# What's an astronomical event which happened about 10,000 years ago, which humans could have noticed?

Bob is older than he looks.

Specifically, even though Bob looks like he's in his twenties or thirties (it's hard to tell), he was actually born in Europe about ten thousand years ago, long before the first word was ever written down.

By coincidence (or not), on the night (or the day) of Bob's birth, something unusual happened in the sky. Several people noticed, and thought that it was an important omen. So, Bob grew up knowing that he was born on the day that this particular event happened.

Furthermore, with the development of astronomy and computing, Bob has finally been able to determine the exact date that this event occurred, and thus, his date of birth.

What could the event have been?

Does it have to be a one-time thing or is it something that could have happened repeatedly? It can be something that happens repeatedly. A total solar eclipse that happened around that time would fit all my criteria.

Just how visible was the event? Visible enough that someone living in Europe ten thousand years ago plausibly could have noticed it and pointed it out to other people. Not necessarily visible enough that multiple people would have noticed it independently.

How precisely does Bob need to be able to determine the time that the event occurred? Down to the day. Some astronomical events (like supernovas) can't be calculated back with that amount of precision, so those wouldn't be suitable. Other events (like solar eclipses) can.

How much information does Bob already have about the date? I figure that Bob knows the season in which he has born, and he has also calculated the exact number of years since he was born. So he already knows the season and the year; he just wants to pin down the exact day within that season.

Are you looking for an actual historical event? Yes, I'm looking for the date that an individual, real astronomical event actually happened.

• Unrelated, but elements of this plot remind me of The Man From Earth (2007): "John's man would have learned as the race learned..." Dec 20 '18 at 22:52
• Most of the astronomic computations have been known for millennia. Bob could probably have computed his date of birth quite a long time ago (if that matters for your plot), especially if based on things such as solar or lunar eclipses or movements of the planets of the solar system. If you want something that he can only compute nowadays (and you would have to define nowadays), it's going to limit the acceptable events a lot more. Dec 21 '18 at 10:24
• Does it have to be the night of Bob’s birth, as specified in the question? If it does, then solar eclipses are out, since they can obviously only occur during the day. If the day of Bob’s birth is also acceptable, perhaps you can edit the question, and we can have the rare occurrence of a question here being edited to validate existing answers rather than to invalidate them. Dec 21 '18 at 18:40
• @MikeScott Edited. Dec 21 '18 at 18:46
• 1) if you seek for such realism, why it's exactly 10000 years? that's quite an anniversary for Bob Dec 21 '18 at 23:49

Evaluating the various options for astronomical events:

• Solar eclipse: This is probably your best bet. Total solar eclipses are brief events, lasting only a few minutes. They're also highly predictable, so you could figure out the timing of a past eclipse with sub-second precision. The problem is going to be figuring out which eclipse you saw: since virtually the entire Earth sees at least one eclipse per millenium, you need to know both roughly when and where a specific eclipse was seen. If there's something unusual about the eclipse (eg. that it was both preceded and followed by lunar eclipse), that would help with narrowing it down.

• Lunar eclipse: Too many of them. There are between two and five lunar eclipses a year, they're visible from half the Earth, and there's not much to distinguish one from another.

• Supernova: The initial rise in brightness is certainly fast enough to get your desired one-day precision: SN 1006 and SN 185 were both recorded in the Chinese chronicles as stars that were there one day but not the day before. Earth-visible supernovas are rare enough (about one per thousand years) that you'll have no difficulty figuring out which one it was if you've got even a vague idea of when you were born. The problem here is dating: the Vela remnant, for example, dates from sometime between 11000 and 12300 years ago.

• Comet: Doesn't meet any of your criteria. Matching up "great comets" in the historical records to 6300+ known astronomical bodies is highly error-prone, and comets are visible for months at a time. For a comet 10,000 years ago, getting a correct match is hopeless.

• Meteor impact: This certainly gets the precision you want, with an event lasting just minutes. The problem, like with supernovas, is dating: we can predict future impacts to less than a minute, but past impacts can only be dated through geology, with the attendant lack of precision.

• Meteor near-miss: An asteroid grazing the upper atmosphere has the same brief duration as an impact, and leaves the it intact for backwards prediction of its trajectory from modern observations. The problem here is that the small size of an asteroid means it's subject to perturbations of its orbit that can be ignored when back-predicting the movement of a planet or moon. This is something of a long shot, especially compared to a solar eclipse, but it might work.

• I think a total solar eclipse is probably my best bet. I'll say that Bob has kept track of his age in years as well as the season of his birth, so that will allow him to figure out exactly which solar eclipse it was. Now all I need to do is find a total solar eclipse that happened about 10,000 years ago and would have been visible in Europe. Dec 21 '18 at 3:13
• You could also have a combination of the above phemomena: a total solar eclipse right after a lunar one while a comet blazed brightly on the newly formed meteor crater would be quite easy to date and quite the Omen. Dec 21 '18 at 8:05
• "Sub-second precision" for solar eclipses is highly unlikely for the scale of 10,000 years as this answer maintains.Too many other bodies in the system add effects that accumulate exponentially. Dec 21 '18 at 14:52
• @TannerSwett I think you might be better off with the Vela remnant, because it's more remarkable. But it's your story. Dec 21 '18 at 18:53
• The really neat thing about a solar eclipse here is that it's spectacular. A civilization without an idea of astronomy would definitely notice it. Dec 21 '18 at 19:21

Perhaps one of the best astronomical options for you would be a rare planetary conjunction (e.g. two visible planets exactly intersect with each other, or, say all five visible planets come very close to each other, or adopt some specific arrangement, perhaps also involving the moon). Advantages:

• Software exists to make these predictions for you (e.g. http://shallowsky.com/blog/science/astro/predicting-conjunctions.html )
• Estimates likely have the sort of precision you need (unlike comets, estimates of past planetary orbits are probably precise enough over the 10000 year time frame you need).
• They were of note to the ancients, as the motions of the planets against the fixed stars were of great interest: unlike in modern society, it's likely that pretty much everyone would notice a rare arrangement, because pretty much everyone would be looking at the sky each night. Planetary motions were the television series of that era.
• Rare conjunctions or other unusual arrangement (like of all five visible planets) have already been used to date things like the start of ancient calendars and have been associated with specific past historical events (like the onset of the black death): http://adsabs.harvard.edu/full/1994JBAA..104..293D https://solarsystem.nasa.gov/news/133/astronomers-solve-ancient-mystery-of-the-chinese-calendar/
• If involving the more rapidly-moving inner planets (and/or the moon), they would provide the sort of duration you need (i.e. an arrangement that peaked on a specific day), rather than some of the other proposed candidate events. e.g. even though the onset of a supernova can be rapid, it is difficult to date accurately in retrospect, while other events (like the appearance of a comet) happen over a time span of months.

Another relevant Stack Exchange answer: https://astronomy.stackexchange.com/questions/11141/how-to-calculate-conjunctions-of-2-planets

• Playing around with the conjunction calculator you found, it looks like extremely-close conjunctions (a tenth of a degree or less) would work about as well as solar eclipses: there are typically one or two a century, with specific pairings being even less common. Mercury-Saturn looks like an especially good candidate: besides being symbolic, there's only one close conjunction in the next thousand years.
– Mark
Dec 21 '18 at 1:46
• @Mark Sounds to me like they would work better, since solar eclipses are highly dependent on location and these would be much less so. You'd need at worst a very rough idea of what latitude you were born at to go by the stars, I think. Dec 21 '18 at 2:28
• Bob was born in a hunter-gatherer, or early-agriculture society. Astronomy became important later to know the date, so one could know when to plant crops. It's not quite so obvious Bob's society would have been that interested. Dec 21 '18 at 8:27
• For a specific example of this, Jupiter was occulted by Venus when both were visible in the June evening sky shortly before 1AD. (The Stellarium date is -1/06/17, but the pyephem date is -2/06/17; ancient date conversions are tricky.) There is a story, still told annually, of an important person's birth around this time which was heralded by a "new star." But in the modern version of the story the "new star" was in the east and appeared near the December solstice. Details are tricky.
– rob
Dec 21 '18 at 15:07
• A major conjunction during a solar eclipse would be pretty memorable and portentious. Dec 21 '18 at 19:54

Solar eclipse is a good idea. They do not happen often, and are visible only from a small area (google "eclipse map"). If he knows approximate location of birth, all he needs is to check a list like one of those: https://en.wikipedia.org/wiki/Lists_of_solar_eclipses They do not go back to 10,000 years (that would be 80th century BC, i think), but it is obvious the math is there. You do not really have to pick an actual eclipse, it's not like any of your readers will sit down to do the computations.

Another idea is solar+lunar eclipse in same year, or two lunar eclipses in one year.

Comets are not a good idea since they arrive so frequently, and there is no way to tell which of many comets it was.

• I agree that a solar eclipse is likely to be the way to go, but I don't know where to find information about solar eclipses which happened around 10,000 years ago. Dec 20 '18 at 20:53
• Virtually every solar eclipse is paired with a lunar eclipse two weeks earlier or later, and two lunar eclipses a year is normal.
– Mark
Dec 20 '18 at 22:51
• "it's not like any of your readers will sit down to do the computations." - Not true at all. Someone will, and it's likely that they will post it all over TVTropes as another example of lazy writers making "science" up as they go along. Dec 23 '18 at 3:50

A volcanic eruption might be a good choice. Widely noticed and can be very accurately dated in the modern era by, for example, layer counting in ice cores from polar ice caps or lake sediments. There are not many volcanoes in northern Europe, but ash from volcanically active Iceland fell in northern Europe at various times. For example, the Vedde Ash originated in Iceland, is very accurately dated to 12,100 years ago, and occurs at a number of sites in Europe. There may be some others from Italian and Greek Island volcanoes in the appropriate time range. In some cases, perhaps in the ice core situation or for example if you have trees near the volcano where you can see the effect of the eruption in the tree rings, you could potentially determine in what season or month the eruption happened.

• That won't get you down to the required precision, however. You could get it down theoretically to the season by, say, cross-referencing ice core data with dendrochonology. It will give you an exact year, but you'd only be able to narrow it down to summer or winter, possibly to early or later in the season, but certainly not down to a specific date. Dec 21 '18 at 5:27
• Well, maybe it occurred in a particular season that co-occurred with one of the other events like planetary conjugations, comets or eclipses; so we get the rough date from the carbon date and the fine date from the astronomical schedule! Dec 21 '18 at 17:49
• Eruption + lunar eclipse should do it. Ice cores or dendrochronology would get you the year, simple memory would give you the season, and the eclipse would give you the day.
– Mark
Dec 21 '18 at 21:02

We need an event which...

1. happened 10,000 years ago
2. visible from Europe to the naked eye
3. is pretty unique
4. looks the same to all of Europe
5. can somehow be dated down to the day

I don't think this is possible. Make that the story.

Why does Bob need to know the exact day he was born on some calendar that didn't exist when he was born? Make the story about Bob's journey from obsessing about his place in history, to coming to accept it's going to always be a little ambiguous.

Maybe Bob has counted solar eclipses since he's been born. He gets very excited about the new science of astronomy in the Renaissance, finally they answer his question! The Renaissance astronomer shows him the fascinating new models of the Solar System and how he can predict eclipses forward and backward. Yes, they give Bob a date, but note it's an approximation. Disappointed, Bob continues his search for his birthday. Maybe he learns a ton of math, astrophysics, and archeology in the process.

Fast forward to modern times. Bob is excited about super computers and the ever increasingly accurate models of the Solar System. He talks to a modern astronomer who says yeah, they can calculate Bob's eclipse down to the minute! But Bob gets the same speech; don't be seduced by the precision of the computer model. It's still an approximation. While the answer might seem precise there's still error bars of days. They can't give Bob any assurance this is very precise looking calculation is really Bob's birthday.

Bob, frustrated and angry after searching for centuries, lets loose on the astronomer. What good are all these computers and all this precision all this math and science if they know every answer is wrong?!

He tells Bob all the imprecision and approximations in his eclipse counts. He has to trust that Bob remembered all of them correctly. That it was never cloudy on an eclipse day. And that what's a full eclipse in one part of Europe might be a partial in another.

Even if Bob's observations were correct, the astronomer explains chaos theory as it applies to astrophysics. No matter how powerful the computer there will also be some details they're leaving out of the model, details that could effect the timing of Bob's eclipse.

Bob finally realizes that science isn't about the search for certainty. Religion provides certainty, but it does so by over-simplifying and detaching itself from messy reality. Science accepts reality for what it is, and they need to figure it out through a million imprecise observations. A good scientist accepts their answers are inaccurate while continuing to chip away at those inaccuracies. A good scientist knows the math is a model of reality, not reality itself. A good scientist knows their job isn't to find The Truth, but to produce answers which are good enough to get the job done.

The modern astronomer says the calculations are done, "would you like to know when the computer says you were born?" "No", says Bob, "I already know it well enough." Bob digs out a scrap of parchment from his notes and shows it to the astronomer. It's the date the Renaissance astronomer gave him centuries ago. "You're invited to my birthday party."

Maybe it could be ash from a volcano causing the full moon to appear ominously red. If Bob already knows the year and season, knowing it was a full moon will be enough to pin down the exact day.

A possible candidate volcano would be the Grímsvötn eruption in Iceland, 8230 BC.

Sodom and Gomorrah are good candidates. No proposed explanation was accepted so volcanic activity is unlikely. Maybe a meteor shower?

A much more recent one (out of your time scale but may have happened earlier many times) is the famous SN_1054. (Abbreviation of Super Nova of the year 1054), in the crab nebula, observed by Chinese astronomers the same year.

Here are more ancient records.

• Sodom and Gomorrah are horrible candidates, because there's zero non-Biblical evidence for them.
– Mark
Dec 20 '18 at 22:56
• Trinity Southwest University is a Bible college that abides by the Chicago statement that the Bible is inerrant and infallible. In other words, Collins and his co-workers already decided there was a Sodom and Gomorrah because the Bible said so and the Bible is never wrong about anything, thus they're only only looking for evidence to confirm what they already claim to be true. In other words, not precisely what I'd consider unbiased reliable sources. Dec 21 '18 at 16:21
• If Bob said "Yeah, I remember them. I was in the next city over when they done got smote. Shame. Great parties", the Bible gives no date for Sodom and Gomorrah. And in Biblical "history" the Earth is only 6000 years old, we need 10,000. Dec 21 '18 at 18:55
• @Schwern actually, that's Bishop Ussher's timeline of the Bible (with Creation in 4004 BC). Not all believers follow it. Dec 23 '18 at 3:52

The saros is a period of 18 years, 11 days, 8 hours (or 6585.3211 days), and represents the amount of time after some given eclipse when an identical eclipse will happen again.

As with any estimate it's not completely accurate, but we can and have estimated many lunar and solar eclipses to the day.

As for Bob's birthday, 555 saros periods would represent about 10,006.441 years. Legends state that he was born on a total lunar eclipse when the moon was descending, so if he starts from the date that type of eclipse last happened and puts in the effort to take into account leap days he could make a pretty good guess as to when he was born.

1Just a rough estimate, (6585.3211 * 555) / 365.25 = 10006.4427392 years

• A saros series lasts no more than 1550 years. The eclipse on Bob's birthday long pre-dates any current series.
– Mark
Dec 20 '18 at 23:00

Hale–Bopp (images) last seen 1997 when it was visible for 18 months (the last time it came around before that was about 4200 years ago), it was very visible & a very impressive sight for a long time & would certainly have been noticed each time it's visited in the past.

I had a short cruise around Google to try & find other previous dates it visited to try & find you the one closest to your 10,000 year mark but not a lot of joy on that so far.

By far the most impressive space related event I've witnessed myself & hung around long enough it couldn't possibly be missed so I'm convinced this would be an excellent choice for what you want.

You want a single day so how about the first day it would have been visible to the naked eye, Bob is born on a night when a new small pin prick of light is visible in the night sky (it's noted but no one really thinks that much about it), it grows over the following weeks & months into a really impressive sight, everyone decides Bob must be special, then it slowly fades & disappears from the night sky.

Of course maybe it wasn't the first day it was visible to the naked eye, maybe it was just the first day anyone in his tribe noticed this new object growing in the night sky.

Which means Bob will never know for sure the exact day he was born.

Is it absolutely integral to your plot that Bob has to be able to pin down exactly the day of his birth in the modern calendar? because I don't think that's going to be possible with any astronomical event.

Of course you could fix this with some additional information, he was born on the Summer (longest day) or Winter (shortest day) Solstice of a year the comet visited & it's been back twice since then.

That ^ would give you an exact date.

It doesn't have to be on the exact day of either Solstice either, simply knowing an exact number of days before or after either Solstice or Equinox that he was born along with a notable astronomical event to identify the year with will give you an exact date.

• "because I don't think that's going to be possible with any astronomical event" solar & lunar eclipses are just too common for you to pin it down to any specific one that far back in time & pretty much everything else that might qualify is visible for days, weeks or months. Dec 22 '18 at 16:36
• Comets don't come with labels. How does Bob know that it was Hale-Bopp that was visible and not, say, a long-period comet that hasn't yet come around for a second pass? Or worse, a comet on a hyperbolic orbit that isn't going to come around for a second pass?
– Mark
Dec 22 '18 at 20:35
• @Mark : Bob doesn't, but Bob probably does know within a certain time frame when it was, comets aren't that common, really bright impressive ones (as viewed with the naked eye) as apposed to faint smudges in the sky you have to squint to see that if your note familiar with the stars you wouldn't even be sure was something new (like Halley's comet last time it came by) are really quite rare. Dec 22 '18 at 20:57
• ^ if Bob can tell an astronomer very roughly when it was with a description of how bright it was & how long it was in the night sky then even with a window as vague as 400 years or so chances are more than good said astronomer can narrow the likely suspects to just one. Dec 22 '18 at 20:59
• ^ of course if it just passed through from outside the solar system your going to be buggered when it comes to identifying which one it was, just thought of that <scratches head> :/ Dec 22 '18 at 21:03

I'm gonna go with Halley's comet.

It's visible from the Earth, might have been recorded as 467 BC. So it has probably been cycling for longer.

It has a cycle of 75.32 years, and was last seen in 1986. So, if I can still do some basic maths, it passed by 9974 years ago.

It can be seen with the naked eyes from the Earth, can be easily noticed and can be well dated.

• "There was a large, bright comet the month before and after my birth, but with orbit changes over this kind of time frame, there's no way to be sure which one, or even if it was a short-period repeater (vs. a long-period, seen once in millennia like Kohoutek). Dec 20 '18 at 20:42
• Well, the Wikipedia article about Halley's comet says that "Researchers in 1981 attempting to calculate the past orbits of Halley by numerical integration starting from accurate observations in the seventeenth and eighteenth centuries could not produce accurate results further back than 837". That makes me think that we probably don't know the date that it passed by about 9,974 years ago. Dec 20 '18 at 20:51
• There are 6300 known comets, with more being discovered every year. Halley's Comet is merely one of the brighter ones. If Bob was born during the passage of a Great Comet, there's no good way to figure out which comet it was. (Further, comets lasting more than 50 close passes of the Sun are rare; the brighter a comet is, the less likely it is to survive.)
– Mark
Dec 20 '18 at 23:05

Perhaps a Meteor Air Burst. Though they're anything but regular normally, they can only be observed from certain areas and you could arrange for their regularity.

Though its exact cause is still up for debate, the Tunguska Event was highly-noticeable and is perhaps the most famous example.

Physically-destructive events often leave a great deal of evidence. Perhaps a town or simply a building was destroyed, buried, and since uncovered with some easily-dated items inside

• propaganda from or art depicting a ruler
• precision tools or weapons showing the metallurgical science of the day

This could show up as exciting science news, reaching Bob via modern means, but not (yet!) have been discovered today.

• How do you propose to date the air burst?
– Mark
Dec 20 '18 at 23:09
• The best dating technique we've got for 10,000-year-old objects, dendrochronology, has an error margin of a year.
– Mark
Dec 21 '18 at 1:58
• @Mark With the caveat that the oldest known tree is currently about 5000 years old. Doubling that seems unlikely, especially with only 15 known to have exceeded 2000 years.
– Bob
Dec 21 '18 at 3:54
• @Bob - You don't need standing trees for dendrochronology. Branches on the ground work too. Wikipedia says the "Currently, the maximum span for fully anchored chronology is a little over 11,000 years B.P.", so depending on where the other Bob was born, dendrochronology could work (but, of course, this still doesn't give the day of the month). Dec 21 '18 at 8:20
• 10,000 years ago is a long time. It is about the start of agriculture. There may have been some towns, but that is a subject of scholarly debate. Dec 21 '18 at 8:24