"You are safer outside than inside," he said. "I do not see how anyone's going to get his virus if they keep two metres away from someone and I don't see how anyone's going to get it if they sit on a park bench."

Beach parties should certainly be broken up, Professor Collignon, of the Australian National University medical school, said. But it made no sense to stop people walking on the beach, or pausing on the beach to rest.

"These are panic decisions not decisions based on data," he said.

After concern about huge crowds on Bondi in Sydney, beaches have been closed in some parts of the country, and stories have emerged of people being stopped by police when they attempt to walk along a beach. Golf Australia has told clubs to close because they no longer fall under the list of acceptable reasons for leaving home.

Professor Collignon said extreme measures were unwarranted and risked rebellion, especially among younger people least at risk from the virus.

"My real worry is if we overdo it now we will have people particularly in 30s and 40s who will say 'stuff this'," he said.

"I think this is not sustainable for six months. We have to do everything we can to minimise the spread to others, but not do things that don't even make biological sense."

Professor Collignon was speaking as Australia's total reached 4860 at 6am on Thursday, an increase of 8 per cent on the day before, continuing the lower rate of increase seen each day for about a week.

Professor Collignon also took aim at modeling that suggests hundreds of thousands of deaths, saying the same models were used for the SARS, ebola and swine flu epidemics when they had been wrong.

Modellers had predicted tens of thousands of swine flu deaths but in Australia there had been 191.

"All these catastrophic people have been wrong every time they've made predictions," with the modelling done on what he believes is the false presumption that coronavirus has a R0 of 2 (meaning every infected person infects two others). Professor Collignon said the R0 was probably closer to 1.2 to 1.4.

Australia was already seeing the "curve" flattening, with the number of new cases each day on a downward trajectory. Places like New York and Italy had likely had uncontrolled community spread for a couple of months before the alert was raised.

"Just like everybody else, I haven't got a crystal ball, but the current data is reassuring that what we put in place 10 days ago has substantially decreased the spread and we're not going to see an epidemic before winter. All of these additional things will make very difference, but will cause huge social problems," he said.

Australia's success was not because of the new limits, which restrict gatherings to no more than two people, with hefty fines, and tell people they should only leave home for essentials.

Rather, because coronavirus had an incubation period of about five days or 10 at most, the success was due to measures put in place 10 days ago, including the closing of pubs, clubs and restaurants, the social-distancing measures and the quarantining of overseas arrivals. Those measures should continue, Professor Collignon said.

Australia was highly unlikely to eliminate the virus, so must find ways to live with it in small numbers without a long lockdown.

"My view is where are going to have this probably for the next 18 months to two years because that will be the earliest that a vaccine will become available, that's my pessimistic outlook," he said. "We have to come up with ways that minimise the transmissions of this virus, but I don't think we can all become hermits for the next 18 months to two years."

READ MORE:

Asked for a response, Deputy Chief Health Officer Dr Nick Coatsworth, said, "The current measures to limit gatherings are consistent with those adopted in other nations such as the United Kingdom and are necessary to prevent Australia's epidemic worsening."

Australian authorities have so far refused to release the modelling behind their decisions, and while Deputy Chief Medical Officer Paul Kelly said earlier this week that the modelling would be "unlocked" this week, he appeared to backpedal a little on Wednesday.

Asked about expectations that it would be released, he said, "I would like to clarify that we will discuss the modelling and look to make that transparent in coming days."

He did not elaborate on that and the Health Department did not respond to a request for clarification.

Professor Kelly said transparency was important but "modelling can be misinterpreted and we need to make sure that it is presented in a way that is useful".

"The other thing I would say about modelling and modellers themselves would also say this, modelling is not necessarily the truth. It is a way of seeing the world and potentially pointing out how one can influence certain scenarios," he said.

New Zealand released its modelling this week, showing that the country was planning for 6.8 per cent of its population requiring hospitalisation and 0.68 per cent dying. The numbers were described as close to the "worst case" scenario, with New Zealand since going into lockdown to avoid such a dire result.

If extrapolated to Australia and assuming the numbers are directly comparable, the New Zealand figures suggest between 65,000 and 174,000 deaths and 1.74 million hospitalisations here.

Our COVID-19 news articles relating to public health and safety are free for anyone to access. However, we depend on subscription revenue to support our journalism. If you are able, please subscribe here. If you are already a subscriber, thank you for your support. If you're looking to stay up to date on COVID-19, you can also sign up for our twice-daily digest here.