Canon discussion / Essays

Astronomy in the Harry Potter Series


Within the Harry Potter stories are many scenes that can be analyzed astronomically. As both an astronomer and a teacher, these interest me, and so I will present such analysis here. My goal is not to criticize Rowling when she is inaccurate in her astronomy (or, for that matter, to praise her when she gets it right), because she is clearly not concerned with making her tales scientifically accurate—she’s writing a story, not an astronomy textbook.

“Mars is bright tonight,” Ronan repeated . . . . “Unusually bright.” (PS15)

As seen from Earth, Mars varies greatly in its apparent brightness, due to both the changing distance between the two planets, and the changing amounts of sunlit surface Mars presents to us. Roughly every 26 months, the Earth is closest to the red planet, and Mars looks brightest; at these times Mars is said to be in opposition.

However, according to the Lexicon Timeline, Ronan’s remark about Mars’s unusual brightness was made in late May of 1992, about 7.5 months before its January 1993 opposition. On that night in the Forbidden Forest, Mars was in fact only at about 10% of its maximum brightness[1]—hardly “unusually bright.”

[In Diagon Alley, Harry] was sorely tempted, too, by the perfect, moving model of the galaxy in a large glass ball, which would have meant he never had to take another Astronomy lesson. (PA4)

A perfect model of our Milky Way Galaxy—the enormous conglomeration of hundreds of billions of stars to which our Sun belongs—would be very cool indeed, but as Professor Sinistra’s class appears to focus only on our solar system (just the Sun, its planets, and their moons), it is hard to understand why this galactic model would be so helpful to Harry. If the glass ball containing the scaled-down Milky Way was three feet in diameter, our solar system would be an invisible speck within it, about a millionth of an inch in size. Perhaps, when Rowling was describing this item, she confused the terms “galaxy” and “solar system”—a common error.

“Did you check the lunar chart and realize that I was always ill at the full moon?” (PA17)

The average length of time from one full moon to the next is 29.53 days. Three dates when we know Lupin was ill are November 5, December 25, and June 6. However, since those dates are not multiples of 29.53 days apart, they cannot all have been full moons.[2]

“the hippogriff Buckbeak, hereafter called the condemned, shall be executed on the sixth of June at sundown” (PA21)

The time of that fateful sunset can be estimated by using the events in the story. At 11:55 PM that night, Harry and Hermione went back in time by three hours. They heard their earlier selves crossing the entrance hall with Ron and leaving for Hagrid’s cabin, walking slowly under the Invisibility Cloak (PA21). If it took the Trio fifteen minutes to cross the grounds, then they arrived at Hagrid’s at 9:10 PM. According to Lupin, they were in Hagrid’s hut for twenty minutes (PA17), and therefore left around 9:30 PM. The execution party arrived at the same time; they read the official notice, signed it, walked outside, and saw Buckbeak gone; then the executioner threw his axe (PA21). If all of that took five minutes, then the axe was thrown at approximately 9:35 PM. At that moment, “the very last rays of the setting sun were casting a bloody light over the long-shadowed grounds,” (PA17) so it was just before sundown.[3]

At first glance, this might seem like an unusually late sunset, but we must remember that (1) Britain is on Summer Time in June (what Americans call Daylight Saving Time), (2) the date is just before the summer solstice, the longest day of the year, and (3) Hogwarts is quite far to the north, in Scotland (according to this web site referenced by the Lexicon).

Mainland Scotland lies roughly at latitude 55-58º N and longitude 2-6º W. On June 6 in that region, sunset ranges between 9:40 PM and 10:20 PM British Summer Time[4], so the timing roughly agrees with Rowling’s chronology (the agreement gets worse the farther Hogwarts is to the north and west).

“Jupiter’s biggest moon is Ganymede, not Callisto,” [Hermione] said. . . . , “and it’s Io that’s got the volcanoes. . . . Europa’s covered in ice, not mice . . . .” (OP14)

This information about Jupiter’s four largest moons is indeed correct.

A twinkling red star winked at [Harry] from overhead. . . .  “Mars, bringer of battle, shines brightly above us.” (OP27)

Mars does look like a red star to the naked eye, but since it is a planet, it does not appear to twinkle as the stars do. Also, the Lexicon Timeline estimates that Firenze’s first Divination class happened in early March of 1996, which was midway between Mars’s 1995 and 1997 oppositions. At that time, Mars was shining with only about 10% of its maximum brightness.[5]

Harry’s O.W.L. Astronomy practical (OP31)

In the discussion that follows, I will assume that this exam began at 11 PM British Summer Time[6] on June 24, 1996,[7] and lasted for 90 minutes.[8] I will also (somewhat arbitrarily) place Hogwarts about midway between Edinburgh and Aberdeen, at latitude 56.5º N, longitude 2.5º W. Using these parameters and the online interactive planetarium Your Sky (implemented by John Walker), I have generated maps showing which stars would have been visible in the sky at the beginning (Figure 1.) and end (Figure 2.) of the exam (the moon and the planets are not drawn, as they change their positions from year to year).

Beginning of O.W.L.s - 2300
Figure 1.: Beginning of O.W.L.s – 11 PM

End of O.W.L.s - 0030

Figure 2.: End of O.W.L.s – 12:30 AM

. . . a perfect night for stargazing, cloudless and still. The grounds were bathed in silvery moonlight . . .

There are two reasons why this night would in fact be less than perfect for stargazing. First, if the grounds were bathed in moonlight, then the sky would have been quite washed out, making it difficult to see faint stars. Astronomers usually try to observe when the moon is below the horizon, or barring that, when the moon is not very full.

Second, the summer sun would have set around 10 PM,[9] only an hour before the beginning of the exam. At 11 PM, the sky would have still been aglow with twilight, and only the moon, the planets, and a handful of the brightest stars could have been seen with the naked eye. Indeed, at that latitude near the date of the summer solstice, the sun never gets lower than ten degrees below the horizon, so the sky is never truly dark.

As Harry completed the constellation Orion on his chart . . .

Orion is a constellation best seen in December. In June, the stars of Orion are almost directly behind the Sun as seen from the Earth. Consequently, at that time of year it is impossible to view Orion—not only in Scotland, but almost anywhere on the planet.

Harry put his eye back to his telescope and refocused it, now examining Venus.

Harry seems to have observed Venus a bit more than an hour into the exam—that is, just after midnight. Since Venus is usually only visible either shortly after sunset or just before sunrise (due to its proximity to the Sun), my initial reaction was that Rowling was describing an impossible scenario. However, in a letter to the amateur astronomy magazine Sky & Telescope,[10] astronomer Kevin Krisciunas pointed out that this is indeed possible in certain years.

Since Krisciunas made his calculations for a slightly different location in the UK than what I have assumed, I have done my own investigation (although the results are not much different). It turns out that in 1991, Venus was visible in the western sky at 12:05 BST on June 25, albeit just a fraction of a degree above the horizon. (Here [Figure 3.] is a simulated view using Your Sky.) As seen from Earth, Venus’s motion in the sky repeats itself every eight years, so this late Venus could also be seen from Hogwarts on the same date in 1999 or 2007. Depending on Hogwarts’s exact location, it might have also been just barely possible to see Venus at that same time and date in 1994 and 2002. (It was not possible in 1996, however.) Thus, the situation that Rowling describes is indeed astronomically possible: Venus can sometimes be seen at midnight.

Venus during the O.W.L.s
Figure 3.: Venus during the O.W.L.s

. . . Harry . . . noticed that he had mislabelled Venus as Mars.

It’s a good thing that Harry noticed, because this is quite a mistake to make: Venus and Mars are very different in color, as well as brightness. Venus appears bright white in the sky, while Mars is reddish and somewhat dimmer.

. . . [Harry] saw, for mere seconds, a vision of the main street in Hogsmeade, still dark, because it was so much farther north.


Harry has this vision at Shell Cottage, which is located “on the outskirts of Tinworth” (DH23), and is therefore (according to Bagshot’s A History of Magic) somewhere in Cornwall (DH16), along the extreme southwestern coast of England. According to DH24, the sun rose at Shell Cottage just before Harry spoke to Griphook and Ollivander, and it rose at Hogwarts (in Scotland) just after these interviews. Perhaps a half-hour passed between the two sunrises, and Rowling explicitly states that it rose later at Hogwarts because it was farther north than Shell Cottage. Does this work out astronomically?

To answer this question, we need to know the date when Harry’s vision took place. Fortunately, Rowling has given us enough clues to determine that the sunrise in question happened between March 14 and April 1.[11] During this period, which is roughly within a week of the spring equinox, two locations that lie on the same line of longitude will experience sunrise at more or less the same time. At the latitude of the U.K., the more northern location will see sunrise only a few minutes later during the week before March 21, and a few minutes earlier during the week after. Thus, it seems that Rowling has made an error—the fact that Hogwarts is farther north than Shell Cottage cannot explain why its sunrise is a half-hour later.

The time of sunrise also depends on longitude. Close to an equinox, when sunrise is insensitive to latitude, it happens four minutes later for every degree of longitude farther west. Therefore if Shell Cottage was significantly east of Hogsmeade (by about 7.5°), then the two sunrises could have occurred as Rowling relates, though for a different reason. However, since Shell Cottage is supposedly in Cornwall (longitude 4.5-6° W), it is in fact farther west than much of mainland Scotland (longitude 2-6° W). So unfortunately, we cannot reconcile Rowling’s sunrises with either astronomy or geography.

It was a relief when six o’clock arrived and they could slip out of their sleeping bags, dress in the semidarkness, then creep out into the garden. . . . The dawn was chilly, but there was little wind now that it was May. Harry looked up at the stars still glimmering palely in the dark sky. . . .


Is it possible for the stars to still be visible at shortly past 6 AM BST on a May morning in the U.K.? Alas, no: sunrise in that part of the world is always before 6 AM at that time of year.[12]


[1] This was determined with the Freeman edition of the astronomical software Starry Night Backyard, ©2000 Canada Inc.

[2] November 5: First Hogsmeade weekend is Sunday(?), October 31 (PA8); Gryffindor vs. Hufflepuff match is next Saturday, November 6; Snape substitutes for Lupin the day before (PA9).

December 25: “I’m afraid the poor fellow is ill again . . . . Most unfortunate that it should happen on Christmas Day” (PA11).

June 6: “the hippogriff Buckbeak … shall be executed on the sixth of June” (PA21); Lupin transformed that evening.

See also Troels Forchhammer’s essay “Mapping the Harry Potter Timeline,” where he does basically the same analysis.

[3] See also Hollydaze’s “Time Line: The End of PA.”

[4] This was determined with the US Naval Observatory’s online sunrise/sunset calculator..

[5] Starry Night Backyard, Freeman edition, ©2000 Canada Inc.

[6] OP31 gives the start time as 11 PM, and all of Britain is on Summer Time in June.

[7] The Lexicon Timeline guesses that O.W.L.s began on either June 8 or 15, which would place the Astronomy practical (nine days later) on June 17 or 24.  I will adopt the latter date for the following reason.  In Half-Blood Prince, we find that “Harry remained within the confines of the Burrow’s garden over the next few weeks,” and then he left the Burrow for Diagon Alley a few days after his July 31 birthday (HBP6); thus he arrived at the Burrow around July 10.  On that day, Dumbledore asked him, “I gather that you have been taking the Daily Prophet over the last two weeks? … Then you will have seen that there have been not so much leaks as floods concerning your adventure in the Hall of Prophecy?” (HBP4)  This adventure therefore took place around June 26, and the Astronomy practical was two days before that.

As a further check, in HBP1 Fudge visited the Muggle Prime Minister on a day “in the middle of July.”  During that visit, he said, “I was sacked three days ago!  The whole Wizarding community has been screaming for my resignation for a fortnight”—presumably since Fudgefirst announced the return of Voldemort two days after the Astronomy practical.  Thus, if the practical was on June 24, Fudge’s visit would have been roughly two weeks and five days later on July 13, which is indeed near mid-July.

[8] “Half an hour passed, then an hour,” then Harry sees Umbridge and her crew cross the grounds and enter Hagrid’s cabin.  Shortly after that, the examiner calls “twenty minutes to go.” (OP31)

[9] This was determined with the US Naval Observatory’s online sunrise/sunset calculator.

[10] Sky & Telescope, “Rowling Gets It Right,” p. 12 of December 2003 issue (Vol. 106, No. 6), Sky Publishing Corporation, Cambridge, MA.

[11] The relevant timeline is as follows: on the night Ron managed to tune in Potterwatch (DH22), the Trio were captured by the Snatchers and brought to Malfoy Manor (DH23). The sunrise in question was the following morning.

We are told that the Potterwatch broadcast took place on an evening in March, so the sunrise could have been no later than April 1. We also know that Draco was on Easter holidays at the time. The earliest possible date of Easter is March 22 (see the U.S. Naval Observatory’s article “The Date of Easter”). According to Diana Summers’s essay “British Schooling in the 1970s,” Easter holidays always begin two Fridays before Easter Sunday, so they never start earlier than March 13. Assuming that Hogwarts is on the same schedule, the sunrise in question could therefore have been no earlier than March 14.

[12] This was determined with the US Naval Observatory’s online sunrise/sunset calculator.


Pensieve (Comments)

Tags: geography moon