Why Does the Night Sky Look Like a Dome?

When we look at the sky it definitely looks like a dome to us, stretching from the very faraway parts of the horizon, many kilometers away, to just a few hundred meters above our heads. We get this impression whether it is daytime or night-time. And, in daytime, it is quite obvious that that’s the way it should be: there are clouds in the sky, and those clouds that are near the horizon look really-really (and are!) far away (tens of kilometers, in clear atmospheric conditions); whereas the clouds above our heads start at around 500 meters approximately, and may extend much more above that, but what we usually observe are the closest ones to us. So, at daytime, the sky really is like a dome. Take a look at the following image:

The image above is a partial screen-capture from Google Earth. Suppose you’re standing on the surface of the Earth somewhere at the middle of where the red line is. What would you see if you looked up, into the sky? To help you out, the portion of the sky (the Earth’s atmosphere) that you’d see has been copy-pasted above, in the dark region of the image (the outer space), and an eye-icon has been put there signifying you, the observer. What you’d see is above that eye-icon. It doesn’t just look like a dome; it is a dome. Clouds may help us perceive the daytime dome better, but even in the total absence of clouds the portion of the atmosphere that we can see is actually a dome.

This, by the way, has been captured in the manner in which some religions design the domes of their churches or mosques. The picture that follows is from the Greek island of Santorini:

We see that the tradition of the Greek Orthodox Church wants their churches to be capped with a dome, which is supposed to represent the sky (or “heavens”) under which God placed us and let us live. On the island of Santorini, they really paint their church domes sky-blue, just as the sky appears over that island during the best part of the year.

All right, so in daytime there really is an objective reason for which we perceive the sky as a dome. But what about during night-time? We don’t see the atmosphere at night. (There is no sunlight to be scattered in the atmosphere and let only the blue color of the entire spectrum of colors pass through, reach our eyes, and let us see a blue dome.) At night, our atmosphere becomes transparent. We see through it, and we see all the way to the stars. (We can even see a handful of galaxies.) But when we look at the stars, they also appear to be arranged on a dome! Why is that? Come to think of it, the stars are at enormous distances from the Earth. Moreover, they are at enormously different distances with respect to each other. Sirius, for example (the brightest star on the night sky) is only 8.6 l.y. (light years) away from us. But Deneb, a 1st-magnitude star that can be seen from northern latitudes during most times of the year, is at a whopping ~1,500 l.y. away! Why do both Sirius and Deneb (as well as every other star) appear to be placed on a “dome” at night time, just above our heads, in spite of Deneb being around 175 times farther than Sirius?

For a more concrete example, consider the stars of the “Southern Cross”, which appear on the flags of Australia and New Zealand:

On the left, we see the flag of Australia, which includes a very stylized depiction of the Southern Cross on its right side, and probably Rigil Kentaurus (or Hadar) just under the Union Jack. On the right, we see the actual placement of those stars in the night sky. The distances from Earth of several of the stars are also shown on the right, as numbers followed by “l.y.” (light years). We observe that the distances have huge differences. For example, whereas Rigil Kentaurus is only 4.4 l.y. away (the closest star system to Earth), Hadar (to its right) is 525 l.y., and the stars of the Southern Cross also vary widely: Mimosa is at 278 l.y., Acrux at 321 l.y., Gacrux at 88 l.y., and a very faint star called 35 Crucis is at 1832 l.y. away! Why do we see all those stars as if they’re little decorative lights on a dome (which is the way some religions described them; e.g., that’s what “Allah” says they are in the Qur’an of Islam: decorative lights) even though in reality their distances differ widely?

To get a better understanding of how stars are positioned in relation to us, maybe the following “stereogram” will help:

You may see the above stereogram in actual 3D if you manage to watch the left side of the image with your left eye, and the right side of the image with your right eye.

It shows several stars of our neighborhood in their locations around the Sun. It’s not that the Sun is at the center of the universe, but it is at the center of the above stereogram (both left and right). And it’s not that the rest of the stars revolve around our Sun, but the stereogram presents them that way, to let us see them from all angles. (It’s as if we go around the stationary distribution of stars around the Sun and watch them from every possible angle.) Thus, we see that stars exist in all possible locations and at all possible distances and angles around the Sun (therefore around the Earth, too) — nothing like the “dome” impression that we get when we look at a starry night sky.

So, to recapitulate the question: if the day-time “blue dome” is objectively justified, how is the night-time “black dome” justified?

The answer follows. And the answer has to do with the limitations of human vision. Specifically, with the ways in which we perceive depth. We’ll examine the mechanisms by which we perceive depth, and find out that they all fail in the case of the starry night sky.

1. Binocular Disparity

The term “binocular disparity” refers to the disparity between the two images that the brain gets from the left and right eye. The images differ slightly, because each eye looks at the world from a different angle. Take a look at the following diagram:

On the left side of the diagram, the two eyes focus on a nearby object. For that reason, the angle formed by the two lines that lead from the eye lens to the nearby object, converging on it, is relatively large. If the object is faraway, the angle is narrower (right side of the diagram). A larger angle (left) causes a larger disparity between the two images in the brain and, likewise, a smaller angle (right) causes a smaller disparity. From this disparity, the brain deduces the distance of some objects.

Now, for the above mechanism to work, the object must be no farther than about 50 m. Beyond this distance the disparity becomes so tiny that the brain is unable to detect it.

Clearly, stars on a starry night sky are much, much farther away than a mere 50 m. (And this is an understatement. The closest star, Proxima Centauri, is 4.24 l.y. — or about 40 trillion kilometers — away.)

So: binocular disparity is no good for understanding star distances.

2. Eye Lens Focusing

When we look at a nearby object the lens of the eye focuses on it, keeping it under sharp focus while blurring the more distant objects. Similarly, when we look at a more distant object the lens of the eye focuses on that one, blurring the more nearby objects. This can be understood by the following pictures:

On the left, the focus is on the faraway objects. We can see clearly the objects on the hill and the islet in the distance. However, the stones that form the foreground, as well as the yellow lichens on the stone in the middle-lower part of the picture (still a nearby object) are all blurred.

On the right, we have the opposite situation: the focus is on the nearby objects. We can see the stones and the yellow lichens in every detail, but the faraway hill and the islet can hardly be seen — they’re blurred.

That’s how the lens of the human eye works. Even lenses of cameras work that way, but in our times the lenses of mobile phones are made so tiny that they keep everything under sharp focus (not allowing artistic photography, but keeping the billions of people who use them happy). One needs a relatively large lens in a camera (a professional one) to observe the blurring effect of focusing.

This mechanism, again, works for short distances only; no more than 20–30 m. Once we focus on objects farther than about 30 m, all objects at longer distances appear sharply focused.

So this mechanism, too, fails for objects like stars, which are trillions and quadrillions of kilometers away.

3. Object Overlapping

Take a look at the following image:

The nearby igloo partially overlaps the slightly more distant igloo. Overlapping is another mechanism by which we understand depth. An object in the background clearly cannot overlap a foreground object, and our brains know that, drawing conclusions about which object is near and which is far.

But stars are point-like objects. They cannot overlap one another. So this mechanism, too, is useless and fails with stars.

4. Shadowing

Consider the following picture:

Shadows are another great way by which we understand depth — provided they’re present. In the picture, above, the person with the red turban casts a shadow on the ground, which helps us understand he’s standing, and also at what angle he does so relative to our line of sight. So does the other person who is sitting, and so do the bowls and the rest of the objects on the ground. Shadows are consistent, cast in the same direction from all objects (otherwise we would have trouble understanding where the light source is).

But stars are not objects that cast shadows on their background (the night sky). So this mechanism is not applicable on stars, either.

5. Size Varying by Distance

When an object of known size is nearby, it appears large. When it is faraway, it appears smaller. Here is the idea, pictorially:

The truck (an object of more-or-less known size) appears large when it is near, covering a large part of the retina and, correspondingly, of the visual cortex in the brain. The farther away it goes, the smaller its image becomes on the retina and the visual cortex.

This is another mechanism by which we understand distances; but it requires objects the size and shape of which are familiar to us.

Obviously, stars, being point-like lights on the sky, have neither size nor shape. So, again, this mechanism fails as well.

6. Perspective

Take a look at the following picture:

The railroad tracks on the ground appear to converge at a point “infinitely” far in the distance. That’s because they are parallel. The phenomenon is called “perspective”, and it took a long time for painters of the Medieval Times to master it and come up with paintings that looked more realistic. By the times of the Renaissance, painters had completely mastered the principles of perspective.

But stars on the sky are point-like, they don’t give us any chance to apply perspective principles on them. Cancel this mechanism, too.

7. Wavelength Absorption

In some cases, the absorption of certain wavelengths in our atmosphere can help us understand distances. This can be understood by the following image:

The nearby mountain can be seen in all its colors. The farther away we see, the more of our planet’s atmosphere stands between us and the objects. This causes the longer-wavelength colors (red, orange, yellow, green) to be absorbed, leaving only the shorter wavelength colors (blue, purple). That’s the reason the mountains at long distances have a purplish hue, and are depicted as such in paintings.

Unfortunately, once again, there is no difference in the amount of atmosphere that stands between us and the stars. In addition, the above phenomenon can be observed only with the daylight. So this mechanism, too, is not helpful when it comes to stars.

As we saw, each of the seven mechanisms by which we perceive distances on Earth is non-applicable in the case of stars.

The result is that we cannot understand how far or close the stars are. We have exactly zero information regarding their distances.

All right. But then, why do we see them on a dome? Why don’t we see them lying on a very large plane?

This has to do with the way we perceive the ground of the Earth, the landscape around us; as well as with perspective.

If we perceived stars as arranged on an infinite plane above our heads, the stars at the farthest regions of that plane should be all faint. The farther away we looked, the fainter the stars should appear to be. But this is not what we observe. We see stars as having an equal average distribution in brightness, no matter whether they are near the horizon or at our zenith. This can only happen if the stars are arranged on the inside of a sphere.

But still, objects at the horizon (mountains, etc.) are very far, and we know that. Therefore, stars near the horizon have to be at least as far as the horizon objects. In contrast, stars at our zenith don’t have any nearby objects to compare them with, so we judge that they aren’t as far as stars near the horizon. The net effect of all this is that we conclude that stars are placed on a dome: the inside of a sphere, the zenith of which is closer to us, and the horizon of which is at the farthest distance from us.