If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

Main content

Apsidal precession (perihelion precession) and Milankovitch cycles

Apsidal Precession (Perihelion Precession). Created by Sal Khan.

Want to join the conversation?

Video transcript

We've learned that axial precession, it's not a change in the tilt or the obliquity of our rotational axis, it's a change in the direction. And over a long period of time, 26,000 years, it kind of traces out a circle. And the main affect of that is that if we wait long enough that our rotational axis, or you could say almost the North Pole, will be pointed in a different direction. And so if our rotational axis is pointed in a different direction after a long enough time, then the absolute point in our orbit, if we use the sun as our frame of reference, the point in our orbit when we are most pointed away from the sun, or when the Northern Hemisphere is most pointed away from the sun, will be earlier in the orbit. Now I emphasize that that won't necessarily mean earlier in our calendar, because our calendar, by definition, takes into consideration, I guess, or it's more based on when we are furthest tilted away from the sun or furthest tilted towards the sun. So even though if we wait 1800 years, like the example I gave, we will be most tilted away from the sun. The Northern Hemisphere will have its winter equinox at an earlier point in the orbit. According to our calendar, it will still be December 22nd. If our calendar instead was based-- and it's not based on this-- but if our calendar was based on the exact point in orbit, then our year would be about 20, 25 minutes longer every year and then the date for the start of winter actually would go back. 1800 years later, the date of the start of winter would be November 22nd. But that's not how we measure our calendar. Our calendar is actually measured from equinox to equinox. From December 22nd or 21st, there's slight fluctuations depending on the calendar, but that will always be the date that we are most pointed away from the sun. That will not be necessarily the date at this exact position relative to the sun itself. And that's why the actual perihelion does change. Because if this is always December 22nd, and if we, at first, assume that the perihelion is always at the same fixed point in space relative to the sun, although that's not exactly the case, but if we make that assumption, then it will be further and further after that December 22nd, further and further after that time, that we are most pointed away from the sun. And that's why you have this kind of pushing back of the perihelion. Now, what I want to add to this video is that the perihelion itself is also changing. So if I draw the sun again, and right now our orbit looks something like this, and I'm going to exaggerate the eccentricity of it so that the perihelion and the aphelion are a little bit clearer. So right now this is the perihelion. This is the aphelion based on the way I drew it right over there in different colors. I don't want to show that's necessarily where Earth is. Perihelion and aphelion, there is also a rotation of this, of the perihelion. And sometimes this is called the precession of the perihelion, or perihelion precession, or apsidal precession. These are all very hard to say. And so if we wait several thousands of years our orbit might look a little bit like this. The actual perihelion will have rotated. So our orbit will look like this. The actual ellipse would have rotated a little bit. You wait a little bit longer, it might look like this. And obviously, I'm once again talking about over thousands and thousands of years. From a year-to-year basis, you really wouldn't notice the difference. But what that does is, is we talked about the axial precession, that this change in direction of our rotational axis, it takes 26,000 years to complete one period. So 26,000 years from today, our polar axis, if we don't think about our rotational axis, if we aren't too concerned about the actual change in tilt, which there will be some small change in tilt, but 26,000 years from now, our pole will roughly point in the same direction again. We would have completed one whole period of axial precession. However, it does not take 26,000 years for whatever our date of perihelion is today. So it's in January. I actually don't know the exact date. You can look that up. But whatever that data is in January, it will not take 26,000 years for it to be that date again. And it would have taken 26,000 years if the perihelion itself were not changing, if it always stayed fixed over here, if we did not have this apsidal precession. But since it is also changing, you could kind of say it is over 1,000 years moving in that direction while our January date is moving in that direction, they will actually meet sooner, so that the precession will be back on whatever date it is on January, less than 26,000 years from now. And actually the exact time, and I haven't done the calculation, but this is what I've read, is that it will be 21,000 years from now. And then on top of that, if that's enough for you that not only is the direction of Earth's rotational axis changing and the tilt is changing and that the perihelion and the aphelion are also rotating around, it's also the case that the eccentricity of the orbit itself is changing. So over long periods of time, Earth's orbit becomes more or less eccentric. And we've learned that almost circular has-- well if you're circular you have no eccentricity and then you can become more and more eccentric, which means you're more and more of kind of this flattened out ellipse. And these cycles occur, these eccentricities cycles occur over approximately 100,000 years. And so to revisit the Milankovitch cycles-- and once again this is a theory. We're not sure whether this is necessarily causing our ice ages or whether this is necessarily a major influence over long-term climate change, but the theory of Milankovitch cycles is that over long periods of time, if the eccentricity changes enough and if it coincides with when the perihelion and the seasons also coincide, maybe that's enough to start an ice age. Or maybe that's enough to take us out of an ice age. And actually, if you want to throw something even more on that, the actual plane of our orbit also changes over time mainly because of interactions with the outer planets. Anyway, I'll leave you there. As you could imagine, this is a very complex topic, but hopefully you now have an appreciation of all the different ways our orbit can change and maybe start to think about how that might affect our weather. Although, we don't necessarily know how it does or whether it even really does affect going into or out of ice ages.