John M. Barry is the author of “The Great Influenza: The Story of the Deadliest Pandemic in History and distinguished scholar at Tulane University’s School of Public Health and Tropical Medicine.

In the near future, likely before Christmas, 90 percent or more of the U.S. population will have considerable — but not perfect — immunity to the novel coronavirus, SARS-CoV-2.

Close to 70 percent of Americans have had at least one vaccine dose, while a very substantial if not precisely known percentage of the unvaccinated have recovered from infection. The so-called herd immunity target will be achieved. But what will this actually mean?

For a virus such as measles, our immunity is so good that it almost never infects a protected individual. Herd immunity means that so many individuals have achieved protection that their collective immunity prevents the virus from circulating in that population. For SARS-CoV-2, herd immunity is unfortunately more illusory. Even in the best of circumstances, covid-19 will continue to cause sickness and occasionally death.

Where does that leave us? While no one can predict with certainty what will happen, there are several reasonable scenarios in the near and medium terms. Over the long term, it is harder to say.

In the best short-term scenario, variants will continue to emerge but won’t have significant impacts. Delta now causes 99 percent of U.S. cases; the once-feared alpha variant is nowhere to be seen, and neither are variants that once seemed threatening after they emerged in New York, California, South Africa and elsewhere. Delta, in this scenario, will continue to outcompete any variant that emerges, and severe disease will be limited to the steadily decreasing few who still lack any immune protection. Breakthrough infections will still occur, but the immune systems of those vaccinated or previously infected will protect them against severe disease and death.

In this scenario, based on current death rates in breakthrough infections, covid-19 might very well become less threatening than seasonal influenza. This could happen sooner rather than later.

A less favorable short-term scenario would see the emergence of as-yet-unknown variants better at evading our immune systems than those we have encountered so far. In vulnerable populations lacking any immune protection, delta has elbowed such variants aside. But once the vast majority of people have acquired significant immunity, the competitive advantage could shift to variants better geared to evade immune protection — at least for a while.

That is largely what happened in 1918. As the virus adapted itself to humans, it caused three waves. The second wave was far more transmissible and lethal than the first, and it was followed by a third wave in 1919 that had no problem infecting people previously sickened. This third wave was less lethal, probably because of a combination of viral mutations and moderate protection provided by the immune system. By 1921, the once-lethal virus was transformed into seasonal influenza, killing relatively few over the following decades and having no impact on daily life.

Something like this could occur with covid-19. If so, at least one more wave is coming. It’s likely our immune systems would still protect us against severe illness and death, but, as we do with influenza, we would need to update vaccines to keep pace with mutations. Eventually, life would return to pre-pandemic normal.

In the worst case, a new variant would appear that generates enormous viral loads — much like delta — and also evades the immune system. Such a variant could surface through either random mutation or “recombination,” which occurs when two different SARS-CoV-2 viruses infect the same cell and trade genetic information, creating a hybrid with the worst of each. This could expose even populations that have achieved so-called herd immunity to a new round of risk.

That frightening possibility drives home the need to minimize the chances of it occurring, by vaccinating the billions in the developing world and, longer-term, developing a vaccine effective against all coronaviruses. It also drives home the need for therapeutic drugs to treat those who do get sick.

Is there a best-case scenario in the long term? Yes: Like the 1918 virus, the coronavirus infects cells in both the upper respiratory tract, which makes it easily transmissible, and deep in the lungs, which makes it potentially lethal. Replicating in the upper respiratory tract allows viruses to transmit between people; replicating in the lung does not.

That means evolutionary pressure could push the virus away from the lungs and toward the upper respiratory tract. It may already be happening. Delta creates viral loads in the upper respiratory tract more than 1,000 times greater than earlier variants, but it is not at all clear that the viral load in the lung is comparably higher. If it were, one would expect many more deaths than have occurred.

It’s conceivable, therefore, that the virus could lose its ability to infect the lungs and, quite possibly, other parts of the body. That would be a blessing. How long that would take is not knowable, but this process may have already taken place once. There is evidence that the 1889 “Russian flu” pandemic, which killed about 1 million people globally, was actually caused by a coronavirus known as OC43.

Today, OC43 kills no one. It causes a cold. We can only hope.