Mohammad Sajadi and Anthony Amoroso are associate professors of medicine at the Institute of Human Virology at the University of Maryland School of Medicine and members of the Global Virus Network.

As physicians and researchers, we have spoken to doctors across the country over the past few weeks and kept hearing variations on this description of the novel coronavirus’s spread: It’s like a chemical bomb went off. One day everything seemed normal, and the next everyone became sick: patients, nurses, family members.

The pattern is scary, but familiar. Seasonal respiratory viruses, most infamously influenza, can act in the same explosive fashion. In temperate areas, every fall and winter, as the weather cools and humidity drops, influenza ferociously takes hold and retains a strong grip until spring. There are four coronaviruses — viruses that are common in humans, usually causing moderate respiratory illnesses — that are also seasonal.

With covid-19, there is an even more frightening pattern of exponential growth than exists for influenza. At times it seems to be ripping through populations that have no previous immunity. But there also appears to be a distinct geographical pattern. The virus has spread globally with clusters of infection, often spread by travelers, in most countries. But certain locations were hit much harder than others: Wuhan in China, Daegu in South Korea, Milan, Tehran, central France, northern Spain, Germany, Seattle, and now New York.

Why not Hong Kong, Singapore or Bangkok? They were directly in harm’s way, relatively near Wuhan, where the pandemic began, and highly connected to it by trade and travel. Are we to believe, as many assert, that these places were simply superior at implementing quarantine procedures? And what about cities well north of China, such as Moscow or Nur-Sultan, Kazakhstan, that have also not been much affected?

When we noticed this pattern, we began to suspect that the coronavirus causing covid-19 was spreading like a seasonal virus. We launched a study at the Institute of Human Virology, part of the Global Virus Network, along with colleagues at the University of Maryland at College Park and researchers in Iran, to try to find out.

Our research determined that all the major affected areas in January through early March were in the Northern Hemisphere between 30 and 50 degrees north latitude. And the areas had nearly identical temperatures of 41 to 52 degrees Fahrenheit and low humidity in the weeks before they recorded their first covid-19 death.

Those conditions are similar to the laboratory conditions that maximize survival for seasonal respiratory viruses and are most conducive for transmission. Cooler temperatures are known to increase the survival of respiratory viruses on surfaces, help stabilize (along with low humidity) the aerosol floating in air, and even increase viral replication in some instances.

The notion that the virus causing covid-19 will behave as a seasonal respiratory virus is a controversial one, because a whole season’s worth of data isn’t available yet, and because it risks stirring false hope that might inspire shortsighted policymaking. Even though we believe that covid-19 is behaving as a seasonal respiratory virus, that view comes with an enormous caveat: The study of past pandemics, from the whooping cough in 1400s Persia to the 2009 H1N1 virus in the United States, shows that new infections that eventually are regarded as seasonal afflictions can behave erratically (due to a lack of widespread immunity) until they settle into their characteristic pattern.

The campaigns against covid-19 in areas now affected should certainly not be tailored in the expectation of a seasonal retreat that might not come. But bolstering preparedness in northern regions where the pandemic could soon spread would be wise.

We’re working on models right now to validate and refine our findings and identify the communities where likely temperature and humidity levels in April and May could make them a covid-19 target, and those likely to be hit in the coming fall and winter. That could greatly aid preparation efforts, including the deployment of testing and protective equipment.

Could covid-19 disappear permanently, as SARS did, in the summer? Will it be controlled by public health measures, as MERS was? Or will it, like the four other seasonal coronaviruses, retreat into the tropics and Southern Hemisphere, set up a home, and become the fifth human coronavirus, reemerging again in the fall for another season or longer? Will the extraordinary current global efforts send this pathogen into oblivion?

To answer these questions, it is essential to conduct widespread serological testing so scientists can fully understand the pandemic, establishing a profile of who has and has not been infected and allowing improved modeling of covid-19’s spread. The world isn’t helpless, and the fightback is well underway, but a vital weapon will be finding out whether covid-19 will indeed act like other human coronaviruses and change with the seasons.

Read more: