The team’s findings, published on Wednesday in the journal Nature, prove that a classical power law – where a relative change in one quantity results in a proportional relative change in another – called the Initial Mass Function (IMF) is not universal but varies with time.

First proposed in 1955 by Austrian astrophysicist Edwin Salpeter, the IMF is one of the most important concepts in modern astronomy. It describes the demographics of a newly formed galaxy – how many stars of each size are created when their parent gas cloud collapses under gravity.

The IMF underpins astronomers’ understanding of virtually everything, from star formation to the evolution of galaxies, supernova explosions and the behaviour of black holes.

Liu Chao from the National Astronomical Observatories in Beijing, lead author of the paper, said the idea of a universal IMF – which was built on direct observations of the Milky Way – is counterintuitive.

For instance, the number of stars being born could easily vary with the temperature of the collapsing gas cloud, and the IMF of our home galaxy does not necessarily represent the IMF elsewhere in the universe, he said.

Scientists have been using increasingly powerful telescopes to measure smaller, fainter stars to test the IMF’s assumption that a new stellar population always consists of a fixed proportion of relatively few massive stars, a moderate number that are sun-sized, and numerous stars that are smaller than the sun.

One of these is LAMOST, a reflecting telescope near Beijing with a diameter of 4 metres (13.1 feet) which can take the spectra of thousands of stars at once.

“LAMOST is very good at measuring the basic properties of stars in the Milky Way, including their temperature, velocity and chemical composition. It also measures metallicity, which is the abundance of elements in a star that are heavier than helium,” Liu said.

How much dark matter is in the Milky Way? Maybe less than we thought

As the most efficient ground-based telescope of its kind, LAMOST has collected the spectra of millions of stars in the past decade. “This data, combined with the stars’ position and distance information from Europe’s Gaia space telescope, made our research possible,” he said.

Astrophysicist Pavel Kroupa from the University of Bonn in Germany said the scale of the survey underpinning the Chinese research was unprecedented. The previously unavailable large sample had made it possible to study subtle differences in the mass distribution of stars, he added.

“Some results were known from previous observations, but this is the first ever study which verifies such systematic differences on a star-by-star observation,” said Kroupa, who was not involved in the research.

“This result will have a major impact on how astronomers interpret the light they observed with, for example, the James Webb Space Telescope, which was emitted when the universe was forming its first stars,” he said.

“Meanwhile, the evolution of galaxies needs to be recomputed with the now verified variation of the IMF.”

Liu and his colleagues from the Chinese Academy of Sciences and Nanjing University started their research in 2015 by selecting 93,000 nearby stars and grouping them according to their metallicity The team then calculated the distribution of mass in each group.

“We used low-mass stars because their lifespans can be well over 100 billion years, and their mass distribution does not change over time,” Liu said.

The researchers found that the universe’s oldest stellar populations lacked low-mass stars, in sharp contrast with the younger galaxies observed today, which are full of them.

They also found that the proportion of stars in young populations varied with their metallicity. The more heavy elements are contained in parent gas clouds, the more low-mass stars they will produce, the paper said.

Chinese telescope catches one of biggest gamma ray blasts to hit Earth

Liu said the team spent a lot of time ruling out possible biases from star selection and calculations to make their results robust.

However, the findings still need to be cross-examined with data from other surveys before they can be widely accepted as the basis for constructing a more accurate history of our own galaxy, the Milky Way, he said.

In particular, Liu is looking forward to next year’s launch of the Hubble-class Chinese Space Station Telescope (CSST), with imaging power over a wide range of wavelengths.

Liu said CSST will provide scientists with an even larger sample of stars in and outside the Milky Way – especially those which are more massive than the sun but relatively rare and hard to detect with existing facilities.