Astronomers hope that one day soon we'll obtain a spectrum of light that might tell us whether or not an Earth-sized exoplanet harbors life. This spectrum could be of starlight filtered through the planetary atmosphere, or of reflected and emitted radiation. In either case it would probe the chemical composition of an alien world.

The detection of an imbalance, or disequilibrium, of atmospheric components has long been considered a smoking gun for a planetary biosphere. For example, a rich mix of oxygen and methane in a warm environment is not a stable state. Methane oxidizes relatively speedily to form water and carbon dioxide. So detecting the presence of both oxygen and methane would suggest an active replenishment mechanism. Life (as we know it) represents an excellent candidate for supplying these ingredients.

So far so good. Find that Earth-sized planet, gear up our best technology, and sniff for a spectral biosignature.

Except there may be a hitch. It's possible for nature to throw us a curve ball. In a new paper in the Proceedings of the National Academy of Sciences, Rein, Fujii, and Spiegel explore a disquieting possibility. They ask what would happen if this juicy looking exoplanet also happens to have a substantial moon that itself harbors an atmosphere, yet both objects are devoid of life.

If the Earth-sized exoplanet has an oxygen rich atmosphere (but nothing that would react with that gas), and the exomoon has a Titan-like methane rich atmosphere, we might be duped into thinking that we're seeing a single, out of equilibrium, biologically driven environment - an inhabited planet. That's a problem, and it arises because, like it or not, our data is going to be comparatively low-fidelity.

It's going to be extremely difficult to distinguish between a planetary spectrum originating from one object or two closely spaced objects, but this is as good as our near term (or even long term) technology is going to give us. Oxygen and methane represent just one example of such confusion, other chemical pairings would be similarly affected.

What's the solution? The authors suggest two simple ways forward. One is that we might be lucky enough to identify a true 'Earth-twin' within some 30 light years distance - close enough to give us a shot at beating this kind of false-positive with careful measurements. The other is to just forget about looking at Earth-twins. Other potentially habitable planets, such as those around low-mass stars, or so-called super-Earths, could allow us to test for this kind of confusion by making the spectral measurements easier.

Of course, we don't know how common such 'spoof' systems might be, but when we're talking about determining whether or not we're alone in the cosmos it would pay to be very, very certain about what we're seeing.