No Rayleigh scattering gives yellow skies to exoplanet WASP-79b

Here’s a catch-up on a press release recently put out by NASA, Hubble and Johns Hopkins University, who led an analysis of WASP-79b. Lead author of the paper, Kristin Sotzen, combined spectroscopy from the ground-based Magellan II telescope in Chile with data from the HST and Spitzer satellites.

As explained in the press release: “The surprise in recently published results, is that the planet’s sky doesn’t have any evidence for an atmospheric phenomenon called Rayleigh scattering, where certain colors of light are dispersed by very fine dust particles in the upper atmosphere. Rayleigh scattering is what makes Earth’s skies blue by scattering the shorter (bluer) wavelengths of sunlight. Because WASP-79b doesn’t seem to have this phenomenon, the daytime sky would likely be yellowish, researchers say.”

“This is a strong indication of an unknown atmospheric process that we’re just not accounting for in our physical models.” said Sotzen.

WASP-79b also was observed as part of the Hubble Space Telescope’s Panchromatic Comparative Exoplanet Treasury (PanCET) program, and those observations showed that there is water vapor in WASP-79b’s atmosphere. Based on this finding, the giant planet was selected as an Early Release Science target for NASA’s upcoming James Webb Space Telescope.

The press release has led to national media coverage in the US and the UK, including by The Sun and Fox News.

A circumbinary planet transiting a WASP binary star

Circumbinary planets are planets that orbit a binary star, rather than a single star, in contrast to most planets known. Possibly they formed out of the same disk of material that formed the two stars. So long as the planet is sufficiently far from the binary, its orbit can be stable.

Several circumbinary planets were discovered by the Kepler mission, but the first such planet found using TESS has recently been announced by Veselin Kostov et al.

Here is the TESS lightcurve of “target of interest” TOI-1338:

Six-percent-deep eclipses recur every 14.6 days, when a small, M-dwarf star eclipses its Sun-like companion. But, when one looks more closely at portions of the lightcurve, one also sees:

Near Day 1483 is a deep eclipse (truncated in this plot). Near Day 1577 is a shallower secondary eclipse, when the Sun-like star eclipses the M-dwarf star. But, in addition, and highlighted by green bands, are two transits where a Saturn-like planet in a wide, 95-day orbit, transits the Sun-like star.

The interest, from the WASP point of view, is that the eclipsing binary had already been found by the WASP survey. The primary eclipses were too deep to be caused by a Jupiter-size planet, and were instead found to be caused by the M-dwarf, but such systems are interesting in their own right, and thus formed the “EBLM” programme offshoot from WASP, led by Amaury Triaud.

Since the eclipsing binary was already known to be an interesting system, this meant it could be observed with 2-min cadence in the TESS survey (instead of the default 30-min cadence data), and that made the detection of the circumbinary-planet transits much easier.

Above is an illustration of how the planet’s orbit precesses around the inner binary over time. The inner binary is illustrated in black and grey. The coloured dots show how the orbit of the planet changes over 2000 days. When the planet passes within the grey ellipses, transits can be seen.

The two planets of WASP-148

Even though WASP has found nearly 200 planets we are still announcing systems that are unlike any previous ones. WASP-148 is an example, as described in the discovery paper by Guillaume Hébrard et al.

WASP first detected transits of the hot Jupiter WASP-148b in an 8.8-day orbit. Spectroscopic observations with OHP/SOPHIE, aimed at measuring the planet’s mass, then found that there was also a second massive planet in a longer, 35-day orbit:

The orbits of both planets are eccentric, likely because they are perturbing each other by their gravitational attraction. Further, the gravitational perturbations mean that the transits of the inner planet vary in time by 15 mins.

We don’t yet know whether the outer planet, WASP-148c, also transits (since its longer period means that there are gaps in WASP’s coverage of its orbit), but this patch of sky is currently being observed by the TESS satellite. The space-based photometry from TESS will be good enough to detect any transits of WASP-148c, to map out transit-timing variations, and to look for additional planets in the system that are too low mass to have been detected in the radial-velocity data. WASP-148 is thus an important system for studying an unusual planetary-system architecture, with two massive planets in relatively close orbits in resonance with each other.

Detecting helium envelopes around WASP planets

A new paper by Shreyas Vissapragada and colleagues reports a new technique for detecting material boiling off hot-Jupiter exoplanets. The idea is that helium atoms in escaping material should be strong absorbers of light at the wavelength of 1083.3 nm, one of the transitions of neutral helium. Thus, if one records a transit in an ultra-narrow-band filter around that wavelength, the planet should look bigger and so the transit should be deeper.

Vissapragada et al pointed the 200-inch Hale Telescope at a transit of WASP-69b. Here’s the result:

The blue line is the usual transit depth expected in continuum light. The data and fitted red line are the transit observed in the 1083.3-nm helium line. The authors compute that the extra depth of the transit implies that 30 million kilos of material is evaporating off the planet each second, as a result of stellar irradiation. This sounds a lot, but adds up to only a few percent of the planet’s mass over the host star’s lifetime.

Aluminium oxide in the atmosphere of hot-Jupiter WASP-43b?

WASP-43b is one of the favourite planets for atmospheric characterisation, being in such a tight, short-period orbit that it is heated up by its host star, such that the molecules in its atmosphere should be easy to discern.

A new paper by Katy Chubb et al re-analyses observations made with the Hubble Space Telescope and concludes that the observations show signs of aluminium oxide.

It is important to realise that this work is not easy, being right at the limit of what can be done, even with Hubble. Neither the spectral resolution nor the signal-to-noise ratio of the data are sufficient to unambiguously discern features of particular molecules. Instead, the art is to guess the molecules that might be present, simulate the resulting spectrum if the guess were right, and then compare that to the observed spectrum. This leads to figures such as this, from Katy Chubb’s paper:

The grey lines are the data (shown as error bars). The coloured lines are the calculated model (with the coloured bands then allowing for uncertainties), and the grey diamonds are where error-free data would be if the model were perfect. The x-axis is wavelength, and the y-axis is the effective radius of the planet’s atmosphere at that wavelength, which tells us how good it is at absorbing light of that wavelength.

The bottom panel (orange) fits the data with water vapour only, while the upper panel (blue) includes both water and aluminium oxide. The later gives a significantly better fit. The authors write that, in addition to water, “AlO is the molecule that fits the data to the highest level of confidence”, while “We find no evidence of the presence of CO, CO2, or CH4“.

However this could be a puzzle, since: “AlO is not expected from the equilibrium chemistry at the temperatures and pressures of the atmospheric layer that is being probed by the observed data. Its presence therefore implies direct evidence of some disequilibrium processes with links to atmospheric dynamics.”

As with all current characterisation of exoplanet atmospheres, we await the James Webb Space Telescope (which has been designed to do this work; Hubble was designed before exoplanets were even known), to tell us how reliable the current results are.

WASP-4 is accelerating toward the Earth

Here is a plot of the timings of the transits of WASP-4b, taken from a new paper led by Luke Bouma:

The curve in the plot shows that the transits are occurring progressively earlier as time passes. One possible explanation is that the planet’s orbit is decaying under the influence of the tidal interaction between the star and planet. This is expected to occur in most hot Jupiters, though how quickly is debated.

However, Bouma have also obtained radial-velocity observations of the system, which show that the star is accelerating towards us. This can result from it being in a wide orbit with another object (the authors suggest a wide-orbiting companion of 10-to-300 Jupiter masses at a distance of 10-to-100 AU). Since the system is accelerating towards us, the light-travel time is decreasing, and this (not orbital decay) means that the transits occur earlier.

Wide companions are expected in hot-Jupiter systems, since, in most theories for the occurrence of hot Jupiters, the gravitational perturbation of a distant companion is needed to shrink the hot-Jupiter orbit down to the current values of only a few days.

Bouma et al recommend continued radial-velocity monitoring of hot Jupiters in order to distinguish orbital decay from accelerations caused by orbiting companions.

The morning and evening terminators are different

Hot Jupiter exoplanets are “phase locked” by tidal forces, meaning that the same face of the planet always faces the star. Being blasted by radiation it is far hotter than the night side. This means that strong winds must be racing around the planet, redistributing the heat.

And that means that the “evening” terminator (where winds flow from the hot day-side face to the cooler night side) will be much hotter than the “morning” terminator (where winds flow from the night side to the day side). Here’s an illustration from a new paper by Ryan MacDonald, Jayesh Goyal and Nikole Lewis:

Of course the terminators are exactly the regions of the planet’s atmosphere that are being sampled by atmospheric-characterisation studies, since that’s the regions that are seen projected against the host star.

As Ryan MacDonald et al point out, most atmospheric-characterisation studies assume that the two limbs are the same, since that’s the easiest thing to do. However, the authors argue, while doing that might produce an acceptable fit to the data, the resulting parameter values could be very wrong.

Thus, the fitted temperature profile could be “hundreds of degrees cooler” than reality. As a result, the fitted abundances of molecular species could also be wrong. MacDonald et al conclude that: “these biases provide an explanation for the cold retrieved temperatures reported for WASP-17b and WASP-12b” and say that: “to overcome biases associated with 1D atmospheric models, there is an urgent need to develop multidimensional retrieval techniques”.