The transit method of detecting extrasolar planets has been a well used and successful tool, yielding many a planet around other stars. Recent years have seen better equipment recording higher quality transits, and with them far more detail that take us beyond the original ideas of transit theory. In a paper now on arXiv, David Kipping discusses how the depths of transit curves in the infrared can vary due to atmospheric effects complicating the original theory of transits.
The original theory is quite simple, the planet blocks quite a bit of light from the host star as it passes between us and it. The light blocked is determined by the apparent size of the disc of the planet blocking it, which depends on the planet’s radius and it’s distance from the star. The result was a graph of light emitted from the star against time that showed a line interrupted by regular flat bottomed, sloping sided dips as the planet got in the way.
Initial results from Kepler showed that the line wasn’t quite as flat as originally believed. As the planet moved round the star it exhibited phases as the dayside reflected light. The phase went from ‘new Moon’ when the planet was doing its blocking, through crescent and gibbous close to ‘Full Moon’, but not quite as the planet would then vanish behind the star. This means the flat line actually exhibits a rise in recorded light, followed by a drop whilst the planet is behind the Sun, then a return to the pre-drop amount and a decline before the main dip.
The planet itself also doesn’t present a solid black disc. Any space based imagery of the curve of the Earth will show a thin layer of translucent gas – our atmosphere – glowing as light is transmitted through it. The type of planets that are most readily visible through the transit method are ‘Hot Jupiters’ – large gas planets close in to the star. Their size means they block a lot of light relative to others, their proximity means they swing round the star quickly, meaning astronomers don’t have to wait years to record their repeated dips. Being gas, their atmospheres are rather larger than ours, and they are translucent for a larger fraction of their radius. This also affects the transit curve graph.
Information can be got from light reflected by or transmitted through a gas. Spectroscopy is the study of the amount of light emitted at different wavelengths of light (blue has a different wavelength to red etc) and other electromagnetic radiation photons – gamma rays, x-rays, ultraviolet, infrared, microwaves, submilimetre and radio waves. The full electromagnetic spectrum (including the visible rainbow of light between UV and IR) extends from the highly energetic and short wavelength gamma rays to the longer wavelength and low energy radio waves. Each and every material – gas, liquid, solid – has the capability to emit or absorb a selection of lines cut out from this spectrum. This is referred to as the spectrum of that material, or it’s spectral signature or fingerprint. It’s one of the many things astronomers look for in data recorded by their telescopes. Looking at the reflected light or the light transmitted through the atmosphere has allowed astronomers to pick out individual gases in a planet’s atmosphere (both exoplanets and planets in our own solar system).
The spectroscopic principle however dictates that there are two types of radiation that can be emitted by objects – that corresponding to the spectral signature and continuum, or in this case black body, emission caused by thermal emission. Old style streetlights produce a yellow glow as the sodium inside it is heated enough to emit the colours associated with its spectral signature – yellow. But newer streetlights, high pressure sodium lamps, emit white light – all the various colours together – due to continuum emission. In fact, they both emit spectral and continuum light, but by altering the properties of the gas inside the lamp, the preferred method of emission can be altered (in this case because the energy that would go into the spectral emissions gets lost due to collisions in the higher pressure mixture).
Turning away from the streetlights and back to planets, what does this mean? Well a planet in orbit of a star will receive light emitted by that star. The light will pass through the atmosphere and strike the surface. Some will be absorbed by the atmosphere or the surface, some will be reflected. The radiation absorbed carries with it energy, meaning the absorption heats up the planet. Now as the Earth hasn’t burnt to a crisp or turned into a small star, the planet must then be able to get rid of the energy it absorbs. It does this through ‘black body’ radiation. This refers to a thought experiment involving an ideal absorber, which is also an ideal emitter of radiation. This creates a characteristic smooth emission curve (measure of energy emitted at each wavelength) known as a black body emission profile, to which thermal emission profiles such as stars, planets and even light bulbs can be compared to. The peak of the curve is determined by temperature and allows us to determine the ‘black body temperature’ of an object, one of several ways to determine temperature using an object’s emission spectrum. The peak emission of the Sun, for example, is in visible light. The colours of other stars are also determined by the peaks of their emissions, with blue stars representing bodies with higher black body temperatures than red stars. For planets close in to their host stars, the black body emission peaks in the infrared.
Taking these sources of emission together, it becomes possible to see that in some parts of the spectrum, an exoplanet will be brighter than others. As the star itself varies in emission rates across the spectrum, this means the ratio of planet to starlight radiation varies considerably.
This is already well known; part of the field of exoplanet research has been the search for spectral windows where a given planet is easiest to see. But in this new paper, Dave and Giovanna Tinetti look at how the emissions can affect how the transit curve looks when recorded in different wavelengths, specifically, does the black disc glow a little providing a slightly incorrect transit depth, in turn affecting the derived radius of the planet?
The paper begins by noting the emissions from the night side can be directly observed. As mentioned earlier, light due to the phases of an orbiting exoplanet has been observed. This light includes both light reflected from the star on the dayside and light emitted directly from the planet on both the day and nightside. Studying the emissions over the course of an orbit and knowing the accurate phase of the planet should work, allowing anisotropic emission profiles to be uncovered (eg more thermal emissions on the dusk side than predawn etc). The planet HD 189733b is examined in this way in the next section. The results show that the transit depth in the infrared is altered enough to be detected by the Spitzer space telescope (meaning it will be very measurable in the upcoming James Webb Space Telescope).
The next section is theoretical. It starts off by noting that variations in the dayside flux over time could be mapped to nightside variations and seeing what effect this has on the results. Then it moves onto potential sources of pollution from nightside emissions on other observables. Transmission spectra looks at light shining through the translucent atmosphere surrounding the main body of the planet. That planet, like our own, will be emitting thermal radiation. That atmosphere, like our own, will be emitting radiation due to photochemical reactions (recombination of ionised gas) and chemical reactions, as well as heat transported by winds (ions like H3+ can act as thermostats, radiating heat from the planet spectroscopically). For this, some knowledge of the atmospheric composition is required in order to know what spectral lines will be radiating (this is the opposite to the method used to discover the composition – fit various modelled gas spectra to observations). The authors plot a spectrum that neglects nightside pollution, one that includes it and one that notes the difference between the two in order to determine where the effect is greatest. They determine the effect is large enough to be measured by Spitzer.
Next, they look back at work done previously and as whether the determination of the composition of the atmosphere of HD 189733b was affected by this. Modelling the star and the planet as black bodies, they determine the thermal profile of the planet and note that just from this first order approximation, they discover a measurable deviation at 8.0 microns. A future paper will include the atomic and molecular lines and determine how each is affected in greater detail.
Finally, they note that a similar effect will lead to underestimation of the secondary eclipse depth (the slight drop in light due to the reflecting planet going behind the star).
They conclude by noting we’re moving into an era of higher precision measurements, meaning stricter definitions on what is being measured are required and considerations of atmospheric composition will become more important.
This field is maturing.