New Zealand's next Covid-19 wave could take off at the back-end of winter, modellers say - and it's possible the country's eased restrictions could drive a bump in cases within weeks.
But picking the earliest signs of another major surge is fraught with complexity – and a dearth of detailed data doesn't help experts who are closely tracking case trends.
What indicators could we turn to for a warning?
The new numbers the Ministry of Health reports each day can give us a rough picture of what Covid-19 is doing in our communities – but it's far from a complete one.
Going off the seven-day average of daily reported cases, we can see that the Omicron outbreak peaked, with just over 4000 cases per million of population, around the second week of March.
Since sharply rising to that peak, that value has been gradually coming down – the latest seven-day average had case numbers hovering around 1450 cases per million – although regions have been seeing their own peaks as the outbreak fanned out beyond Auckland.
Experts point out reported case numbers aren't a perfect indicator, given factors like asymptomatic infection and the move to self-reporting through rapid antigen tests would hide much of the iceberg of infection.
Covid-19 modeller Professor Michael Plank has said official figures might only represent one in three, or four, actual infections.
As well, shifts in own behaviour – whether taking fewer precautions like wearing masks, or just spending more time inside as the seasons change – could also blur the picture, and cause the case tally to bounce around.
"People may have been limiting their contacts during the peak of the wave, and maybe now they're starting to relax - but there's no real way for us to directly measure that."
If there happened to be a sudden spike in cases, Plank said it would still be hard to immediately deduce another wave was upon us.
"It could be a little blip, or it could be a bigger wave, so it would be hard to say just based on an upturn of cases."
Plank didn't expect to see another national Omicron surge in the immediate term – high vaccination coverage and exposure to the virus has given the population a good degree of immunity for now – but it was still "quite possible" we might see cases climb gradually in the next few weeks.
"It could be that we see a bit of an effect in the change to orange and the relaxation of peoples' behaviour," he said.
"But that's probably not likely to cause a massive increase in case numbers: it's more likely to be a bump."
Reopening our borders to the world wasn't expected to have much of an effect on cases, he added, as visitors were arriving into a country where the virus had already been widely spreading.
When an actual second wave did begin – Plank's current best guess was perhaps some time around August - he said the trend would likely be a slow build-up in the rolling case average.
"I'd be surprised if we saw the sorts of growth rates we did back in February and March, where we were getting cases doubling, twice a week."
In the absence of better data, epidemiologists have pointed to hospitalisations as a useful gauge for outbreak patterns.
That's due to the simple reason that hospitalisation figures were less prone to the variability of case reporting.
But like deaths, these didn't give a clear snapshot of actual infection: and many of these figures merely reflected patients who'd turned up at hospitals with something unrelated to Covid-19, to find out they'd caught the virus.
Another reason why this data didn't make an ideal indicator was because of the oft-highlighted lag – hospitalisation numbers tended to trail case numbers by a fortnight.
"By the time you see hospitalisations start to tick up, everything else has already gone up too," said Plank's colleague at Covid-19 Modelling Aotearoa, Dr Dion O'Neale.
"If you hadn't already seen that things were rising, you'd have missed a bunch of indicators."
Plank added that demographic factors also complicated things.
"It's important to remember that hospitalisations are very sensitive to age, in particular," he said.
"What we're seeing at the moment is there's been a gradual shift from younger age groups into older ones over time – and that's contributing to keeping our hospital and death numbers relatively high."
Among younger Kiwis, Plank said the cumulative attack rate had been high – something that merely reflected the sheer amount of Omicron infection so far.
"Whereas in older age groups, there are still a lot more people who haven't been infected yet."
Thus, it could be that, by the time the second wave kicked off and case counts began climbing again, hospitalisation numbers themselves might still be tracking at relatively high levels.
Plank and O'Neale singled out the most obvious coalmine canary: a jump in people suddenly contracting the virus again.
Elsewhere in the world, Covid-19 waves have been clearly set apart by reinfections as waned immunity and fresh variants have left populations repeatedly susceptible.
Because Omicron's original sub-type BA.1 and the faster-spreading BA.2 hit New Zealand in relatively close order, our country hasn't experienced the same one-two punch that Omicron has delivered in other places.
Immunologists have reported that infection with one sub-type of Omicron afforded good cross protection against another – and that it was rare for people to contract both within a few months.
But given Omicron has only been circulating in the world for about four months, scientists are still learning about how long we can expect our natural and booster immunity to hold against it.
Plank said such uncertainties around waning immunity continued to make it difficult to pick the timing of the next wave.
"Nonetheless, I think how many reinfections we're getting is going to be an increasingly important indicator," he said.
"If we start to see a significant increase in these, then that would point to the possibility of a second wave being driven by waning immunity."
Despite New Zealand's Omicron outbreak launching from Auckland, Plank said the city wouldn't necessarily be the first place the next wave began from.
Perhaps the bigger question was how we'd be able to actually observe these reinfections, given the Ministry of Health wasn't yet releasing that data in its regular case reporting.
Here, O'Neale said an infection prevalence survey – like that run by UK's Office of National Statistics – could help enormously.
The ministry was planning to establish a New Zealand survey over coming months, but it wasn't yet clear whether it would run longitudinally, like the UK's.
"Our next wave will potentially be driven by reinfection with Omicron - or a new variant showing up," O'Neale said.
"If that happens, we might be able to use international data to look at things like how much previous infection protects us against it.
"So, it'd be really good to know how many people have actually been infected, and when they'd been infected.
"In missing that, we remain really limited in what we can say about protection from reinfection."
The incident was reported about 7.30pm.