Star Forming Clouds Live Shorter Lives in Busy Galaxies

The Universe’s Fastest Demolition Crew

On the outskirts of a small, scrappy galaxy, a cloud of gas and dust the size of a city block is about to collapse. It has been gathering material for millions of years, slowly pulling itself together under its own gravity. Soon, it will ignite a star. And then, almost immediately, that star will begin to destroy the very cloud that gave it birth.

This is not a tragedy. It is the normal lifecycle of matter in the universe. But a new study suggests that the speed of that lifecycle depends heavily on where a cloud lives.

In busy, star forming galaxies, clouds die young. In quieter neighborhoods, they linger. The difference is not subtle. And it is forcing astronomers to rethink how galaxies actually build their stars.

Jaeyeon Kim and his colleagues at the Max Planck Institute for Astronomy, along with collaborators from the PHANGS survey, have measured the lifetimes of giant molecular clouds across 54 different galaxies (Kim et al., 2022). It is the largest systematic study of its kind. And the numbers are striking.

Clouds that are visible in carbon monoxide (CO) emission survive for between 5 and 30 million years. That is the equivalent of a few dozen orbits around the galactic center. But the actual star formation, the part where new stars become visible, is over in a blink: 1 to 5 million years.

The short version: a molecular cloud spends most of its life doing nothing visible, then collapses, makes stars, and gets torn apart by those stars before you can say "stellar feedback."

Why Some Clouds Die Faster Than Others

The central finding of the Kim paper is that cloud lifetimes correlate strongly with galactic environment. In less massive galaxies, clouds live shorter lives. In more massive galaxies, they last longer.

At first glance, this seems backwards. You might expect big galaxies with lots of stars and radiation to be more hostile to clouds. But the opposite is true. The key is not the violence of the environment, but the composition of the cloud itself.

Kim and his team found that the CO visible cloud lifetimes become shorter with decreasing galaxy mass, mostly due to the increasing presence of CO dark molecular gas in such environments (Kim et al., 2022). In other words, small galaxies have more gas that is molecular but not bright in CO. That gas is harder to see, but it still participates in star formation. The clouds we do see in CO are just the tip of the iceberg, and they are the ones that get destroyed fastest.

This is a subtle but important point. Astronomers have long used CO as a tracer of molecular gas. It is the workhorse of the field. But if CO is only telling part of the story, and if that part is biased toward short lived clouds, then our entire picture of star formation might be skewed.

How to Measure Something You Cannot See

The methodology behind this study is worth understanding, because it is clever and because it reveals the limits of what we can know.

The PHANGS survey uses the Atacama Large Millimeter Array (ALMA) to map CO emission across entire galaxies. This gives the location and velocity of molecular clouds. Then they use narrow band H alpha imaging from the Hubble Space Telescope to find young, massive stars that have recently emerged from their birth clouds.

The trick is statistical. You cannot watch a single cloud form and collapse. That takes millions of years. But you can look at thousands of clouds at different stages of their life and infer the timing.

Kim and his team measured the separation between CO clouds and H alpha regions. If a cloud is close to a young star cluster, it is probably still forming stars. If it is far away, it has already been dispersed. By mapping the spatial distribution of these pairs across 54 galaxies, they could calculate the average time it takes for a cloud to go from CO bright to H alpha bright to nothing.

The numbers: cloud lifetimes of 5 to 30 million years. Star formation efficiency of 1 to 8 percent. Dispersal time of 1 to 5 million years once the stars become visible.

These ranges are not measurement errors. They reflect real physical variation from galaxy to galaxy (Kim et al., 2022). Some galaxies are just faster at making and destroying clouds than others.

The Feedback Loop That Kills Clouds

The reason clouds die so quickly is stellar feedback. Once a massive star turns on, it emits intense ultraviolet radiation, drives powerful winds, and eventually explodes as a supernova. All of these processes inject energy into the surrounding gas, heating it, ionizing it, and pushing it away.

In the Kim study, the authors found that clouds are efficiently dispersed by stellar feedback within 1 to 5 Myr once the star forming region becomes partially exposed (Kim et al., 2022). That is fast. Faster than many theoretical models predicted.

The implication is that star formation is a self limiting process. A cloud cannot convert all its gas into stars. It only uses a small fraction, 1 to 8 percent, before the feedback from those newborn stars shuts down further formation. The rest of the gas gets blown away, recycled into the interstellar medium, and eventually might form new clouds elsewhere.

This is the engine that drives galaxy evolution. Gas cycles from diffuse atomic clouds to dense molecular clouds to stars, and then back to diffuse gas. The speed of that cycle determines how fast a galaxy can turn its gas into stars. And that speed depends on the environment.

The Galaxy Mass Effect

The most interesting correlation in the paper is between galaxy mass and cloud lifetime. Lower mass galaxies have shorter cloud lifetimes. Higher mass galaxies have longer ones.

Why? The authors point to the increasing presence of CO dark molecular gas in low mass environments (Kim et al., 2022). In small galaxies, the gas is less metallic, meaning it has fewer heavy elements. CO emission depends on carbon and oxygen, so low metallicity gas is harder to detect in CO. But it is still molecular. It still forms stars.

This means that in low mass galaxies, the clouds we see in CO are a biased sample. They are the ones that happen to have enough metals to be visible. And those clouds are also the ones that get destroyed fastest. The invisible clouds, the CO dark ones, might live longer.

Or they might not. The data cannot tell us yet. But the correlation is strong enough that it changes how we interpret CO surveys.

What This Does Not Prove

This is a correlational study. It shows that cloud lifetimes vary with galaxy mass and other environmental properties. It does not prove causation.

It is possible that some third factor, like the overall star formation rate or the galactic shear, drives both the cloud lifetime and the galaxy mass. The authors control for some of these variables, but not all.

Also, the study only examines main sequence galaxies, meaning galaxies that are actively forming stars at a normal rate for their mass. It excludes the centers of galaxies, where the environment is extreme and the physics might be different. And it excludes quiescent galaxies that have stopped forming stars entirely.

So the results apply to the disks of normal star forming galaxies. That is a large and important population, but it is not the whole universe.

Another limitation: the timescales are statistical averages. They tell us about populations of clouds, not individual clouds. A single cloud might live for 50 million years or 2 million years. The average is 5 to 30 million years, but the spread is real.

The PHANGS Survey: A New Way of Seeing Galaxies

This paper is part of the PHANGS survey, which stands for Physics at High Angular resolution in Nearby GalaxieS. It is a collaboration that combines ALMA, Hubble, and the James Webb Space Telescope to study star formation at the scale of individual clouds.

The sample of 54 galaxies is unprecedented. Previous studies looked at one or two galaxies. The Milky Way, the Large Magellanic Cloud, M33. Each gave a snapshot. PHANGS gives a family portrait.

By looking at so many galaxies, Kim and his team could separate the effects of local environment from global galaxy properties. They found that cloud lifetimes correlate with galaxy mass, but also with the local gas surface density and the velocity dispersion of the clouds.

The result is a more nuanced picture. Cloud evolution is not determined by a single parameter. It is a complex interplay of gravity, turbulence, feedback, and metallicity.

The Role of Turbulence

One of the key timescales in the paper is the turbulence crossing time. This is the time it takes for a turbulent eddy to cross the cloud. In molecular clouds, turbulence provides support against gravity. Without it, clouds would collapse in a free fall time, which is only a few million years.

But clouds live for 1 to 3 turbulence crossing times (Kim et al., 2022). That means turbulence is not just a passive feature. It actively regulates the cloud's lifetime.

In environments with high turbulence, clouds might live longer because the turbulent support prevents rapid collapse. In quieter environments, clouds collapse faster. But the data show the opposite trend: clouds in low mass galaxies, which tend to have lower turbulent velocities, live shorter lives.

This suggests that other factors, like metallicity and feedback, dominate over turbulence in setting the cloud lifetime. The turbulence is important, but it is not the whole story.

Why This Matters for Galaxy Evolution

The lifecycle of molecular clouds is not an obscure detail. It is the central mechanism by which galaxies turn gas into stars. If you want to understand why some galaxies are blue and star forming while others are red and dead, you need to understand how long clouds live and how efficiently they form stars.

The Kim paper provides the first systematic measurement of these timescales across a large sample. The numbers are now in hand. The next step is to build models that reproduce them.

For theorists, this is a challenge. Most simulations of galaxy formation use subgrid models for star formation. They assume that a certain fraction of the gas turns into stars per free fall time. But the data show that the efficiency is low, 1 to 8 percent, and that the timescales vary with environment.

A good model should predict these variations. If it does not, it is missing something important.

What This Actually Means

• ▸Clouds are short lived. A typical molecular cloud in a star forming galaxy exists for 5 to 30 million years. That is about 0.1 percent of the age of the galaxy. Clouds are not permanent structures. They are transient phenomena, constantly forming and dissolving.

• ▸Star formation is inefficient. Only 1 to 8 percent of the gas in a cloud actually turns into stars. The rest is blown away by feedback. This means galaxies are wasteful star makers. They could form more stars, but they do not, because the stars themselves shut down the process.

• ▸Small galaxies are different. In low mass galaxies, the molecular gas is harder to see in CO. The clouds we do see live shorter lives. This biases our view of star formation in dwarf galaxies. We might be missing most of the action.

• ▸Feedback acts fast. Once a massive star becomes visible, it destroys its birth cloud within 1 to 5 million years. That is almost instantaneous on galactic timescales. Feedback is not a slow process. It is violent and quick.

• ▸Environment matters. Cloud lifetimes depend on where the cloud lives. In busy, star forming regions of massive galaxies, clouds last longer. In the outskirts of small galaxies, they die young. The local conditions, not just the cloud's internal properties, set the clock.

The universe is not a uniform factory. It is a patchwork of different environments, each with its own rules. Molecular clouds obey those rules. They live, they form stars, and they die. And now, for the first time, we can measure how long each act takes.