Authors: Inayoshi, K. and Maiolino, R.
First Author’s Institution: Kavli Institute for Astronomy and Astrophysics, Peking University, Beijing 100871, China
Status: ePrint [open access]
Since its launch in late 2021, the James Webb Space Telescope has discovered all kinds of weird and wonderful objects. Its impressive sensitivity to infrared wavelengths has allowed astronomers to peer billions of years into the past and discover previously unseen populations of early galaxies. One distinct group of new galaxies were dubbed “Little Red Dots”, or LRDs for short and were observed to be red, compact galaxies with a distinctive “V” shaped spectrum (you can read more about LRDs here and here).
There’s been a lot of speculation about what kinds of galaxies LRDs might be. One of the most popular interpretations is that they’re galaxies hosting a supermassive black hole that’s being fed by a rapidly rotating disc of gas. These kinds of galaxies are known as active galactic nuclei or AGN. One of the key signatures of an AGN is the presence of broad Balmer emission lines in the galaxy’s spectrum.
Electrons in an atom can only inhabit specific energy levels, and to jump down from one level to another, they have to emit a photon with the exact same energy as the difference between the two levels. Each element has unique energy levels, allowing astronomers to attribute different emission lines to specific transitions within specific elements. Wavelengths in the Balmer series are emitted when an electron in a hydrogen atom jumps from a higher energy level to the second energy level.
But just having Balmer emission lines doesn’t tell us much – it just indicates that there’s hydrogen in the galaxy. Hardly surprising given that it’s the most common element in the universe! Normally, an emission line is narrow since light is being emitted at just one wavelength. But, if the gas is moving relative to the observer, then the Doppler effect kicks in, shifting light to different wavelengths. The breadth of the Balmer lines in LRD spectra can only be produced by a spinning disc of hydrogen (see Figure 1). From your point of view, one edge of the disc is moving towards you, and the other edge is moving away from you. As a result, light from the edge moving towards you will be blueshifted, and light from the edge moving away from you will be redshifted. Adding up the light from the entire disc results in a broad emission line, hence why a broad Balmer line is a hallmark of an AGN (although not all AGN are observed to have broad lines, you can learn more about these kinds of AGN here).
But there are other possible explanations for what LRDs might be. One explanation that’s garnered some attention is the possibility that LRDs are not AGN, and are instead very dusty starburst galaxies. This explanation is supported by the presence of a Balmer break (sometimes also called a Balmer jump) in the spectra of some LRDs. A Balmer break refers to a significant dip in a spectrum for wavelengths shorter than the Balmer limit – or, the maximum wavelength of light which can ionise a hydrogen atom with an electron in the second energy level. Observing a Balmer break means that a significant fraction of hydrogen atoms with electrons in or above the second energy level have been ionized by high-energy photons. Typically, a Balmer break is associated with recent star formation, since you need lots of hot stars to be emitting photons beyond the Balmer limit.
There’s a problem with this explanation of LRDs. If LRDs are in fact dusty starbursts, then their spectra are consistent with stellar masses of tens of billions (or even up to hundreds of billions) of solar masses. In the local universe, these kinds of masses are pretty normal, but local galaxies have had 13.8 billion years to grow – an LRD at redshift 7 has not even had 1 billion years to grow. Our current understanding of the universe makes it seem pretty unlikely that this could happen.
Luckily, today’s authors are on the case and have shown how an AGN spectrum could have a Balmer break, allowing astronomers to assume a much lower stellar mass for LRDs. The authors suggest that if LRDs are AGN covered by a thick blanket of dense gas, then we could expect to see a Balmer break.
To test this idea, the authors use a photoionization modelling code called Cloudy, which essentially calculates how many electrons should be in each energy level, given the temperature and density of the gas, as well as the light source illuminating the gas. The authors model the gas in the LRD and surrounding the AGN as a single slab of low-metallicity (10 times fewer heavy elements than in the Sun) gas at a uniform temperature and density and use an AGN spectrum as the light source. They vary the density of the gas between 10 million atoms per cubic cm and 100 billion atoms per cubic cm.
At low densities (see the magenta line in Figure 2), there’s no Balmer break because there just aren’t many electrons in the second energy level. As the density increases, collisions between particles become more common, and some electrons will excite to the second energy level due to these collisions. As a result, there are more electrons at the right energy level to absorb photons bluewards of the Balmer limit and photoionize. In Figure 2, you can see that the strength of the Balmer break increases as you go from 108 cm-3 to 1010 cm-3.
At the highest density tested by the authors (1011 cm-3, the yellow line in Figure 2), the strength of the Balmer break actually decreases. This is because the equilibrium temperature associated with this density is slightly lower (7800 K instead of 8000 K), resulting in less frequent collisions and fewer electrons in the second energy level.
You can see from Figure 3 that the authors’ simulated spectra produced Balmer breaks that are just as strong as the Balmer breaks seen in LRDs. This means that the picture of LRDs as AGN surrounded by very dense gas is consistent with observations! The authors also show that such dense gas can produce absorption features at the Balmer wavelengths and an oxygen emission line, which are also sometimes observed in LRD spectra.
Further observations are needed in order to definitively say what LRDs are, and it’s possible that not all LRDs are the same kind of object. The results of today’s paper show that we don’t have to invoke large stellar populations in order to understand Balmer breaks in LRD spectra, but Balmer breaks are only seen in 10-20% of broad-line AGN observed by JWST, so astronomers will need to understand the different physical scenarios that produce the full range of LRD spectra.
Astrobite edited by Storm Colloms
Featured image credit: NASA
Discover more from astrobites
Subscribe to get the latest posts to your email.
When you say “Typically, a Balmer break is associated with recent star formation, since you need lots of hot stars to be emitting photons beyond the Balmer limit”, I got confused by the timescale: in the first 100 Myr of star formation the spectra is dominated by OB stars, which do not have strong Balmer break. After OB stars A-stars enter the game, increasing the Balmer break. But these stars have timescale in the order of 1 Gyr. So if there is ‘recent’ star formation you won’t see the Balmer break, only after 100 Myr you should start to see the break, until A stars are alive. Does it make sense?
It’s true that OB stars do not intrinsically have strong Balmer breaks in their spectra. However, the hydrogen in the interstellar medium will absorb photons bluewards of the Balmer break, creating a break in the overall galaxy spectrum.