The lights are on, and someone’s home

Something I’ve often wondered about Earth – would our civillisation be detectable from another star, by the light of our cities? It may sound silly at first, but hear me out.

Most of the street lights in the world are sodium lamps. They work by passing an electric current through sodium vapour, which excites the sodium atoms and causes them to emit that familiar orange glow, characteristic of cities all across the world. That orange glow, however, has a very distinctive spectrum. There’s no natural mechanism through which a planet like Earth would be able to emit light like it. Thus our city lights are quite an obvious indicator that this planet is inhabited, and that the inhabitants have some level of technology at their disposal. Whatever the form of artificial lighting, it would be very obviously artificial in nature. Interestingly, a couple of scientists named Abraham Loeb and Edwin Turner have recently proposed a similar concept to try and look for civillisations on other worlds.

Science fiction writers have hypothesised the concept of the ecumenopolis – a planetwide city. The most well known example is Coruscant from the Star Wars universe. It’s quite uncertain whether or not such a place might actually ever exist, but it’s not unreasonable to suppose that a sufficiently advanced civillisation might be able to construct a planetwide city. If population growth on Earth continues at the rate it currently has, the inevitable consequence would be increasing amounts of conurbation. Ultimately, a society without any sense of environmental conservation might decide to build up their entire world.

Just supposing a place like Coruscant might exist, whatever form of artificial illumination they use would stand out like a handful of sore thumbs against the spectrum of the planet. Perhaps the idea of detecting an extraterrestrial civillisation this way isn’t so unreasonable. It’s certainly no less unusual than the idea of detecting alien vegetation. In fact conceptually it’s very similar. Both Earth’s cities and forests are likely to be easily detectable from light years away. Provided, of course, that anyone was looking in the right direction and using a sufficiently powerful telescope.

That is, of course, assuming any extraterrestrial civillisation needs artificial lighting in the same way that we do. Assuming they even build cities the way we would recognise them. Assuming that they do, in fact, even have eyes.

Images via mo-webo @ instagram and Wookieepedia.

About Invader Xan

Molecular astrophysicist, usually found writing frenziedly, staring at the sky, or drinking mojitos.
This entry was posted in astronomy and tagged . Bookmark the permalink.

4 Responses to The lights are on, and someone’s home

  1. Scott near Berkeley says:

    Goat Guy has it right. The window for any planet-wide civilization broadcasting broadly into space in the electromagnetic spectrum (be it light, or radio/tv waves, etc) will occur for a very tiny interval in their history, in my opinion. As an example, we have cable in place of large television towers, marking the move toward efficiency, and pan-planet, pan-area broadcasts that can be detected extra-terrestrially will soon (forty years??) disappear. Street lighting, same thing, because of required efficiency. Most every planet will NOT have gone through a carbon-build cycle (in my opinion), creating centuries of cheap fuel to be mined and lavish lighting broadcast to the heavens.

    Basically, this is why any and all SETI will not succeed. The window is tiny, and the targets obscure. And, as soon as other ETs figure the same probabilities, out go the usable broadcast semaphores.

    • invaderxan says:

      Hmmm… Interesting. So you think that an advanced technological society would have absolutely no background emission detectable? Because in my opinion, it seems a bit unlikely that there would be nothing at all. Note that I’m not talking about broadcasts or similar here. Only background radiation emitted as a consequence of technological activity.

      And while carbon is certainly not an absolute (and may not work nearly as well in all atmospheres), some form of energy generation is likely if any technological civillisation is to exist, surely.

  2. GoatGuy says:

    Well… let’s see…

    There’s the opportunity-window. We’ve been using sodium lamps basically since the 1970s. Before that the whole world was on high-pressure mercury vapor (with an equally distinctive spectrum). Before 1950, incandescent – in all its inefficient glory – was the streetlight of choice. No distinct spectrum. And, post 2020 … I’m guessing the majority of city lights will evolve to broad-spectrum LED … since they last, are durable, becoming cheap, and turn on instantly.

    So. Civilization’s opportunity window for mercury-vapor emission was about 30 years. Sodium vapor (which – mind you – still has all that mercury vapor too!) lamps are now upwards of 40 years in service. They’ll ALL be gone by 2030, to be sure. So, that’s 30 + 40 + 20 = 90 years.

    What’s the chance that we’d SEE another civilization when its going through its 90 year metal-vapor-emission phase? Civilizations (we have to presume) last tens of thousands of years. Maybe more – we’re too damned young to know. But let’s say 25,000 years, just to be generous. 90/25000 = 0.0036 (0.36%) of a civilization’s existence.

    So, looking at all the candidate stars, all the candidate planets, with magnificent optics that can detect all the most curiously vapor-produced mercury and sodium spectra … AND assuming that ALL the planets being looked at have civilizations (which they wouldn’t of course), 0.36% of them might show-the-glow.

    Now, take into consideration that that 25,000 years of civilization (ours, say) is at the terminus of 4,500,000,000 years of planetary evolution, reformation, sprouting of living oozes, evolution of species, volcanism, various asteroid impacts, development of buggies, bean counters and bellfries, … that’s just 25,000 / 4,500,000,000 … 0.00056% (woo!) of the life of said planet. We’re but a hair’s breadth of icing on top of a house-sized cake of cement and stone.

    I doubt – given such statistics – that we’re going to “get lucky” and see the emmission-spectrum of a civilization that just happens to be in its metal-vapor-lighting window, any time soon. We could look at 100,000,000 candidate planets (to keep things sober), in esquisite spectrographic detail, and only find TWO on the average. The odds of winning the State Lottery are 1:14,700,000 … or about 3 times better.

    I’m gonna buy a lottery ticket.


    • invaderxan says:

      Hmmm… All excellent points and well made. Though there are a couple of facts you’re missing which broaden things somewhat from the ~90 year margin you mention. :)

      Incandescent bulbs do indeed give off continuum emission without any discrete transitions. With all their inefficiency, even these would give an excess in the near infrared (around 1 micron). However, it would be difficult to deconcolve that from a planetary spectrum.

      More interesting is your mention of LED lighting. I’m not sure I agree that all sodium lamps will be gone by 2030, for one big reason. White LED lighting uses a mixture of gallium nitride and indium nitride in its makeup. Neither gallium nor indium are particularly common elements, and it’s questionable whether Earth actually contains enough of these two elements to create enough LEDs to illuminate every single part of the world. Until those asteroid miners start bringing home the proverbial bacon, I have my doubts that LED production will be able to continue increasing at its current rate.

      Secondly, there’s the matter that white LEDs are, in fact, not white. All LEDs actually emit over a very small range of frequencies. As you can see in this spectrum, the GaN/InN component of the LED emits brightly in the blue part of the spectrum, causing a huge peak around 460nm. This blue light is then absorbed and re-emitted by white phosphors in the LED, giving the appearance of white light – and the Stokes shifted light re-emitted by the Ce:YAG phosphor in most “white” LEDs is also fairly atypical in its emission profile. Overall, it still has a rather abnormal looking spectrum. Nothing in any planetary atmosphere which we know of will cause an emission feature like that.

      So in this case, I’m willing to stand by my original statement with a single caveat. I will still say that any artificial lighting will be very obviously artificial in nature and easy to spot – with the caveat being that a civillisation which somehow decides to light its cities with incandescent lighting will be harder to see.

Comments are closed.