The abstract from the paper[1] linked to from the article is a good read and short, so I think it is worth quoting here:
"The existence of a giant planet beyond Neptune -- referred to as Planet Nine (P9) -- has been inferred from the clustering of longitude of perihelion and pole position of distant eccentric Kuiper belt objects (KBOs). After updating calculations of observational biases, we find that the clustering remains significant at the 99.6\% confidence level. We thus use these observations to determine orbital elements of P9. A suite of numerical simulations shows that the orbital distribution of the distant KBOs is strongly influenced by the mass and orbital elements of P9 and thus can be used to infer these parameters. Combining the biases with these numerical simulations, we calculate likelihood values for discrete set of P9 parameters, which we then use as input into a Gaussian Process emulator that allows a likelihood computation for arbitrary values of all parameters. We use this emulator in a Markov Chain Monte Carlo analysis to estimate parameters of P9. We find a P9 mass of 6.2 (+2.2, −1.3) Earth masses, semimajor axis of 380 (+140,−80) AU, inclination of 16±5∘ and perihelion of 300+85−60 AU. Using samples of the orbital elements and estimates of the radius and albedo of such a planet, we calculate the probability distribution function of the on-sky position of Planet Nine and of its brightness. For many reasonable assumptions, Planet Nine is closer and brighter than initially expected, though the probability distribution includes a long tail to larger distances, and uncertainties in the radius and albedo of Planet Nine could yield fainter objects."
Or is it just a matter of us not spending the time/effort to find it. I know time on instruments like Hubble is very, very tight and scheduled out years in Advance. But they have been talking about Planet X since like the 80s at least, and probably earlier.
Even Pluto was detectable relatively early and it’s on a crazy orbital plane and elliptical orbit.
It could, the relevant part of the whole thing is the object's mass (and therefore its gravitational influence) and its orbit.
Lex Fridman did a podcast interview a few months ago with Konstantin Batygin [0] about the topic of Planet Nine and he poses this exact question of whether or not it could be a black hole [1].
More specifically, he asks if it could be a primordial black hole which would make it a double discovery since primordial black holes are still only theoretical.
It won't even start to decay until space expands enough that it's warmer than the cosmological background, not that that's more than a rounding error to your calculation.
Counter argument against the black hole hypothesis includes the lack of powerful x-ray activity where the black hole should be. Also, its size would be quite unusual, which in that case could be a primordial black hole, which in itself is theoretical and never been observed.
If it were a black hole, which would necessarily mean a primordial black hole, we would never spot it. We would look, and look, and eventually give up and say it must be a statistical fluke. A coal-black body would also be hard to spot. Surprisingly many asteroids are black, for poorly-understood reasons.
We might not spot it from earth, but if we’re confident enough of where we’ll find it, its orbital parameters, its mass, a probe could be sent, and it could be detected through its SOI, and perhaps even direct observation.
Sure, ~300AU is a long way, but we have the technology, and it’s our closest black hole candidate by many orders of magnitude
If it is a choice between, (1) Spend $500M just to see whether there really is a black hole there, or (2) it is probably a just statistical fluke -- it is not hard to guess which will be chosen.
Probably. It narrows down the range of locations and therefore the search area. And given the potentially very low brightness of the object, search area and hence telescope time are limiting factors.
Probably not. This planet would be about 10 times as far away from us as Pluto, and estimated brightness is about 1/1000th that of Pluto. And its low velocity would make it much harder to pick out against the stellar background.
In Cartoons (and in science), wasn't this called Planet X and in other hypothesis 'Nemesis'?
I get that Pluto was demoted, but it's still a planet, if dwarfish, so it should still be Planet X (X as in ten, but X can also mean undetermined number, which is perfect).
I love the way this is written. It's like the author, as a scientist, is in the service of humanity. The tone, language, the tongue in cheek apology...
Something very noble and approachable to it. Good job. Find that planet for us! We appreciate it.
The writing is great, but I keep wondering: Why do so many physicists use Jet or other non-perceptually uniform rainbow-like colourmaps for their graphs? Shouldn't that be part of great science communication, given that about 1 in 12 men have some form of colour blindness?
There are so many other visually appealing options [0] that are attractive while still being perceptually uniform. There's even Google's Turbo [1] colourmap for those who refuse to give up rainbow-like colourmaps.
Short answer: because none of them is the default in standard plotting tool.
There is so many alternative that it is hard to know which replacement you should pick. If you pick the wrong one you are a moron, if you stick with the bad default it's way less noticeable, even if it is a strictly worse choice.
The default issue is certainly keenly felt. My first scientific publication graphs were done with Jet. But since switching to some of the perceptually uniform ones (most often Viridis), I've never had a complaint.
But your point "if you stick with the bad it's way less noticeable" is understandable.
Since it's showing a probability density, wouldn't it make most sense to just use a grayscale in this case? As the probability density approaches zero the visual would disappear as well. This is more intuitive, as the probability is zero off the edge of the graph/page
I think you only need this color vomit if you have a change between two states, or you have multiple intermediary states and want some visual "clustering".
You can actually see the problem very clearly b/c it ends up having a weird dark-blue to white cut is some of the plots.
The advantage of using a rainbow scale as opposed to a gray scale is that a rotation in hue space brings out more 'detail' than a simple black to white gradient.
I still think it's a bit visually confusing when you're representing the existence or non-existence of something (the planet is in this region, and not anywhere else)
You could in theory do both.. rotate in hue and fade to white. If you did it right, when if you printed it out in B&W you'd get the grayscale equivalent :)
That was a very interesting read. Thank you for sharing. I could really use some of the other color spaces presented too.. in some other situations.. I'll just need to figure out how to plug them into my workflow :)
It's strange but the first linear color space (CET-L03 red-yellow-white) seems to go to a lightness of 100. On the far right it looks white as expected. However the other one (CET-L20) looks very nice, but it doesn't seem to reach 100. Even though in the plots and explanation it suggests it does. Strange..
The more I look into this the more it seems this inevitably (accidentally or intentionally) misrepresents data. Even the author highlights how it does this. All the options are in some ways effectively thresholding/binning ... but in non-systematic ways through the vagaries of the human visual system. Unless this is an art project, there are more systematic and honest ways to bin things.
I can see how it can be useful or even inescapable in some situations.. The most obvious being maps with different content that needs to be distinct and of course these bimodal "divergent color maps". But maps are already half-art :) and bimodal data you're effectively vaguely "pre-binned" the data into two equal halves. There are also situations where the color is adding an extra dimension (like these "cyclic colour maps"). The result with the fingerprint is honestly hard to visually interpret, but I don't see any clear alternative (other than a grid of arrows I guess)
But reading all this stuff leaves me with the impression that colors are dangerous. They really should be last resort. In simple/common situation like the probability densities of the mystery planet - grayscale looks like it's the only really honest option. Side by side with the grayscale, the linear colors don't seem to be adding all that much and are creating artificial "islands" (not to mention the colors are a real blast to the eyeballs)
Veritasium did a video about Planet 9 [1] with one of the authors of this article, Konstantin Batygin. It also features David Jewitt with opposing arguments regarding the existence of Planet 9.
It sounds like the opposing arguments are: (1) if you want something to exist, you'll be biased in your observations, (2) we might simply be unaware of a bunch of Kuiper objects, and those unknown objects might be exerting their influence, and (3) we haven't actually seen P9 yet, visually, so skepticism is healthy until then.
Yes, thank you. It's been a while since I've watched the video, so calling it "opposing arguments" was a stretch. More like a healthy dose of skepticism.
> We find a P9 mass of 6.2 (+2.2, −1.3) Earth masses
This wouldn't even be half the mass of the smallest gaseous planet, but would be much larger than the largest terrestrial planet. Which do astronomers think it would be?
Yes, and modern methods would be really easy to incorporate because of awesome open source libraries.
If you got really lucky, there could be public data from telescopes that had already recorded whatever regions of space were of interest, and evidence of planet nine might have been disregarded as noise.
Is this something NNs could do? The output might just be a probability field of orbits, which would not tell you anything more than known analytical methods would.
People are being conditioned to apply NNs to every problem, regardless of their suitability. Having to only think about a single tool makes everything easier.
(No, not the original Bode's Law, which was just a first approximation! Mathematician Mary Blagg did the real work in the early 20th century, but was mostly unacknowledged until nearly 70 years later.)
It works, not just for the solar system, but also for the satellite systems of Jupiter and Saturn; but no one knows precisely why...
Unless they were a double-planet, similar to Pluto/Charon. (Those orbit a common center not inside either body.) To my knowledge, they are unique, in that, among known bodies, though it would not be surprising if there were a few tiny double asteroids.
Reminds me of a documentary (which I can't find anymore) about New Horizons and the hunt for Ultima Thule, where the scientists did the exact same thing!
The Hill Sphere also changes over time as stars pass us. The galaxy is more of a soup - the stars in the neighborhood don't have a static position to one another.
...so I would imagine that anything even 1ly out would be unstable, over several hundred million years. Solar systems may even exchange very outer outer "planets".
It seems to me that something that's a couple of lightyears out can't have any sort of stable orbit. I suppose that an Oort cloud essentially merges with an adjacent star's so there's no hard definition.
The only physical evidence we have they exist are comets that occasionally show up in the inner solar system where we could observe them which have orbits that are not interstellar but would put them as going incredibly far out. From that we can extrapolate that they likely exist around every star but there is no way we could verify that with our current technology since any light reflected off the objects would be far too faint to detect.
What if it's a planet temporarily captured by our solar system which is currently deataching due to an unstable orbit? Moons are temporary, why not planets, even suns.
Orbits don't just detach. Even at the predicted aphelion distance of 460 AU, P9 would have an orbital velocity of 1.4 km/s and a solar escape velocity of 2 km/s. So to detach from solar orbit you would need a minimum delta-v of 600 m/s. 600 m/s on a planet size object is only going to happen if a star happens to wander through the solar system.
It's possibly a transient object - ie not in orbit around the sun, but just passing through.
For all we know it formed billions of years ago in another system, and then was ejected from its home system from some near miss with its sister gas giant.
That looks like an asymmetric error bar, meaning the best estimate is 6.2 but could be 2.2 above or 1.3 below.
Essentially it's a common physical scientist way of expressing a point estimate with a confidence interval. (In the life and social sciences they tend to just provide the point estimate and confidence interval explicitly.)
Seems like an uneven error interval? Meaning if it’s not our estimated valid of 6.2, we have less confidence in the upper bound than the lower bound, or our estimates are distributed nonuniformly
> Put these two plots together and you get a 99.6% chance that the objects are clustered, rather than uniform. That sounds pretty good to me.
That's not even three-sigma. I'm not sure if this is real evidence of a planet nine. Just a chance alignment of orbits seem like the most plausible explanation.
Three sigma is pretty damn good for something like this. It means it's not worth sending a probe out yet, but is more than enough justification for serious telescope time.
Say there is a planet out there, might it have a protective effect from things hitting earth, similar to how Jupiter is said to help reduce the chances of things hitting earth? Maybe it's too far away, but multi-body dynamics is complicated! Considering the previous lizard 'lords' of-the-earth for millions of years seem to have been largely ended by such a collision, maybe there's more we should be doing to detect and prepare for such an event?
"The existence of a giant planet beyond Neptune -- referred to as Planet Nine (P9) -- has been inferred from the clustering of longitude of perihelion and pole position of distant eccentric Kuiper belt objects (KBOs). After updating calculations of observational biases, we find that the clustering remains significant at the 99.6\% confidence level. We thus use these observations to determine orbital elements of P9. A suite of numerical simulations shows that the orbital distribution of the distant KBOs is strongly influenced by the mass and orbital elements of P9 and thus can be used to infer these parameters. Combining the biases with these numerical simulations, we calculate likelihood values for discrete set of P9 parameters, which we then use as input into a Gaussian Process emulator that allows a likelihood computation for arbitrary values of all parameters. We use this emulator in a Markov Chain Monte Carlo analysis to estimate parameters of P9. We find a P9 mass of 6.2 (+2.2, −1.3) Earth masses, semimajor axis of 380 (+140,−80) AU, inclination of 16±5∘ and perihelion of 300+85−60 AU. Using samples of the orbital elements and estimates of the radius and albedo of such a planet, we calculate the probability distribution function of the on-sky position of Planet Nine and of its brightness. For many reasonable assumptions, Planet Nine is closer and brighter than initially expected, though the probability distribution includes a long tail to larger distances, and uncertainties in the radius and albedo of Planet Nine could yield fainter objects."
[1] https://arxiv.org/abs/2108.09868