What created more light: the Big Bang or stars?

Although time might seem to pass slowly in some instances, it’s important to remember that our Universe has been around for a long time. It’s been 13.8 billion years since the hot Big Bang, and our entire cosmos has evolved by quite a large amount over that duration. As of right now, our cosmic vision extends for some 46.1 billion light-years in all directions, revealing somewhere between an estimated 6 and 20 trillion galaxies in the process. Among the typical large galaxies, there’s an average of hundreds of billions of stars contained inside; although most galaxies are small and low in mass, this still adds up to a cumulative 2 × 10²¹ stars.

Inside them, each star is made of perhaps an average of around 10⁵⁷ atoms. There’s a lot that’s happened in our Universe, but most of it — including the formation of most stars — is a part of our cosmic past, not our present or future. In fact, we can reconstruct the entire star-formation history of the Universe via a variety of methods, including by surveying the stars and galaxies found at all different epochs throughout cosmic history. One important piece of evidence that corroborates those estimates comes from the Fermi gamma-ray telescope, which in 2018 first measured the star-formation history of the entire Universe throughout all of cosmic time.

Fascinatingly, this type of measurement also empowers us to answer a longstanding question: what made more light, the Big Bang, or the cumulative amount of stars formed over cosmic history? The answer is the Big Bang, and it’s a wonderfully instructive story to find out how we know.

A spiral galaxy typically consists of four main gaseous regions within the disk: diffuse atomic gas, dense molecular gas, stars and star clusters, and ionized regions of matter arising from energy injections from star-forming regions, young stars, and stellar cataclysms. JWST, along with the other PHANGS data sources, helps reveal different aspects of this life cycle, but once a galaxy’s gas is gone and no new gas reservoirs fall inside, star-formation ends permanently.

When you form stars, there are a number of important processes that happen sequentially.

1. A molecular cloud of raw material, mostly hydrogen, will collapse under its own gravity.
2. During collapse, the cloud will fragment, giving rise to star systems and clusters of stars in short order, with star birth officially coinciding with the ignition of nuclear fusion in the cores of those stars.
3. Then, with so much high-energy (i.e., ultraviolet) radiation streaming out from these newborn stars, the molecules remaining in the surrounding cloud get ionized by that energetic radiation, stripping electrons off of their atoms.
4. Once the surrounding circumstellar medium gets ionized, a special type of radiation begins to appear: emission lines, as radiation is emitted when electrons fall back onto the ionized atomic nuclei and cascade down the various energy levels.
5. This starlight then travels through the Universe, including by streaming through intergalactic space, where they interact with all the atoms they encounter, resulting in additional absorption signatures imprinted onto that light.
6. And, finally, that starlight has a finite, non-zero probability of interacting with gamma-rays, which are the highest energy photons, to produce a specific species of new particles: electron-positron pairs.

This isn’t the only set of things that happens, but it leads to the consequence that starlight, indirectly, can eventually give rise to electron-positron pairs.

The production of matter/antimatter pairs (left) from two photons is a completely reversible reaction (right), with matter/antimatter annihilating back to two photons. This creation-and-annihilation process, which obeys E = mc², is the only known way to create and destroy matter or antimatter. If high-energy gamma-rays collide with normal photons from starlight, there is a chance to produce electron-positron pairs, diminishing the gamma-ray flux observed at great distances.

This last point is of particular interest to anyone with a space-based gamma-ray telescope, because when positrons stream through space, they eventually collide with electrons, where they’ll re-annihilate. When they do, they produce photons of a very specific energy, 511 keV, which happens to be a telltale gamma-ray signature of their presence. Now, positrons aren’t the only source of gamma-rays in the Universe, and this poses a challenge for observational astronomers who work on high-energy signatures. For example, there are classes of objects in the Universe — such as active, supermassive black holes — that are very good emitters of extremely energetic particles, including gamma-rays.

With enormous event horizons and large, massive accretion disks surrounding them and infalling onto them as they feed, the material surrounding supermassive black holes, charged particles, generate tremendous magnetic fields as they rotate. These magnetic fields accelerate the charged particles, causing them to interact and emit radiation of extremely high energies. The brightest ones of all, as far as our viewpoint here on Earth is concerned, are the ones whose relativistic jets are pointed right at us. These objects are known as blazars, because they “blaze” down the line-of-sight right toward your eyes.

In this artistic rendering, a blazar is accelerating protons that produce pions, which produce neutrinos and gamma rays. Photons of all energies are also produced. Extreme events in energy are generated by processes occurring around the largest supermassive black holes known in the Universe when they’re actively feeding, as well as around highly magnetic neutron stars and actively feeding stellar mass black holes.

What do blazars have to do with understanding the total amount of starlight, cumulatively, that’s ever been produced in the Universe?

Hang in there; we’re going to get there. There are plenty of interesting features that are emitted from blazars, but in order for them to arrive at our eyes (or instruments), those signals also have to contend with the intervening matter that’s in the way. Whenever you look at anything in the distant Universe, you have to recognize that gas clouds exist, absorbing a fraction of the light; we can account for those by examining absorption lines. There are also galaxies and clusters of galaxies that often intervene; we can measure their brightnesses, densities, and other properties to calibrate each individual blazar we examine. Blazars are located all over the sky, where zodiacal effects from the Solar System and foreground effects from the Milky Way can affect what we see. And each individual blazar, at the source, will have properties of energy and flux that are intrinsically unique to it.

By doing the proper accounting of what exists in the Universe — at the source, along the line-of-sight, and right at the point where we received the light at long last — we can determine the source properties of the blazar we’re examining. That’s a lot of legwork, but in the end, the reward is worth it: we’ll have a well-calibrated starting point to work from.

This illustration of an active galactic nucleus shows the characteristic relativistic jets emanating from the central supermassive black hole. If our line-of-sight happens to be located along the axis of one of these jets, the object will appear as a blazar to us: one of more than 700 presently known.

Now, at last, we can take the fact that we have not just a gamma-ray telescope, but a gamma-ray telescope (Fermi) with all-sky coverage. There’s now a method for measuring all the starlight in the Universe, all put together. Here’s the way to do it.

• First, start by measuring and identifying all the blazars, all throughout the Universe, everywhere that they can be found.
• Then measure the redshift of each blazar, which can be known by measuring at least one absorption or emission line within it, so that you have a good proxy for the distance that it lies from you.
• Once you have that information, then this is where the gamma-rays come in: you measure the number of gamma-rays received by your gamma-ray telescope as a function of two key properties of each blazar, the blazar’s redshift and the blazar’s brightness.
• Remember that gamma-rays, when they collide with the cumulative amount of extragalactic, background starlight, have a certain (starlight-dependent) probability to produce electron-positron pairs.
• And then, at last, instead of looking for that characteristic energy signal of electron-positron annihilation, that 511 keV signal (redshifted from the expansion of the Universe by the time that Fermi sees it), instead use the gamma ray deficit from the source (because gamma rays are being converted into electron-positron pairs) to calculate how much background starlight must be present, as a function of redshift/distance, to account for the loss of gamma rays.

NASA’s Fermi Satellite has constructed the highest resolution, high-energy map of the Universe ever created. Without space-based observatories such as this one, we could never learn all that we have about the Universe, nor could we even accurately measure the gamma-ray sky. The evidence for large quantities of clustered antimatter is nonexistent, but a whopping 739 blazars are identifiable in Fermi’s gamma-ray sky.

Only once the blazars have been properly calibrated, and found at a wide enough variety of distances and along a large enough numbers of lines-of-sight, can we estimate the total amount of starlight produced at each step throughout cosmic history.

This was something that, up until recently, we would’ve been unable to execute. But with the advent of NASA’s Fermi satellite, and in particular, the Fermi-LAT collaboration (where LAT is the Large Area Telescope instrument aboard Fermi), we were able to make these critical measurements, for the first time, for all of the known blazars that appear in the gamma-ray sky: all 739 of them. Blazars aren’t exactly nearby; the closest one that we know of is already more than 200 million light-years away, which means that its light that’s arriving now was emitted from it more than 200 million years ago. These objects are usually found at great distances, with the light from the most distant blazar included in their survey arriving after a journey of 11.6 billion years: from when the Universe was just 2.2 billion years old.

Because of how these Blazars are distributed in space and (lookback) time, we have one more piece of the puzzle to throw in there: to model when the Universe transitions from being opaque to transparent in gamma rays, which the Fermi-LAT team was able to do as part of this work.

This illustration shows a dust-obscured active galactic nucleus, similar to what must exist at the heart of galaxy Messier 77, located an estimated 47 million light-years away. Messier 77 was the first non-blazar, non-supernova to be identified as an extragalactic neutrino source. Blazars are all far more distant, and become transparent to gamma-rays only at late times in cosmic history.

Previous work had been able to measure and/or estimate the cumulative amount of starlight directly: by charting out the star-formation history of the Universe through direct measurements of galaxies and their internal star-formation rates. However, it’s been a challenge to measure, precisely:

• exactly where and when star-formation peaks,
• how significantly star-formation has declined from its peak to the present day,
• and at what rate star-formation occurred in the lead-up to its peak, which arguably has the largest uncertainties.

That’s where the Fermi data comes in and plays such a critical role. Not only were the net results found by this key 2018 study in agreement, within the errors and uncertainties, with all previous studies, but they were vastly able to improve the precision to which we could know these parameters of the Universe’s history. The Universe had its star-formation rate peak when it was approximately 3 billion years old, and the star-formation rate has been declining ever since. Today, the modern star-formation rate sits at just 3% of that early, maximal rate, and the rate that we’re forming new stars in the Universe is still continuing to fall off over time.

The Fermi-LAT collaboration’s reconstructed star-formation history of the Universe, compared with other data points from alternative methods elsewhere in the literature. We are arriving at a consistent set of results across many different methods of measurement, with the greatest uncertainties persisting at the highest redshifts and earliest times. These uncertainties represent less than a 1% uncertainty in the total number of stars formed throughout cosmic history.

But one interesting and novel result to come out of this study really is revolutionary. According to the lead author of the Fermi-LAT study, Marco Ajello:

“From data collected by the Fermi telescope, we were able to measure the entire amount of starlight ever emitted. This has never been done before.”

That’s right: for the first time ever, we’ve been able to measure the entire amount of starlight emitted throughout the history of the Universe. This leads to an estimate for many other important parameters, including:

• the total number of stars that have been birthed throughout cosmic history,
• the number of stars that existed (or had been cumulatively formed) at any moment in our cosmic past,
• and the total number of photons that have been created, due to stars, since the onset of the hot Big Bang.

That last parameter is the key value we need to compare the total number of photons, or quanta of light, that were created by stars with the total number that were created in the aftermath of the hot Big Bang.

The star-formation rate in the Universe is a function of redshift, which is itself a function of cosmic time. The overall rate, (left) is derived from both ultraviolet and infrared observations, and is remarkably consistent across time and space. Note that star formation, today, is only a few percent of what it was at its peak (between 3-5%), and that the majority of stars were formed in the first ~5 billion years of our cosmic history. Only about ~15% of all stars, at maximum, have formed over the past 4.6 billion years. Direct measures of star-formation are important, but the method of Fermi-LAT for measuring the total number of photons produced by stars is superior.

The key result, or the total amount of starlight that was created since the start of the hot Big Bang? It corresponds to a total of approximately 4 × 10⁸⁴ photons, which is an amazingly large number. Remember how we said, earlier, that there are around 10⁵⁷ atoms in each star, and around 2 × 10²¹ stars in the entire Universe? It turns out that there are even more protons, neutrons, and electrons than what we find in stars, and that there are cumulatively around ~10⁸⁰ of each of them within our observable Universe. Even so, even with all of those particles, that means the number of photons produced in the form of starlight within our Universe is literally thousands of times larger than all the protons, neutrons, and electrons present in our Universe combined.

How does that number, though, compare to all the photons that exist in the Universe as part of the leftover radiation from the Big Bang?

That answer comes most directly from measurement of the cosmic microwave background: the leftover glow from the Big Bang. Two of the key pieces of information that come from our observations of that are:

• the photon-to-baryon ratio, which the cosmic microwave background tells us is about 1.6 billion photons for every baryon (i.e., a proton or neutron) in our Universe,
• and from direct measurements of the photon number density itself, which comes out to 411 leftover Big Bang photons per cubic centimeter of space, at present.

If you put all of this information together, you find that the total number of photons that are left over from the Big Bang, within our observable Universe today, is approximately 10⁸⁹-to-10⁹⁰: hundreds of thousands times as many photons as stars have ever created!

At any epoch in our cosmic history, any observer will experience a uniform “bath” of omnidirectional radiation that originated back at the Big Bang. Note that the CMB isn’t just a surface that comes from one point, but rather is a bath of radiation that exists everywhere at once. As each new year passes, the CMB cools down further by about 0.2 nanokelvin, and in several billion years, will become so redshifted that it will possess radio, rather than microwave, frequencies.

Still, it brings up a fascinating cosmic coincidence. The average energy of the photons that arise from starlight is much larger, today, than the amount of energy in a typical Big Bang photon that’s left over: by approximately a factor of 10,000 or so. As space continues to expand, the leftover glow from the Big Bang will continue to fade, as its wavelength will be stretched to greater and greater values. Meanwhile, new stars, even at a low rate, will continue to form, where they’ll continue to produce photons of the same energy that all stars produce today. Even though the number of starlight photons will never equal the number of Big Bang photons, it’s possible that the energy produced by all the stars, in terms of radiation, will someday approach and may even equal the amount of energy locked up in the form of photons from the Big Bang itself.

Now, with the advent of observatories that specialize in other wavelengths, like ALMA at radio wavelengths or JWST at infrared wavelengths, we’re beginning to probe star-formation at distances that are even greater and at times that are even earlier than what Fermi-LAT is capable of seeing. Star-formation, after all, is the process that turns the primordial elements from the Big Bang into the elements capable of giving rise to rocky planets, organic molecules, and life in the Universe. Even though the number of photons, and the amount of light, produced by stars is still just a tiny fraction of the amount of light produced in the Big Bang, the amount of energy in the form of starlight now almost equals the cumulative energy found in the form of leftover Big Bang photons. At long last, what was a decades-old scientific question has now been answered by definitive measurements.