Fomalhaut means “mouth of the fish” in Arabic and it is the brightest star in the constellation Pisces Astrinus. Fomalhaut is an “A star” which means that it is significantly more massive and luminous than the sun. It is also younger than the sun, with an age of approximately 400 million years (the sun is 4.6 billion years old).  Two other, less-massive stars are physically linked to Fomalhaut: TW PsA and LP 876-10.  These stars are very far from Fomalhaut at the present time.

Brief history of my research

I have been searching for a planetary system around Fomalhaut since 1993 with optical ground-based observations published in Kalas and Jewitt (1996).

In a 2005 Nature paper I was the first to succeed in directly imaging light reflecting from a vast elliptical dust belt produced by the collisional erosion of comets and asteroids orbiting Fomalhaut.  The high resolution Hubble Space Telescope images revealed that the belt has a very sharp inner edge and the entire belt is offset from the star, the first time in astrophysics that such an effect had been observed.  Both properties supported a hypothesis that a planet on a non-circular orbit resided within the belt, gravitationally shaping the belt.

In a 2008 Science paper I announced the discovery of a point source called Fomalhaut b located just inside the inner edge of the dust belt. Comparing its position between the 2004 Hubble data and the 2006 Hubble data showed orbital motion, the first time in astronomy where the orbital motion of a planet outside our solar system was directly imaged. The animation below shows that how we have imaged Fomalhaut b moving through the system over time.

Animation made by Jason Wang (Caltech) using several years of my Hubble Space Telescope observations of Fomalhaut.

I published a detailed study of Fomalhaut in a 2013 Astrophysical Journal paper. This new research had the following findings:

  • Fomalhaut b has a highly eccentric orbit, e ~0.8.  This is not unusual given that many other exoplanets discovered by other techniques have similar, highly elliptical orbits.
  • Fomalhaut b’s closest approach to the star (periastron) is approximately 30 au and the orbital period is roughly 1,700 years.
  • Even though the elliptical path of Fomalhaut b appears to cross through the belt in the future, its orbital plane is likely 17 degrees different from the dust belt’s plane.  In other words, Fomalhaut b may not physically pass through the belt.
  • Fomalhaut b may be a planet like Saturn with a large ring system that reflects visible light.   The central planet would be too faint to be detectable by current instrumentation in the infrared.  However, Fomalhaut b could be the first detection of an extrasolar dwarf planet with a ring system.  Comparisons were made to the hypothetical Charon-forming giant impact of Pluto, and the Haumea collisional family where a dwarf planet would be surrounded by a significant amount of material in a cloud or ring.

Interesting questions about Fomalhaut b

Does Fomalhaut b exist?

In the past, some astronomers stated that Fomalhaut b does not exist because it has not been detected at infrared wavelengths.  They implied that what we see in the Hubble images is pure noise and also argued that if it passes through the belt it would destroy the belt.  It is very clear these statements were misleading because:

  • The absence of evidence is not evidence for absence. The absence of an infrared signal does not mean the object is not there.  Instead, every observation has limitations to its sensitivity.  The non-detections in the infrared merely establish an upper limit to Fomalhaut b’s mass—it is less massive than our own Jupiter.  However, a Saturn-mass would not produce an infrared signature strong enough to be detected by current methods.
  • The Hubble data were independently reduced, analyzed and published by two expert teams (Galicher et al. 2013 and Currie et al. 2012) and they confirmed the existence of Fomalhaut b in the Hubble data.  In other words, including my effort, the existence of Fomalhaut b has been confirmed three times over.
  • As discussed above, my 2013 paper shows that the orbital plane of Fomalhaut b is probably not the same as the belt plane and it would not pass through the belt.

Is Fomalhaut b a planet surrounded by a large dust ring or shroud, or is Fomalhaut b purely a puff of dust created by a cosmic collision between two asteroids?

Both are possibilities that have been studied multiple times by several groups.  In principle, the puff-of-dust model has the problem that a cosmic collision happens rarely and the puff of dust lasts a short time, which means it is highly unlikely to be observed.  Circumplanetary dust rings, on the other hand, last a long time, and in fact we can visually see the rings around Saturn, and even the dwarf planet Chariklo has a ring around it.

Here is a brief review of some of the literature:

  • The pure dust cloud hypothesis was first proposed in Kalas et al. (2008) and deemed the least likely explanation (but not ruled out).  The circumplanetary ring/shroud hypothesis was also proposed and quantified as a large ring around a planet.  The illustration above was created in 2008 to depict what we thought Fomalhaut b might look like up close.  A very large ring of dust surrounds the planet.
  • Kennedy and Wyatt (2011) tested and affirmed the feasibility of a circumplanetary ring/shroud model.
  • Currie et al. (2012) wrote the following in italics in their paper: “Unless we are fortuitously (and implausibly) identifying a very recent collision…Fomalhaut b is not an unbound dust cloud.”
  • Janson et al. (2012) reasoned that the brightness variability reported in Kalas et al. (2008) favored the puff-of-dust model.  However, subsequent research did not confirm these measurements of variability (see below).
  • Galicher et al. (2013) studied the dust cloud model and the circumplanetary ring/shroud model, concluding both were feasible.
  • Kenyon et al. (2014) and Lawler et al. (2015) studied the theory of a planetesimal collision creating a dust cloud, the former saying it is an unlikely event, the latter saying it is likely.
  • Kenyon et al. (2014) also studied the model of a circumplanetary ring/shroud, stating that this “…is the simplest mechanism for dust production in Fomalhaut b.”
  • Tamayo (2014) studied both types of model, concluding the dust cloud model is highly unlikely and: “The best explanation found in this paper is that Fomalhaut b is a super-Earth hosting a large population of irregular satellites…”

Is Fomalhaut b extended (larger than a point source)?

Multiple independent groups (Kalas et al. 2008, Currie et al. 2012, Galicher et al. 2013, Kenyon et al. 2014) have analyzed the highest resolution and highest signal-to-noise detection of Fomalhaut b (HST/ACS/HRC/F606W from 2006) and the measurements show Fomalhaut b is not extended.

With an injection and recovery experiment, Kalas et al. (2013) showed that the lower signal-to-noise HST/STIS detections of Fomalhaut b could make the source appear extended.  This experiment showed that an unfortunate placement of Fomalhaut b in a region of high local noise could even make Fomalhaut b completely undetectable.

Fomalhaut b appears extended in the HST/ACS/HRC/F814W images, yet these data are lower signal to noise than HST/ACS/HRC/F606W.  Also, Sirianni et al. (2005) discovered that a halo appears around point sources when they are imaged at wavelengths greater than 0.75 microns.  The F814W filter is centered at 0.814 microns and therefore it is plausible that Fomalhaut b appears extended in these observations because of a known instrumental property (i.e., it is not astrophysical).

In Kalas et al. (2013) I reported that Fomalhaut b appears elliptical in shape with a width of approximately 3 au in data obtained in 2012 with HST/STIS.  However, I suspected this could result from local noise and conducted an inject and recovery experiment.  I injected an artificially created point source (that had the same brightness as Fomalhaut b) at various locations in the data  and discovered that the resulting images could appear extended along various directions depending on the local noise.  Below is Figure 6 from Kalas et al. (2013).  I concluded that Fomalhaut b appears elliptical in the data because of noise.  As shown in the figure, the brightness of Fomalhaut b can vary a large amount as well, and in 9 injection locations Fomalhaut b is not detectable because of the noise.

Below is an image that shows the artificial point sources that were originally implanted to make the simulation above.

Is the light from Fomalhaut b variable?

In Kalas et al. (2008) I reported my measurements of Fomalhaut b’s brightness, stating: “…between 2004 and 2006 Fomalhaut b became fainter by ~0.5 mag at 0.6 microns.”  Currie et al. (2012) and Galicher et al. (2013) analyzed the same data with independent techniques and found no statistically significant change in brightness.  Galicher  et al. (2013) reported photometry showing that Fomalhaut b may have faded at the 1.7-sigma level between 2004 and 2006, but this is not statistically significant.  In Kalas et al. (2013) I reviewed the aforementioned photometry and concluded, “Since the observations are the same, these results suggest that there are systematic photometric uncertainties due to the choices of data reduction and analysis methods for high contrast imaging.”

I have detected Fomalhaut b at three more epochs using HST/STIS.  However, the inject-and-recovery experiment above shows that local noise has a significant impact on measuring the brightness of Fomalhaut b.  No statistically significant change in brightness has been measured so far (work in progress shown below).

We find no significant change in the brightness of Fomalhaut b between 2010 and 2013. This is the measured brightness of Fomalhaut b in different images from the Hubble Space Telescope obtained in 2010, 2012, and 2013. With each image in each year we use four measurement techniques. Fomalhaut b is very faint and the measurements have significant uncertainty.

To sum up, Fomalhaut b is a very faint object next to one of the brightest stars in the sky.  We have detected Fomalhaut b multiple times and can track its orbit, but there is no reliable evidence for variability.

What’s new with Fomalhaut b?

We have imaged Fomalhaut b in 2013 and 2014.  The 2014 observation was half as long as the 2013 observation and the images are too noisy for a 2014 detection.  How noisy is too noisy?  Below we inject an artificial source as bright as Fomalhaut b into the 2014 data.  The experiment asks, “If Fomalhaut b existed in these data, would we be able to detect it?”  The red squares show the injection points, with the first injection into the upper left square.  As the movie plays, the injections go from left to right.  The 2014 data are so noisy, that only half the injections lead to a detection of Fomalhaut b.  Even when Fomalhaut b is detected, its shape, location, and brightness can change dramatically.  For example, Fomalhaut b can appear to be up to 2 pixels away from the point that it was injected into.

However, we have a very nice detection in the 2013 observations, as shown below, using two different methods to process the data.  Fomalhaut b is not found to be extended (it looks like a point source) and it is not variable with respect to the 2010 and 2012 observations made with the same instrument (HST/STIS).  We hope that our pending request for future Hubble observations in 2021-2022 will provide sensitive data greatly needed to ascertain the nature of Fomalhaut b.



References to the scientific research literature:

Currie et al. (2012)

Faramaz et al. (2015)

Gallicher et al. (2013)

Janson et al. (2012)

Janson et al. (2015)

Kalas and Jewitt (1996)

Kalas, Graham, and Clampin (2005)

Kalas et al. (2008)

Kalas et al. (2013)

Kennedy & Wyatt (2011)

Kenyon et al. (2014)

Lawler et al. (2015)

Sirianni et al. (2005)

Tamayo (2014)


The latest paper on Fomalhaut b was published in April 2020.  Gaspar & Rieke (2020) suggest that the HST data from 2004 to 2014 are consistent with the planetesimal collision hypothesis.  Their analysis is at variance with my results as well as that of theory.  For example, the 2010-2014 STIS data are too noisy to show that Fomalhaut b is fading and expanding in size.  As far as theory goes, they assume that the volume number density of planetesimals that could collide with each other is that of the dust belt.  However, Fomalhaut b was first discovered at a location approximately 15 au (2.25 billion kilometers) away from the dust belt.


© 2020 Paul Kalas . Powered by WordPress. Theme by Viva Themes.