going to be ice, Tõnisson can work out how much of the road is likely to be covered by it and how much will be clear, which can determine whether winter tyres are a help or a hindrance.

“In rally, one of the most important questions is what kind of tyres you’re going to use,” explains Tõnisson. Soft tyres grip well in the rain, but wear out fast in dry weather, while hard tyres are apt to skid on a wet road.

There’s more value to Tõnisson’s predictions than just setting up the car though – what matters most is how the driver performs on the day. “The final position and the stage time depend on the driver,” he says, “the more detailed the information is, and the fewer the surprises the driver has to deal with during the rally, the more confidence he has. And when they are more confident, they can actually drive much, much faster.”

From storm chasing to rally driving

When he’s not helping top rally drivers win world championships, Tõnisson is investigating the past few thousand years’ worth of Estonian climate history. He says that thinking through how the weather would affect rally stages “was pretty much related to my work,” which involved “a kind of storm chasing.” Tõnisson would predict when extreme storms would hit the Estonian coast, then head to the beach just before a storm arrived, and study its effects.

Tõnisson is researching how information locked away in Estonia’s ancient coastal formations can lead to better modelling of how future climate change will affect the weather.

He explains that while many countries are losing land to rising sea levels, the opposite is happening to Estonia because the land is rising even faster than the sea. “More than 10,000 years ago, Estonia was covered with a very thick ice layer, more than two kilometres thick. When the ice melted, the land started to uplift, because the weight of the ice disappeared, and now this land is trying to restore the original position – and this uplift is still on-going.”

That means Estonia’s former coastlines are inland instead of under the sea, making them easier to study with ground-penetrating radar. “Those coastal formations always contain information about the conditions when they were formed,” Tõnisson says.

That information helps scientists get a better understanding of what the weather was like thousands of years ago, to improve predictions of what’s likely to happen in the future.

“Currently all the scenarios for the future – for inundations or storms, for example – are mostly made based on observational data, and that data only goes back 100, sometimes 150 years,” says Tõnisson. That makes it difficult to predict the future, especially when the last 150 years were much cooler than those to come are expected to be.

To illustrate his point, Tõnisson gives the example of trying to predict a child’s growth: if you only base your predictions on the first year’s growth, you’d expect the child to be about two metres tall by the time he’s three. The same goes for relying on just 150 years of climate data.

“If you go back to this climate data that we have recorded in the old beaches, it’s like a history book, that has written down most of the major storms during the last 7,000 years,” says Tõnisson.  “The next step, which we haven’t taken yet, is to model what were the conditions during those storms compared to the ones that we have faced here in the recent past. Were they similar, or much more extensive? And if they were much more extensive, how much larger was the magnitude of those storms? And finally we want to reach new risk scenarios – for example, if the climate continues to warm, how likely is it that we will have even larger storms?”