≡ Menu

Lunar Eclipse: A Proxy for Exoplanet Observation

When it comes to detecting life on planets around other stars, my guess is that what will initially appear to be a life signature will quickly become controversial. We might, for example, find ozone in an exoplanet atmosphere with a space telescope like HabEX (Habitable Exoplanet Observatory). That would lead to hyperbolic news stories, to be sure, but ozone can happen when nitrogen and oxygen are exposed to ultraviolet light. The presence of ozone makes no definitive statement about life.

In fact, definitive statements about life may take more than a few decades to achieve. If ozone seems like a good catch, that’s because it implies oxygen, which makes us think of photosynthesis, but oxygen itself is hardly infallible as a biosignature. Oxygen-rich atmospheres can be completely abiotic, with UV from the host star breaking down carbon dioxide. For that matter, an atmosphere rich in water vapor can produce oxygen and hydrogen through the effects of UV radiation.

Better, then, to look for a mix of things. Methane and oxygen detected simultaneously would be interesting because the two would need constant replenishment to appear together, indicating a lack of chemical equilibrium, and surely that’s a life signature. Or is it? Volcanoes can produce methane [although see Brig Klyce’s comment below], and so can infalling carbon-rich debris entering an atmosphere.

No, given the significance of the find and the difficulties in the measurement, I don’t expect anything but ambiguity when we do get the right instrumentation to study terrestrial-class worlds transiting their stars. We’ll be looking at starlight screened through a thin layer of atmosphere, and I thInk we can expect a battle royal among astronomers as we try to decide among the possible causes of the chemical signatures we do find.

And then there’s the question of just where a particular planet is in its development. Our planet when young was rich in hydrogen and helium, but later volcanic eruptions would supply carbon dioxide, water vapor, sulphur. No oxygen, yet, but if we find these signatures in an exoplanet atmosphere, we may be looking at a world that will one day experience an oxygenation event like the one that changed everything on Earth two and a half billion years ago.

How tricky atmospheric biosignatures turn out to be. All this gets more complicated still because of the fact that we need an observing campaign at different wavelengths to make a detection of the different gases involved.

But ozone seems like a logical place to start. Giada Arney (NASA GSFC) is co-author of a paper that addresses biosignatures in a more local way, as I’ll explain after her quote:

“Astronomers will also have to take the developmental stage of the planet into account when looking at younger stars with young planets. If you wanted to detect oxygen or ozone from a planet similar to the early Earth, when there was less oxygen in our atmosphere, the spectral features in optical and infrared light aren’t strong enough. We think Earth had low concentrations of ozone before the mid-Proterozoic geological period (between roughly 2.0 billion to 0.7 billion years ago) when photosynthesis contributed to the build up of oxygen and ozone in the atmosphere to the levels we see today. But because the ultraviolet-light signature of ozone features is very strong, you would have a hope of detecting small amounts of ozone. The ultraviolet may therefore be the best wavelength for detecting photosynthetic life on low-oxygen exoplanets.”

Arney, working with Allison Youngblood (Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder) and colleagues, has explored ozone detection close at hand through new Hubble observations. The idea here is to use light reflected off the Moon during a lunar eclipse to try to detect biosignatures in Earth’s own atmosphere. Consider it a test case for the kind of atmospheric characterization we hope to be doing with much more distant worlds in coming decades. Have a look at how things lined up for this observation.

Image: This diagram explains the geometry of the lunar eclipse. When the Moon is entirely in the Earth’s umbra (known as a total lunar eclipse or umbral eclipse), all sunlight reaching the lunar surface has been refracted or scattered through Earth’s atmosphere. When the Moon is in Earth’s penumbra (known as a penumbral eclipse), illumination comes from both direct sunlight and sunlight refracted and scattered through the planet’s atmosphere. This process is similar to an exoplanet transit observation. Credit: M. Kornmesser (ESA/Hubble), NASA, and ESA.

As you can see, the Moon is reflecting sunlight that has passed through Earth’s atmosphere before being reflected back to Hubble. Make no mistake, we’ve done this kind of thing before — Earthshine has its uses — but this is the first time a total lunar eclipse has been captured at ultraviolet wavelengths and from a space telescope, so it’s a decent proxy for what we’ll do later with HabEX and other instruments. Hubble found a strong ozone signature, a photochemical byproduct of molecular oxygen we’ll hope to be finding on rocky exoplanets.

The strength of the ozone signature demonstrates the problem with working with ground-based observations at these wavelengths, for Earth’s atmosphere absorbs ultraviolet light. This work, along with other observations — ground-based but at different wavelengths — of the January 20-21, 2019 lunar eclipse, helps to develop spectral models for atmospheric characterization. The weakness of the lunar eclipse as a proxy is also clear, as the paper notes:

…observing the moon with HST has different challenges, namely lower spectral resolution and pointing instability. The moon is not a homogeneous surface, and pointing instability does not guarantee that the overall albedo and reflectivity spectrum of the lunar surfaces observed in-transit and out-of-transit are identical. Creating transmission spectra with only Earth’s spectral signatures relies on the in-transit and out-of-transit solar and lunar features being identical. We find evidence for overall albedo variations of the moon between our out-of-eclipse spectra, but defer a thorough analysis to a future paper.

This, too, is helpful, though, for previous papers in the literature have analyzed problems of refraction and transit geometries as they affect the exoplanet spectroscopic signature. The models for future work are only now being tested through events like these.

Image: This ground-based telescopic image of the Moon highlights the general region where astronomers used NASA’s Hubble Space Telescope to measure the amount of ozone in Earth’s atmosphere. This method serves as a proxy for how they will observe Earth-like planets around other stars in search of life. Credit: M. Kornmesser (ESA/Hubble), NASA, and ESA.

The paper is Youngblood et al., “The Hubble Space Telescope’s Near-UV and Optical Transmission Spectrum of Earth as an Exoplanet,” Astronomical Journal Vol. 160, No. 3 (6 August 2020). Abstract / Preprint.


Comments on this entry are closed.

  • Michael Fidler August 11, 2020, 10:33

    Could exoplanet aurora be identified in the spectrum and give us some idea as to how strong the magnetic field is on the close in planets around M and K dwarfs? Have they looked to see if the signature from our aurora has been reflected off the moon during a lunar eclipse and see if the polarization and spectral lines may be different from ozone. This would give us some idea if the range of planets could protect themselves from the effects of severe solar storms. The tidally locked exoplanets are one of the areas least understood in the compact solar systems and high magnetic fields around them would give a higher chance for life to develop. This in itself may be a bio-signature for without it life may not survive in the hostile storms generated by their star, would earth have life without our magnetic cocoon?

    • Paul Gilster August 11, 2020, 16:59

      What an interesting question! I don’t recall anything on auroral effects in this paper but I wonder if it’s been discussed elsewhere.

    • Alexander Tolley August 11, 2020, 17:59

      This in itself may be a bio-signature for without it life may not survive in the hostile storms generated by their star,

      This is not so much a biosignature, but rather a possible necessary but insufficient condion for life. There are auroras on our gas giants but I don’t think that means that they are suitable to harbor life.

      If a magnetic field of sufficient strength is needed, then it is more like the need for the evidence of [surface]water.


      The issue of methane is interesting. It is very dependent on the amount that volcanism can emit. Calculations for teh early Earth have indicated that methanogens are required to ensure a long atmospheric lifetime. If methane is provided by infalling comets, then the absence of carbon monoxide in the presence of plentiful methane is a biosignature of early life. This remains true even today, although gas leaks from “fracking” is now a major contributor to atmospheric methane levels on Earth.

      My guess is that almost any biosignature that is possible before the Phanerozoic and where complex life lives on land (by mid-Paleozoic), will prove prone to dispute. That means that well over 80% of the time that life exists on a world like Earth, the disputed biosignatures may be false negatives. This will push astronmers to look most intently for life on worlds most approximating modern Earth, where orthogonal biosignatures will be most likely to be indisputable and the recognition will go to the first discovery team.

  • Geoffrey Hillend August 11, 2020, 17:32

    Paul Gilster, I recall reading a paper online using the reflection of ultra violet EMR off of our Moon to detect diatomic or molecular nitrogen which has no visible or infra red absorption spectra, but ozone does, so this Moon reflection technique is not needed to detect ozone. Both ozone and nitrogen absorb ultra violet. Nitrogen is more difficult to detect because as a “homonuclear molecule it has no rotational and vibrational transitions and hence no spectral signature in the visbile and infrared wavelengths.” Segar 2010, p. 231, Exoplanet Atmospheres.

    I like the idea of the reflection of ozone and ultra violet light off of the Moon to detect the spectra of an exoplanet. It would have to be an exact Earth twin around a G class star in the life belt though.

    We may also not have to worry about false positives with oxygen because the photo dissociation of water into oxygen and hydrogen makes a lot less oxygen than produced by life and photosynthesis, therefore, the low levels ozone in an atmosphere on an exoplanet with low levels of oxygen might be too difficult to detect with a spectrometer. On the other hand, we don’t have any examples of an exact Earth sized exoplanet without a Moon around a G class star in the life belt. I have to think that there would be water spectra, but would we be able to detect any ozone and oxygen on a abiotic world?

    • Paul Gilster August 12, 2020, 12:04

      Yes, I would agree that it’s likely that abiotic oxygen is going to be more scarce than what life might produce, though the number of imponderables we’re dealing with still points to an ambiguous reading of early atmospheric transmission spectra, in my view.

    • Alex Tolley August 12, 2020, 13:42

      I agree about the O2. It has been suggested that a world might build up an O2 atmosphere due to photolysis, but this assumes there are no other elements or compounds to be oxidized. Methane would be oxidized to CO2. Iron would rust (Fe2O3). Over time, these should draw down any free O2. While Ch4 with the O2 is very definitely an Earth-life biosignature, I am less troubled (in my naivete) about a rich O2 atmosphere without contemporaneous reducing molecules.

      But clearly, low free O2 levels do not indicate the planet is dead, as it may well be in a state before photosynthesis evolved and/or the planet had its own great oxidation event (GOE).

      Claiming life exists on an exoplanet using some biosignature[s] will, at least early on, be one of those “extraordinary claims requiring extraordinary proof”. There will need to be almost no room for doubt to settle the argument from the skeptics. I suspect this will require multiple gas signatures as well as computer modeling of atmospheres to make a rock solid case that life exists.

      I would hope that a after the first proof of exoplanet life, that we will start building telescopes to image such a world (e.g. at the solar gravity focus) as well as fast insterllar probes that will make a flyby within some reasonable time frame acceptable to the funders. It might even generate a new golden age of exoplanet naturalists, trying to catalog new species and understand their evolutionary history from very limited data.

  • Geoffrey Hillend August 11, 2020, 17:43

    We also would not need an eclipse to see the green light of aurora borealis, but just the night side of an exoplanet, but it might have to be direct imaging with a starshade or a sensitive chronograph. It would indicate a magnetic field. I don’t know how sensitive the telescope would have to be to detect the aurora though.

  • Brig Klyce August 12, 2020, 12:01

    Paul, I saw your entry of today. May I suggest a wording change that may eventually matter? Volcanoes _release_ methane. That they _make_ it is an assumption with tenuous support. The strictly abiological manufacture of methane needs catalysts and a clean environment that seems unlikely for volcanoes. Maybe there are deep reservoirs of regular biological methane. In fact, we know there are ones pretty deep.

  • Geoffrey Hillend August 12, 2020, 16:43

    I agree that in potential abiotic oxygen and ozone can have a spectral signature in an exoplanet’s atmosphere, and so can abiotic methane that is if we can see those smaller quantities from many light years. It might be hard to know if it was produced by life or not. I was hoping to rule out the ambiguity and imponderables to make it easier to predict an exact earthlike environment which might have a lot of contingencies and limitations.

    We do have one example which is Earth’s early atmosphere. There was no oxygen and mostly carbon dioxide, water vapor, ammonia, methane and molecular nitrogen and hydrogen. There was no oxygen and ozone to block UV radiation. There was an ocean and water vapor and it’s photo dissociation by UV. There could be some chemical mixing of oxygen to make ozone, but it is uncertain whether that will be enough to be detectable from a great distance. It might also not last long enough before it is destroyed by the same UV EMR. I am hoping that it won’t be detectable, so it would make oxygen and ozone a prime biosignature gases which would limit our search to exact Earth twins with a Moon around a G class star in the habitable zone. This would rule out K stars and M dwarfs if we consider other contingencies like planetary mass, distance from the star, magnetic field, solar wind stripping, and high energy solar EMR.

    On the other hand, if oxygen and ozone spectra are found on all Earth sized planets in the life belt K and M stars that would certainly make a case for the potential hardiness of life and make it look like life is very abundant in our universe. If O2, O3, and CH4 are all detectible, how will we know whether it’s abiotic or biotic? Maybe the biotic spectra of these gases will be stronger in the worlds than the abiotic ones?

    I like the fragility of life idea better with a more rare Earth twin with a Moon crucial for the spectra of biosignature gases. It kind of forces us to pay more attention to our biosphere and ecology.

  • Thomas R Mazanec August 13, 2020, 11:55

    How do we know the crater was formed 20 or 22 million years ago, as opposed to 10 or 35 million years?