I'm working on a comic which is essentially alternate-history fiction, and a big upcoming part of it involves the formation of the Chicxulub crater 63 million years ago. As I am a stickler for accuracy where things don't actually matter, I was hoping to be able to portray the night sky accurately during that time period.

I tried using Stellarium to simulate the night sky back then, but it only allows me to go back about 63,000 years, and not the 63,000,000 I need. Presumably the model isn't particularly accurate beyond a certain point anyway, but I want to be able to say I Tried(tm).

Or should I just not worry about this, given that it's highly unlikely that any readers would recognize the night sky as being Earth's, much less of a particular geological epoch? (For what it's worth, even going 63,000 years back shows a completely different sky, with the constellations distorted beyond recognition.)

  • 2
    $\begingroup$ well, i terms of pollution: clear $\endgroup$
    – Stefto
    Apr 9, 2015 at 8:45
  • $\begingroup$ @Stefto Perhaps in the alternate timeline, but not so much in our reality ;) $\endgroup$
    – fluffy
    Apr 9, 2015 at 8:52
  • $\begingroup$ I'd try the Celestia Project ( celestiaproject.net ) and see if that will let you go back far enough. Granted, there's plenty of reasons below as to why it's largely irrelevant, but it could give you some ideas for major stars/etc $\endgroup$
    – Andon
    May 1, 2017 at 15:05
  • $\begingroup$ There's an old creationist theory that the Earth used to be covered in a "Vapor Canopy", that might help to spur some ideas. $\endgroup$ Mar 30, 2019 at 10:38

3 Answers 3


It's probably impossible to project their positions that far, because on long time scales the orbits of stars around the galaxy are chaotic, meaning they show sensitive dependence on initial conditions (even a microscopic error in your estimate of their positions and velocities at one time will lead to wildly inaccurate predictions about their positions at much later or earlier times). This article says:

It turns out the motion of stars can be approximately described by a simple differential equation called the Henon-Heiles equation. Unfortunately the solution to this equation is chaotic. In other words the solution is very dependent on a star’s initial velocity and position. Determining precise measurements of a star’s position and velocity can be quite a challenge. So usually we have to look at general properties of the solution rather than finding a particular solution.


Since a galactic year is 250 million years, in 63 million years the Sun will have moved about 1/4 of the way around the galaxy, as will other stars in the sky, in slightly different directions and differing rates, plus the up-down movement of the stars, like carousel horses.

In short, the sky will be completely different and unrecognisable.

You might determine the locations of the visible Andromeda galaxy and the Magellanic Clouds relative to the Milky May. Different external galaxies may be visible from the other side of the Milky Way.

If you want a real chart with orbital movements to work back, look at Gaia and data from Hipparcos.

Breaking news: the Gaia data processing team has released a video!

…the stars themselves will move. Combining positional data of unprecedented accuracy for two-million stars taken over years by ESA's Earth-orbiting Hipparcos (now defunct) and Gaia satellites, a future extrapolation of star movements was made over millions years. As shown in the featured video, many stars make only small angular adjustments, but some stars -- typically those nearby -- will zip across the sky. Once familiar constellations and asterisms will become unrecognizable as the bright stars that formed them move around. Not shown are many local nebulas that will surely dissipate while new ones will likely form in different places.

  • $\begingroup$ Yeah, I was just hoping that there was software that would allow me to see this (approximately). Maybe there's a way to convince Stellarium to do what I want. It's open source, after all... (I'm guessing Stellarium uses the data from those two satellites in the first place.) $\endgroup$
    – fluffy
    Apr 9, 2015 at 8:54

The constellations change over time in a chaotic manner so perhaps you would just have to make up new or perhaps that should be old star positions.

However there would have been other changes in the night skies . Over time the Moon has receded so 60 Million years ago it would have been closer and larger. This would also have changed the appearance of solar eclipses. In addition it is very probable that over time there have been changes to the appearance of the Zodiacal Lights and Zodiacal Band and there would have occasions when comets were more abundant.

Going back further in time when the Moon would have been substantially bigger and at times disintegrating comets might have filled the the night skies but it would have required a clear atmosphere to see them from Earth. Perhaps the only advanced intelligent sentient beings around then would have been time travellers.

Galileo was amongst the first humans to use a telescope to discover the moons of Jupiter but certain animals with exceptional night vision should have been able to routinely see such faint features of the night skies. I do not know whether dinosaurs had eagle eyes but even if they could have spied the approaching comets and asteroids of 60 million years ago they would still have been doomed.

Edumnd Halley noted in his paper on astronomical impacts argued that such catastrophic collisions could benefit other animals which might replace the older established orders.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .