It is a question that puzzles both those on the front line fighting Covid 19 and the experts developing strategies to combat its next move: why has London not seen a second flare-up when other parts of the UK are now having to introduce new lockdown restrictions?

“It’s a bit of an enigma, given that London very definitely led during the initial peak,” said Professor David Alexander, who is based at the Institute for Risk and Disaster Reduction at University College London.

Six months ago, the capital was hit hard and hit first by the pandemic. Wards were converted to treat Covid-19 patients and a temporary hospital was opened in London’s Docklands amid concerns that the capital’s health system would collapse.

But talk to doctors who were re-purposed to fight the pandemic back then and the picture now is very different.

“We are pretty much totally back to normal and we’ve virtually no Covid cases at the moment,” said one doctor at a north London hospital, who spoke on condition of anonymity.

It is a broadly similar picture across all of London’s 32 boroughs, where the number of cases per 100,000 people last week ranged from 5.7 in Bromley to a high of 17.9 in Kensington and Chelsea. That compares with recent totals of 37.3 per 100,000 in Manchester, 28.7 in Leeds and 27.8 in Birmingham, the first and last of which have had new restrictions imposed.

Richard Harris, professor of quantitative social geography at the University of Bristol, notes that many of these areas share a characteristic that is not so common in London. “Regional economic factors may be playing a part – the jobs people are working in,” Harris said. “It may be that blue-collar workers in manufacturing are more exposed.”

In contrast, it is possible that more people in London have been able to work from home or have been furloughed, limiting their exposure.

“London is much bigger, more diverse, has more contacts with the outside world – more people pass through it, and so forth – and so you would expect that situation [further flare-ups] to persist in some form, Harris said. “The only thing I can think of is that it must be a result of changes in people’s patterns of behaviour, such as working from home and not venturing out so much.”

This is a theory that Dr Derek Groen, a lecturer in simulation and modelling at Brunel University London, believes may offer a partial explanation. “Because London was hit so hard, people have been more willing to comply with social distancing measures – people have been more cautious,” he suggests.

There are concerns, however, that the capital remains vulnerable should a severe second wave emerge. Leaked government calculations of a “reasonable worst-case scenario” for a second wave in London suggest that there could be 15,100 excess deaths, compared with just under 9,000 seen during the first wave.

The data, seen by the Observer, covers a 38-week period from July to March next year.

The modelling shows an infection rate doubling in August, and again in early September before levelling out until school half-term week in late October. It then rises steeply in November and December, before levelling out again in January and falling by a fifth by the end of March.

The modelling has been criticised by some as unhelpful, because it assumes little will have been learned from dealing with the first wave of infections. The government has said that the leaked information is not a forecast or prediction, but just one element of its planning.

Meanwhile, open-source coronavirus simulations, carried out in seven London boroughs by Groen and his colleagues, suggest that a second wave of Covid-19 is probable in “almost all cases” although they believe that the outbreak will be less severe than the first.

Groen explained that although the UK is nowhere close to achieving herd immunity – when 60 to 80% of the population is infected – flare-ups are “less steep” where more of a region or city’s population has been infected.

Public Health England’s surveillance studies estimated that up to 17.5% of Londoners had caught the virus by late June. In contrast, separate estimates suggest that only between 5% and 7% of the UK population overall has been infected. In some parts of the UK it is as low as 3%.

Rowland Kao, professor of veterinary epidemiology and data science at the University of Edinburgh, believes that seroprevalence – the proportion of a population that has been exposed to a particular pathogen – could be an important factor in explaining why London has not seen the sharp rise in Covid-19 cases experienced in cities such as Glasgow.

“The UK Biobank serology study showed low seroprevalence overall, but significant variation across regions and different ethnicities and social conditions. In particular, London had the highest seroprevalence, and Glasgow, like the rest of Scotland, is likely to be quite a bit lower.

“As the R number [the number of people that one infected person will pass the virus on to] remains low overall – generally near 1 – these relatively small differences in seroprevalence may mean the difference between having substantial outbreaks and being in a position of being able to contain them.”

Kao points out that university students have yet to return to the capital, while tourists have, so far, largely stayed away. But the picture may be very different by Christmas.

The north London doctor certainly does not think London will escape another flare-up. “I think it is still on the way, though social distancing and mask-wearing may make a difference, as well as more protection of people in public-facing jobs. We’re all bracing ourselves for a second wave. But it is later than we expected.”