
How Astrophysics Can Literally Save the World
Season 10 Episode 37 | 17m 23sVideo has Closed Captions
How is it possible to tell if a space rock will one day collide with the Earth?
Giant space rocks are definitely going to hit the Earth again. We actually do know how to deflect them, but only if we find them and correctly assess their risk. But the solar system is a chaotic place. How is it even possible to tell if a space rock will one day collide with the Earth?

How Astrophysics Can Literally Save the World
Season 10 Episode 37 | 17m 23sVideo has Closed Captions
Giant space rocks are definitely going to hit the Earth again. We actually do know how to deflect them, but only if we find them and correctly assess their risk. But the solar system is a chaotic place. How is it even possible to tell if a space rock will one day collide with the Earth?
How to Watch PBS Space Time
PBS Space Time is available to stream on pbs.org and the free PBS App, available on iPhone, Apple TV, Android TV, Android smartphones, Amazon Fire TV, Amazon Fire Tablet, Roku, Samsung Smart TV, and Vizio.
Providing Support for PBS.org
Learn Moreabout PBS online sponsorshipThank you to DeleteMe for Supporting PBS.
Giant space rocks are definitely going to hit the Earth again.
We actually do know how to deflect them, but only if we find them and correctly assess their risk firs.
But the solar system is a chaotic place.
How is it even possible to tell if a space rock will one day collide with the Earth?
Hey everyone!
Before we dive in, a quick update on our experiment.
A few episodes ago, we asked for your help in an experiment to see if more likes, comments, and subscribers could give Space Time a boost in the algorithm—and wow, did you deliver some impressive datapoints!
In the first 24 hours of our latest episodes, we saw over 8,000 comments (a 1043% increase) and 45,000 likes (up 275%).
Overall this helped boost impressions by 143%, spreading Space Time to more of the community.
Even better?
Many of you told us Space Time is popping up on your homepages again!
Honestly, this has been the best 10-year anniversary gift we could ask for.
So, what’s next?
Phase 2.
This time, let’s push even further.
Our goal: 50,000+ likes in 24 hours and a thriving comment section.
That means asking questions, replying to others, and keeping the conversation going.
We’ll track the results and report back soon!
And if you haven’t seen yet, we’ve got special 10-year anniversary merch to celebrate a decade of breaking brains—check it out at the merch store!
Now, let’s get into it.
To you this may look like a moving smudge.
But it’s actually the asteroid 2024 YR4, discovered by the Asteroid Terrestrial-impact Last Alert System (ATLAS).
Based on early observations, NASA calculated a 3.1% chance of this striking the Earth on the 22nd of December, 2032.
Since then, further observations have lowered that impact probability to close to zero.
essentially eliminated the threat Low enough that we don’t need to scramble our spaceship full of oil rig workers, and also low enough that this isn’t an episode about whether 2024 YR4 will hit us.
Rather, this is an episode about how it’s even possible to go from a fuzzy dot hopping across a screen to a precise probability of that dot manifesting as a giant space rock burning through our atmosphere and flattening a city or a continent.
In our recent episode we talked about the OSIRIS-REX mission, which successfully intercepted the asteroid Bennu and returned a sample to Earth.
Bennu is what we call an Earth-crossing asteroid, which means orbital path crosses right through Earth’s path.
That means it has a potential risk if hitting us if that crossing occurs when Earth happens to also be at the same spot in its orbit.
Bennu won’t do that any time soon, although do land on the thing we still needed incredibly precise knowledge of its trajectory through 3-D space, which had to be determined from those fuzzy moving dots.
Now that OSIRIS-REX has finished with Bennu, it’s been renamed OSIRIS-APEX, and it’ll intercept the asteroid Apophis which has a series of uncomfortably close approaches over the next several decades.
But space is very very big.
How is it possible to know the location and orbit of a faint space rock with the precision to land on it, or know whether it’ll land on us?
Imagine that you’re an astronomer doing some astronomy of some sort on a big telescope.
You notice that an unknown blob moves across the patch of sky that you’re observing over the course of a few nights.
Looks like an asteroid because of the speed of motion, and the lack of tail means probably not a comet.
You check against databases of all known solar system bodies and it’s not one of those.
So it's new.
How do you know whether it’s coming our way?
From these grainy images, you don’t know much.
It could be big, far away, and fast, or small, closer, and slower—or anything in between.
We would say that these properties are degenerate—they look the same from this 2-D image.
Now you as an astronomer have seen Don’t Look Up and so you know that time is of the essence when evaluating a potentially threatening object, so you get to work figuring out its orbit.
The first step is to determine the true location of the object.
You know its exact angular position—the direction relative to Earth, because you know where your telescope was pointing in each image and, more importantly, you can calibrate very precisely relative to the well-known stars in the background.
But you have no idea how far away it is, and so we have this degeneracy I mentioned between close, slow orbits and distant, fast orbits.
Fortunately the problem of determining an orbit has been solved for a couple of hundred years by some of the greats of old—Laplace and Gauss in particular.
They in turn used the even older work of Johanne Kepler, who provided the first mathematically accurate description of elliptical orbits.
So with only three observations, reasonably separated in time, it’s possible to determine which Keplerian orbit your moving smudge is on.
These days there are improved versions of the Laplace and Gauss methods, but the principle is the same.
From your few observations, you manage to solve for the six classical Keplerian orbital parameters, which tell you the path the asteroid would follow around the Sun assuming that the only thing affecting that motion was the Sun’s gravity.
Which isn’t the case, but I’ll come back to that.
Right away you can figure out some useful stuff.
The asteroid has an eccentric orbit that takes it from midway between the orbits of Earth and Mars when furthest from the Sun—its aphelion, to midway between Earth and Venus when closest, at perihelion.
This means you’ve found a new Apollo group asteroid—which means it crosses Earth’s orbit and has a “semi-major axis”, or average orbital radius, is larger than Earth’s.
In contrast the other group of Earth-crossers—the Atens, have smaller semi-major axes than Earth.
Either way, these Earth-crossing asteroids are dangerous because they pass through Earth’s path twice during each of its orbits, and if that happens while Earth is in that same spot then kablooee.
You calm yourself by remembering that 2024 YR4 and Apophis, which did not hit the Earth, were both Apollos, but then you recall that the Chelyabinsk and Tunguska objects were also Apollos and they did hit.
But how big a hit would this be?
Depends on the object’s size, which you can also now estimate.
You know that the light you receive from this asteroid is reflected sunlight, and if you assume a typical rocky composition you can guess its albedo—the fraction of light reflected.
Now that you know its current distance from the Sun and from the Earth, you can compute what surface area it would need in order to have the observed brightness.
That in turn gives you the approximate size, with some assumptions about its shape.
You estimate around 40m in diameter assuming that it’s roughly spherical.
There’s a very large uncertainty in this, so it could actually be as little as half or as much as twice that size.
If the former, it’s the same size as the one that exploded over Chelyabinsk in 2014 with the explosive equivalent of a smallish thermonuclear bomb.
But the upper end of that size range is in the realm of the Tunguska impactor that flattened 2000 square km of Siberian forest in 1908 with an explosive power an order of magnitude larger than the Chelyabinsk explosion.
Given a bit of time this size can be determined by taking the spectrum of the object would determine its composition, and so that would give you a much better constraint on its albedo and hence size.
And observations with a more powerful telescope or radar could resolve the object and literally measure the size—but that’ll have to wait for a close approach to the Earth.
For now, you have enough information about this new, potentially hazardous object to keep digging.
Now that you know the orbital elements you could just run a simulation of both Earth and the asteroid’s orbits far, far into the virtual future to see if they ever collide.
And you know that would be useless.
The real solar system is a messy place.
Your asteroid is going to be tugged off its perfect Keplerian orbit by the gravitational fields of the planets.
It’s also going to be pushed ever so slightly by sunlight—by radiation pressure.
And there can even be a significant nudge caused by a sun-heated rock radiating away its own heat in different directions as it rotates—that’s the Yarkovsky effect.
The only way to understand the possible trajectory of this object is to simulate that future motion step by step by step, and at each step figure out the tiny deflections caused by all these influences.
The gravitational pulls of Venus, Mars, Jupiter, Saturn and even sometimes the Moon have small but noticeable effects, but the real wild card is Earth’s own influence.
When the asteroid makes a close approach to the Earth, it’s deflected by the Earth’s gravitational field, sometimes getting slung into a very different orbit.
That deflection depends strongly on exactly where the asteroid passes the Earth on that close approach, and that’s very hard to predict due to the various uncertainties.
You find that the main uncertainties are the effects of solar radiation and the Yarkovsky effect due to the unknown geometry and composition of the object.
These have only a tiny influence in the moment, but that influence and so the uncertainty compounds over time.
But worse than that is the fact that your starting position and velocity for the asteroid are already uncertain, in particular due to the difficulty in knowing the exact distance from the Earth.
Due to these uncertainties, there’s no way to find the exact path—the best you can do is to simulate many, many possible paths that explore the range of the uncertain parameters—in other words, you need run a Monte Carlo simulation.
Doing that, you discover a huge range of possible futures for this asteroid.
On each close encounter with Earth, it passes either in front or behind the Earth, or, sometimes, straight into it.
But you notice something interesting.
Hypothetical trajectories that do end up hitting the Earth tend to be deflected onto those paths by the Earth itself during a previous close encounter.
And during those close encounters, all of the subsequently cataclysmic trajectories pass through small region in space or one of a very few regions.
These regions are called gravitational keyholes, and they can be as small as a couple of hundred kilometers in diameter.
If a trajectory misses a keyhole, Earth is safe for the time being, but if it threads one of the keyholes for a given close encounter then impact at the next close encounter is extremely likely.
Calculating an impact probability is now simplified to calculating the probability of threading a gravitational keyhole.
It’s also true that deflecting a potential impactor is simplified to diverting it from passing through a keyhole if you can catch it before that happens.
That means you only have to offset its position by kilometers at the keyhole, rather than the several thousand km needed to avoid an Earth collision after it already passes a keyhole.
In 2022, NASA’s DART mission proved we can indeed deflect a large asteroid by the requisite amount by slamming a mass into it.
As long as we do it early enough.
But, assuming that keyhole passage happens, now you can at least calculate with some accuracy when and where the asteroid will impact the Earth.
Each different keyhole focuses the asteroid towards an impact corridor—a trail across the planet that’s defined by the fairly precise location that the asteroid crosses our orbit, with the uncertainty in the crossing time stretching that impact point out into an impact line.
So now you have a probability of impact for the next few close approaches and a rough idea of possible impact locations for each.
Further observations over time will narrow all of that down as the orbit of the asteroid becomes better known.
Ultimately you’ll have a very reliable probability of an impact in the next few close encounters.
But beyond that the uncertainties blow up.
So, what did you discover?
Do we need to launch the nukes?
Organize an apocalypse party?
I have no idea.
This whole scenario is just made up.
Feel free to choose your own ending.
But the process I just described is more or less what we do to turn a newly noticed moving smudge into an impact probability.
And it’s more or less what was done with the most recent scale, 2024 YR4—although in this case by the ATLAS robotic survey, not by some astronomer who happened to look at the right spot.
Anyway, as I said, simulations based on our first set of observations revealed a 3.1% chance of a December 2032 impact by 2024 YR4.
The impact corridor was equatorial, and included parts of Africa, India and South and Central America But as further observations were made, the uncertainty in its orbital parameters dropped, and so too did the number of simulated trajectories impact the Earth.
By March this year NASA revised the impact probability to 1 in around 5500, while further analysis by ESA set the chance to nearly one in millions.
It’s actually more likely that the rock will hit the Moon in 2032 than it will the Earth.
For the time being, 2024 YR4 isn’t a threat.
Which is good, because this thing is between 40 and 90 meters diameter, which is around the size of the Tunguska impactor.
It could make a mess.
A more perennial threat is the asteroid Apophis, which is nearly half a kilometer long.
Big enough that it gets to have a cool name - in this case after the Egyptian god of darkness and chaos, which would certainly ensue if Apophis decided to descend upon the Earth.
If it ever hits, it would be with an energy just shy of all existing nuclear weapons, or several Krakatoa eruptions.
That means regional devastation and tsunamis for an ocean hit, as well as some years of global cooling from dust and ash in the atmosphere.
Apophis is too small to be an extinction level impactor—for example, the object that ended the dinosaurs 66 million years ago was around 20 times larger than apophis.
Still, we should be happy that Apophis won’t hit us any time soon.
Although when it was first discovered in 2004, early observations and orbital simulations indicated a 2029 impact chance of up to 2.9%.
Further observations, including radar by Arecibo, slowly downgraded the risk for a 2029 impact, instead indicating a possibility of passing through an 800m wide gravitational keyhole for a 2036 impact.
Happily, yet more observations show that’s no impact is likely for the next century.
Still, Apophis is a great test case for planetary defense.
As I mentioned in the recent episode, that’s one of the reasons the newly renamed OSIRIS-APEX will intercept Apophis on its now-benign 2029 encounter—in particular to assess how its interaction with Earth’s gravitational field changes the structure and rotation of the asteroid—all of which are important for future simulations of NEO trajectories following a close encounter with Earth.
As far as we know, there are no giant space rocks about to drop into our atmosphere.
The fact that we’re still finding new Earth-crossing objects is a bit worrying.
However the Rubin observatory’s LSST survey that starts this year will find many of the missing ones, NASA’s Near Earth Object Surveyor to launch in 2027 or so is supposed to find at least ⅔ of objects larger than 140m.
In 2005 the US Congress mandated NASA to find at least 90% of the estimated 15,000 NEOs larger than 140 meters.
So, you know, we’re getting there.
At least we know that when we do spot a new space rock hopping across the grainy starfield of our survey images, we know exactly how to figure out its potentially Earth-bound trajectory through space time.
Thank you to DeleteMe for Supporting PBS.
DeleteMe was started to help protect its members deal from the work of data brokers – these brokers employ web crawlers to scrape public databases and social media platforms, cross-reference this information with purchased commercial records, and compile complex relational databases of personal data that they then sell.
Whether the risk is doxxing, harassment, or identity theft, DeleteMe was created to help.
DeleteMe uses advanced monitoring tools to locate your data across hundreds of broker sites.
They compile a custom report showing where your information is exposed, and then submit formal requests to remove it.
They keep scanning to catch reappearances over time.
And because data brokers frequently cross-reference related individuals by enrolling everyone in a family plan, DeleteMe membership can help ensure that all your relatives’ personal details are included in the scanning and removal process, reducing the overall risk for the entire family.
You can get 20% off any DeleteMe consumer or family plan at joindeleteme.com/spacetime with promo code spacetime.
There’s a link in the description.