Likewise, the chance of a person infected with Covid-19 in a gathering of 50 people in Somerset is only one in eight (13 per cent), compared to nearly nine in ten (86 per cent) at risk in Liverpool.

The model, originally designed to demonstrate the risks of large gatherings during the US football season, has been expanded to display the risk of event volumes across the United Kingdom, Italy and Switzerland.

People can use it to decide whether they want to visit a local bar or restaurant, or by policymakers to enforce more local closures, said Joshua Weitz, a Georgia Tech professor of biological sciences and physics who helped create the map.

He said, “If you knew there was a one-in-four chance … that someone in that bar or restaurant, or gathering, had Covid-19 … I hope that changes someone’s behavior.”

He also said that low-risk areas should not think they have “an absolute mandate” to meet in very large groups without wearing a mask.

“This will once again facilitate further spread,” he added.

“But this means that depending on the region, there are completely different levels of risk (for large gatherings) and this really reflects the persistent and heterogeneous differences in the spreading infection.”

Scientists used local authority data from Public Health England (PHE).

Elsewhere, the risk of encountering a person with Covid-19 in a 50-person event in Blackburn with Darwin, for example, is about four times higher than in Bromley (85 percent versus 23 percent).

While the chance of someone contracting the virus in a gathering of 50 people in County Durham is about 50 percent higher than in Suffolk (18 percent versus 67 percent).

In Lancashire, the risk of someone contracting Covid-19 at a 25-person event is more than twice that in neighboring North Yorkshire, 47 percent versus 20 percent.