As Covid barrels through China, scientists around the world are searching for clues about an outbreak with sprawling consequences — for the health of hundreds of millions of Chinese people, the global economy and the future of the pandemic.

But in the absence of credible information from the Chinese government, it is a big scientific guessing game to determine the size and severity of the surge in the world’s most populous country.

In Hong Kong, one team of researchers pored over passenger data from five Beijing subway lines to determine the potential spread. In Seattle, a group of modelers tried in vain to reverse-engineer an unverified government leak detailing case numbers from Chinese health officials. In Britain, scientists are coming up with their own efficacy estimates of Chinese vaccines.

Any personal anecdote or social media report from China — scarce medicines, overrun hospitals, overflowing crematories — is possible fodder for researchers’ models.

They are all attempting to understand the same things: How quickly is the virus spreading in the country? How many people are dying? Could China be the source of a new and dangerous variant?

As scientists sift through varied sources of shaky information, they are bracing for potentially catastrophic outcomes. Barring new precautionary steps, some worst-case estimates suggest that Covid could kill as many people in China in the next four months as it has Americans during the entire three-year pandemic.

Without satisfying answers, some countries are putting limits on Chinese travelers, albeit based in part on unfounded fears or political motivations. The United States, Italy and Japan have said they will require a negative Covid test for those coming from China, citing concerns that the surge in cases in China could produce new, more threatening variants.

While researchers and virologists said the new measures would most likely do little, if anything, to blunt the spread, the policies reflect the limited visibility into the outbreak. Scientists’ models generally point to an explosive spread and a high death rate, given how many people in China have little to no immunity to Omicron subvariants. But even their estimates are all over the place.

In the bleakest of several scenarios of what the end of China’s “zero Covid” policy might mean, nearly a million people could die during the early months of reopening, Hong Kong researchers reported this month in a study partly funded by the Chinese Center for Disease Control and Prevention, though the study does not provide an exact time period.

An American group estimated as many as half a million deaths by April and another million by the end of 2023 if China rejects social-distancing mandates. Airfinity, a British-based analytics company, this week offered an even more dire short-term forecast: 1.7 million Covid deaths by the end of April.

Until this month, the world seemed to have a reasonably clear understanding of what was happening with the virus in China. The ruling Communist Party proudly published low daily case numbers and deaths as a testament to its stringent “zero Covid” policy. A countrywide system of lockdowns, quarantines and mass testing largely kept the virus at bay.

But in early December, the government abruptly abandoned “zero Covid,” leaving the scientific community largely in the dark.

“Nobody, nobody has a clue,” said Siddharth Sridhar, a clinical virologist with a focus on emerging infectious diseases.

Predicting the path of the pandemic has always been difficult. Even in places like Britain with reliable data, forecasts have often been far off the mark. But scientists have generally used reported Covid deaths as a dependable barometer to determine the potential size of an outbreak.

The data coming from the Chinese government can’t be trusted anymore. Officially, China has claimed just 12 deaths from Covid since Dec. 1. The country has said it will only count those who die from respiratory failure directly linked to an infection, leaving out vast numbers who died because Covid aggravated underlying diseases or caused heart or liver failure.

Experts say the sheer speed of the spread would suggest a much higher number of deaths. One city last week reported half a million cases in one day. Another reported a million.

There are also indications that officials are pressuring doctors and crematories to avoid categorizing even respiratory deaths as virus related.

One doctor at a private hospital in Beijing said he and his colleagues found a typed note on a hospital desk in recent days urging them to “try not to write respiratory failure caused by Covid” as the primary cause of death. The note was shared with The New York Times.

The doctor said it was not clear if the message was generated internally or sent from government officials. But similar warnings have been circulating on Chinese social media telling doctors not to “carelessly write Covid” on death certificates.

Several modelers have even been skeptical of leaked information from government officials on case counts, which have been used to assess the scale of China’s outbreak. One recent estimate, making the rounds in news reports and on Chinese social media, cited data from national health officials that 250 million people had been infected in the first 20 days of December.

Some scientists said that such massive figures indicated either that China had been suppressing data for months or that it was trying to make it seem like the outbreak had peaked.

“Either they know something we don’t,” said Dr. Christopher Murray, the director of the Institute for Health Metrics and Evaluation at the University of Washington, “or they’re trying to say the worst is already over.”

“I suspect it’s now the latter,” he said, referring to the idea that China was trying to make it look like the worst had passed. It seems unlikely that China would have been able to fake the numbers for months without raising suspicions, he said.

The about-face on China’s messaging is also complicating scientists’ assessments. Just a month ago, China’s state-controlled media was warning about the dangers of the virus. Now, it is saying the current Omicron variant is mild and the outbreak is manageable.

Scientists and public health experts, though, are worried that Omicron has looked less severe in other places in large part because those populations had huge stores of immunity, including from past infections — a set of circumstances that does not hold in China. If China tries to soldier through its outbreak, without reimposing public health measures or ramping up vaccinations, scientists are concerned that many more may needlessly die.

The Hong Kong researchers, for example, found that administering more fourth vaccine doses and antiviral medications and using social-distancing measures could save at least 250,000 lives during China’s reopening. Dr. Murray’s team, too, found that social-distancing mandates could help spare hospitals from a concentrated surge of patients, reducing the death toll by 200,000 by April and by even more when combined with greater masking and antiviral use.

How the Chinese public perceives the threat of the outbreak will also be important for its trajectory. Even if people decide to start taking more precautions for only a short period, scientists said, it could mean the difference between hospitals being able to treat their sickest patients or being completely overwhelmed.

The vaccination rate in the country is another major variable. While 90 percent of the population has received two shots, the booster rate is much lower for older Chinese people. The World Health Organization has said three shots are crucial with Chinese vaccines that use inactivated virus.

Extra protection from additional doses should arrive in less than two weeks for people with previous shots, said James Trauer, an expert on modeling infectious diseases at Monash University in Melbourne, Australia. And he noted that the size of the country means that the outbreak will not reach everyone at the same time, giving some places extra time to get more people inoculated.

Scientists are studying transportation patterns to understand how fast the outbreak might spread, but the picture isn’t clear.

The Hong Kong scientists, in their recent study, analyzed passenger data from a handful of Beijing subway lines. The information, they said, suggested that mobility in the city had dropped to low levels as people stayed home to protect themselves against the virus.

But Yanzhong Huang, a senior fellow for global health at the Council on Foreign Relations, said there were some indications that at least in big cities, foot traffic was picking up and restaurants were getting busier.

“That sort of seems to challenge the notion that people are actually exercising precaution,” he said.

Without better indications of how often Covid infections are turning deadly in China, many scientists have leaned on comparisons with Hong Kong. The Chinese territory, which like China had also been slow to encourage vaccines, was particularly vulnerable when Omicron began spreading there in early 2022.

Some models have assumed that China would experience an infection fatality ratio very similar to Hong Kong’s in the early stages of its outbreak. Back then, nearly 10,000 people in a territory of 7.5 million died within months of Omicron spreading. A comparable toll in China, with its 1.4 billion people, would be far higher.

But there are also important differences. China has stronger vaccine coverage in its older population than Hong Kong did at the start of its surge.

Based on the timing of their respective outbreaks, though, China’s population-wide vaccination drive was earlier than in Hong Kong, meaning the effects of inoculations had longer to wane. Hong Kong also provided the option of Western vaccines with newer mRNA technology, while China relied exclusively on homegrown, less effective vaccines. Hospitals may also have a harder time handling the surge in some parts of China.

The general lack of clarity has led to worries that the size of the outbreak could create more opportunities for the virus circulating through China — imported versions of Omicron — to mutate into a more dangerous variant.

But scientists are skeptical of such a scenario in China’s current outbreak.

Variants similar to those that China has reported were largely outcompeted months ago in the United States by more contagious or more elusive Omicron subvariants. After Italy mandated testing for travelers from China, it said the first cases it sequenced were all caused by an Omicron variant already present in Italy. European Union health officials said on Thursday that screening travelers from China was unjustified.

“We’ve had a huge number of infections internationally,” said James Wood, an infectious disease expert at the University of New South Wales in Sydney, estimating that most people globally had caught the virus. “That’s a lot more infections than have occurred in China alone.”

Additional reporting and research by Zixu Wang, Joy Dong and Elisabetta Povoledo.