The equinox is not when day and night have equal lengths | Hacker News Reader

The equinox is not when day and night have equal lengths

scientificamerican.com

142 pointsBrajeshwar110 comments

I find it more satisfying and clearer to say it's when day/night are the same length everywhere on earth/for everyone on earth.

For me that prompts thinking about the earth with its tilt in relation to the sun and the two points in its orbit where that could be true.

Within that framework you could bring up pendantry how "definitions" of day/night mean this doesn't imply day and night are each 12 hours. Actually I find such pedantry more interesting in this framework; it's just clearer that we're saying globally there is always more day than night under this particular definition.

It also IMO prompts the even more interesting pedantry discussed in this thread about it crossing this point at some random time during the day, we don't hover in that point of the orbit for 24 hours.

Haha yeah his definition is from a perspective of need of structural consistency the rest of us wants the definition of day to be useful and represent when it’s light

It's my go to fun fact whenever the situation is suitable for it. People often have never realized this.

(My other fun fact is that South America in its entirety lies East of Florida, and not below North America, most people memorize South America as being completely South of North America.)

Florida is also further west than most people realize, as Miami is longitudinally close to Cleveland, OH rather than New York City.

And, while on the subject - Norway is squarely in Western Europe, no?

Yet the town of Kirkenes in arctic Norway (somewhere around 70 degrees latitude) lies to the east of Istanbul.

What were the odds, got this tweet today https://twitter.com/Locati0ns/status/1705703590181187640?t=2...

> The entirety of continental South America lies east of Michigan" means that, in terms of geographical positioning on the Earth's surface, all of the landmass of South America is located to the east of the state of Michigan in the United States. This is a factual statement based on the Earth's geography. It highlights the vast eastward extent of South America compared to the location of Michigan in North America.

My geographic fun fact is my tiny Canadian province of Prince Edward Island (PEI) is the only Canadian province that doesn't border another country (but there is one territory that also doesn't border any country).

From British Columbia (BC) on west coast to New Brunswick (NB) and Nova Scotia (NS) on the East coast they all have a southern border with the USA. NS has a maritime border with the USA.

Newfoundland and Labrador border two countries although technically each is not what you think. To its south the island of Newfoundland borders France via the small cluster of islands which are a territory of France called St. Pierre and Miquelon. The Labrador border is an eastern maritime border with Denmark/Greenland.

Not a province but the territory of Nunavut also border Greenland/Denmark. It even has a land border via Hans Island a now shared island which Canada and Denmark used to both "fight" over.

The Yukon in the north-west borders the USA state of Alaska.

That leaves Northwest Territories (NWT) it has no border with any nation just its territory neighbours and the prairie provinces to the south.

> South America in its entirety lies East of Florida

Miami, FL: 80.2W Manta, Ecuador: 80.7W

It's close, but not quite.

I was curious, so I calculated what this distance would be in kilometers. At 2 degrees lattitude (lattitude of Ecuador), a difference of 0.5 degrees longitude translates to 55.6 kilometers.

Similarly ... the entire state of Maine, even its very north-most point, is to the south of Seattle's even southern-most point. A lot of US maps are tilted a few degrees, especially schoolroom maps, making the East coast and Maine seem much further north than it is.

It’s amazing how far north Seattle is, and at the same time, how temperate it is. Seemingly less than 8 hours of daylight in the winter and 16 hours of daylight in the summer are amazing.

After growing up in Wisconsin, Seattle seasons were so refreshing.

It’s the marine climate and most weather fronts move in from the ocean. When it does get relatively cold, it’s because of a front moving down from Canada.

The ocean helps. Vancouver, BC is even a bit further north and it’s the same.

One year I was there the flowers were blooming in late January. The middle of a Canadian winter.

Vancouver is Canada for beginners, weather-wise.

My more local fun fact: I live on the North Shore of Montreal, and I have friends that live on the South Shore, but they actually live more north than me.

My favorite geographic fun fact is that Maine is the closest state to Africa.

Assuming a perfect sphere which earth is not ?

Equinox is when the Earth's axis of rotation is perpendicular to the path from Earth to Sun.

That is, when the poles are neither tilting "towards" the Sun (which happens in summer) nor "away" from the sun (which happens in winter).

> I find it more satisfying and clearer to say it's when day/night are the same length everywhere on earth/for everyone on earth.

By the definition given in the article, that isn't true either. They point out that the equinox is in fact the point when day and night are of equal length as long as you define them in a consistent way. But if you instead define day as "whenever the sun is visible" [or would be visible in the absence of clouds], day and night aren't equal on the equinox.

But, if that's your definition of "day", it can never be the case that day and night are the same length everywhere on earth. The presence of mountains will affect when it is or isn't "day".

It's the point where day and night would be the same length everywhere IF the earth stopped moving around the sun and just kept spinning.

Each point along a line of latitude is going to experience slightly different lengths of the day or night when the earth crosses the equinox point.

> I find it more satisfying and clearer to say it's when day/night are the same length everywhere on earth/for everyone on earth.

...except at the poles! (There the sun will appear to be moving along the horizon, and there is no day/night in the usual sense.)

And when you take the pedantry into account (top vs centre of sun, refraction) the length of the day/night is no longer the same around the globe, since the effect in terms of time is different at different latitudes.

1. It's nowhere near night time the moment the sun sinks halfway past the horizon. It's around early evening. Only on a planet with no light-scattering atmosphere does night being at the completion of sundown. And even then, since the star is larger than the planet, more than 50% of the planet is illuminated: night is nevertheless shorter.¹

2. Did you consider the poles?

1. Someone claiming to be a mathematician posted some calculations on quora.com claiming that 50.23% of the Earth is illuminated directly due to the relative size of the Sun and the Earth. https://www.quora.com/What-percentage-of-planet-earth-is-ill...

For an article this pedantic they still miss several details. “If we measured the day starting and ending when the sun’s center breached the horizon, we’d be fine; day and night would be equally parsed.”

It then list a few caveats but that’s not sufficient as the sun is going to cross the the equator at some random point in the day say 2:15:45 pm which means it’s not over the equator at sunrise and sunset thus distorting the day. The horizon is generally defined based on local terrain not the more abstract true horizon. Earth isn’t a perfect sphere rotating at a constant rate etc etc.

It’s got some amusing bit’s of trivia, but I don’t think anyone’s definition is going to actually change.

Isn’t that addressed in the first sentence?

> On Saturday, September 23, at 6:50 A.M. UTC (2:50 A.M. EDT or 11:50 P.M. Friday PDT), the sun will be directly over Earth’s equator, which is how astronomers define the equinox.

They mention it, but not the implications of it occurring at a specific time rather than a full 24 hour period.

Do they need to mention the implications? I’d never considered it being a single moment (I’d always thought of it was being a 24 hour period), but I distinctly remember, after reading that sentence, saying to myself, “Oh, it isn’t a 24 hour period but a moment in time.”

Need to no, but even Wikipedia’s mentions them:

“The word is derived from the Latin aequinoctium, from aequus (equal) and nox (genitive noctis, plural noctium) (of the equal nights). On the day of an equinox, daytime and nighttime are of approximately equal duration all over the planet. They are not exactly equal, however, because of the angular size of the Sun, atmospheric refraction, and the rapidly changing duration of the length of day that occurs at most latitudes around the equinoxes.”

Isn’t that what paragraphs 5-9 are about?

It mentions this bit: “ They are not exactly equal, however, because of the angular size of the Sun, atmospheric refraction,”

But not this bit: “and the rapidly changing duration of the length of day that occurs at most latitudes around the equinoxes.”

An article about something should IMO be more comprehensive than its freaking Wikipedia page.

Isn’t that part also discussed, starting in paragraph 18?

I hadn't thought of this part. I don't think that's the only limitation, though.

Even with a spherical planet, no atmosphere, and 0-diameter sun, there might at most be a circle on the planet that experiences equal day and night.

The planet, after all, is still rotating around the sun as it is rotating around its axis. It doesn't stop rotating around the sun at precisely the moment that the Earth's axis is orthogonal to the Earth-sun gravitational vector. Hence the day and night lengths are continuously changing, not changing once per day.

So I imagine that there is at most a path/line/circle on the Earth's surface that has equal night and day by this measure, and at a minimum, a point, but I'm not mathematically inclined enough to calculate this.

Edit: Same line of reasoning: https://news.ycombinator.com/item?id=37629367

That circle (or "longitudinal line") would only share the same length of day and night if it were measured from the center of the Sun crossing the horizon.

This is because of another point glossed over the the article. The length of time it takes the sun to traverse the horizon varies and gets longer the further you are from the equator during the equinox (or the further from perpendicular the suns apparent path is with the horizon). This means that on the equinox, days nearer the poles are longer than days at the equator because we measure days as when any part of the sun are over the horizon.

This becomes clear when you think about how at the equinox the north and south pole are in the middle of a week or two long sunrise/sunset.

The author is Phil Plait, Bad Astronomer. He is writing for an audience that is not as technically well-informed as most of HN, so it’s no surprise to see his writing called out as “pedantic” by… a bunch of pedants! He’s helping people understand the world a bit better and that’s a great thing to do.

I recognized the style before I noticed his name. Phil was writing online as The Bad Astronomer at least as far back as 1999. I started reading his blog after it appeared on my WordPress “Blogs of the Day” ranking site in 2005.

http://www.badastronomy.com/info/whois.html https://web.archive.org/web/20230922000000*/http://www.badas... https://web.archive.org/web/20051212173430/http://blogsofthe...

Okay so it is what I think it is. Equal day and night. Just with a few equally fascinating pedantic details that make it not quite exactly true.

I could go for more articles like this. In the little details that usually don’t matter are motes of fascination.

I was relieved it was what I thought it was, too. Learning equinox means 'equal night', and solstice means 'sun still' will remain stuck in my head for the rest of my life, though.

It seems I forgot enough Latin to not be misled.

The article is missing the most disappointing fact of all, though.

If there was a whole 24-hour period during which the sun was on the celestial equator, then 12 hours would be day and 12 night (going by the time the centre of the sun's disc rose and set).

But the sun is always moving along the ecliptic, so the day we call the equinox is just the 24hr period within which the sun crosses the equator, which is an instantaneous event.

nit: The sun is not moving along the ecliptic or crossing the equator. It is the Earth that is actually doing all the moving.

One can choose to pin a frame of reference to the Earth, it just won't be an inertial one (and quite noticeably not.)

Perhaps you would accept phrasing the sentence in question as saying the Earth-Sun axis is doing these two things?

The Sun and Earth orbit around a central point between the two bodies, which happens to be very close to the center of the sun. But that doesn't change the fact that they are both orbiting around each other.

This is somehow more wrong than what you answered, which was already very wrong. (Wrong in an aesthetic sense, since both are ill-defined in physics, and only carry aesthetic value in Engilsh.)

The point is neither central, nor solely determined by the earth and sun. The moon and Jupiter influence the orbits of the earth and sun as well (of course as do all the other bodies to a much smaller degree)

Depends on the inertial frame, the sun is flying through our galaxy :)

If I search for ‘ecliptic’, I get 2 different definitions: one is the “apparent path of the sun” (where the observer is the reference point), and the other is the plane of earth’s orbit around the sun (where the sun is the reference point). A good portion of the time, the plane of earth’s orbit is referred to as the ‘ecliptic plane’ and not just ‘ecliptic’. Maybe using ‘plane’ as a qualifier is a nice way to distinguish these ideas. Google’s chosen quick-definition, the one where it shows you the result without linking to a website, is the apparent-path definition.

> On Saturday, September 23, at 6:50 A.M. UTC (2:50 A.M. EDT or 11:50 P.M. Friday PDT), the sun will be directly over Earth’s equator, which is how astronomers define the equinox.

Article starts out by specifying the time of the event

As has been noted in another comment, the article makes that statement, but skips over the implications of the constant apparent motion of the sun for the length of day.

It gives the impression that, geometrically speaking, there is a day on (or near) the equinox which has equal darkness and daylight. Of course, this isn't guaranteed at all, and certainly not for every point on the earth's surface.

Speaking of pedantic details, I'm surprised it doesn't mention how the fact that Earth's orbit is an ellipse rather than a circle messes up the dates quite significantly (it's not February's fault).

I expect the same effect on a smaller scale would also mess up "equal day and night".

It’s all Kepler’s fault.

That Copper Knickers guy has a lot to answer for too!

I agree with that, but I do not like the tone of "no you are actually wrong" from the article, in the way that a scientist could be explaining to you that a tomato is a fruit all day long and you'd still eat it with your salad.

This sort of 'conversational journalism' has become endemic, and it's terribly obnoxious. Scientific American used to be a quality publication, but their editorial standards have sunk to the lowest common denominator over the last 2 decades.

Journalism lost over 90% of its funding after classifieds were gone and tech companies gobbled up all the advertising money.

The current low quality is a natural consequence of the economic incentives.

>If Earth had no air, this wouldn’t be an issue (though there would be other uncomfortable consequences).

They don't mention in the article what those 'uncomfortable consequences' are, but if you asked the average person what would happen if all the air around the earth suddenly disappeared, they would say 'we would all suffocate'. This is of course true, but we would likely succumb to the effects of suddenly being exposed to the vacuum of space before we died from the lack of oxygen.

Having a scuba mask and tank on when the air disappeared would not prolong your life at all. Air pressure is something we tend to take for granted if we are not climbing Everest or in an airplane.

If you genie-wished the atmosphere away in one instant, I wonder what type of explosion the oceans would do as the new at atmosphere emerges, from the gases and liquids that didn't count as atmosphere before suddenly find a new equilibrium.

This is exactly the kind of question Randall Munroe answers in his What If <https://what-if.xkcd.com/>. Unfortunately, this specific one seems not to be answered there?

Well they would boil instantly, basically. Presumably the result would be a lot like the old atmosphere, just... mostly water?

Even if we had no oceans (or other bodies of water, or ice) we'd instantly get a new atmosphere from the resulting vacuum-induced outgassing.

To be a bit pedantic, the dangerous effect of the vacuum of space is the lack of oxygen and rapid suffocation, apparently within seconds: https://en.wikipedia.org/wiki/Effect_of_spaceflight_on_the_h.... The other effects (like ebullism and freezing/overheating/sunburn come "much" later in minutes).

There's a few more Wikipedia pages on the theme, and it seems the weirder effects of decompression are more fictional exaggerations than fact. Lack of oxygen incapacitates in something like 10-15 seconds.

> https://en.wikipedia.org/wiki/Effect_of_spaceflight_on_the_h...

> https://en.wikipedia.org/wiki/Soyuz_11

> https://en.wikipedia.org/wiki/Uncontrolled_decompression#Exp...

> Air pressure is something we tend to take for granted if we are not climbing Everest or in an airplane.

Or dive towards the wreck of the Titanic.

Why would the author even make a point about the different seasons between hemispheres and then immediately get it wrong? He tells the reader to "reverse the seasons and add six months to the dates," but doing both of these operations cancels them out. "The summer solstice in June" becomes "the winter solstice in December," which is still the northern perspective.

Besides, not all dates must be corrected. This one for example is true in whichever hemisphere you live:

> On June 21 or so every year, Earth’s North Pole reaches its greatest tilt toward the sun.

We can fix the linguistic ambiguity by either having the Earth orbiting at much larger radius or spinning much faster around its axis. This makes the length of day/night less relevant.

Ofcourse the strictly correct result is only obtained in the asymptotic limit of infinite orbital radius or infinite rotation speed.

The first option seems problematic. While the rotational period becomes infinitesimal vs orbital, in the limit there is no orbital motion at all, not to mention that cosmological considerations come into play.

But the second option brings other difficulties, like people being ejected into space from the infinite centripetal force.

Its a tough one.

> But the second option brings other difficulties, like

The speed of light limit.

Wouldn’t the simpler (and safe) solution be to reduce Earth's obliquity to zero?

The article shared some interesting facts, but equinox is still basically what I originally thought.

So IMHO the title is a bit clickbaity.

Yeah, a more accurate but boring title would be, "the equinox isn't what I think you think it is".

"Isn't what a few of the couple dozen people I spoke to thought it was"

> On June 21 or so every year, Earth’s North Pole reaches its greatest tilt toward the sun. When that happens, the sun’s path across the sky takes it the highest above the horizon all year, making for the longest days of the year and giving the sun maximal time to heat the ground. We associate this time of year with warmth and sunshine.

The duration of daylight can't be the complete explanation for why summers are warmer because otherwise the North Pole and Scandinavia, having super long days, should be hot during summer. He forgot to mention that the angle of the suns rays, depending on where you are, is also a factor for the warmth, right?

Now that I'm thinking about it, which is really the bigger factor in having a warmer climate: your angle to the sun (your latitude) or the duration of the day? (I realize that they are connected, but I'm trying to imagine separating them.)

I read somewhere that the hottest mix of the two is around 30 degrees.

> around 30 degrees

The Horse latitudes. [0]

Having spent a significant amount of time ~at this latitude, I'm inclined to believe it.

[0]: https://en.wikipedia.org/wiki/Horse_latitudes

In the summer you have both effects - you have more direct sunlight (more vertical angle/less atmosphere attenuation) as well as the longer days. The higher apex of the sun in summer is a sign of the former as well as the latter.

> which is really the bigger factor in having a warmer climate: your angle to the sun (your latitude) or the duration of the day?

This graph from wikipedia might be a good start. It is showing the trajectory of the sun on different days of the year at 56 degrees north latitude (e.g. Edinburgh, Scotland).

https://commons.wikimedia.org/wiki/File:Solar_declination.sv...

The horizontal axis is the azimuth of the sun, but it works as time as well. So that tells you the duration of the day.

The vertical axis is the elevation angle of the sun. To turn this into a solar power factor, you would actually have to take the sine of the angle. But I don't think that would change the overall shape of the graph much.

So you can think of the area under the curve as an approximate measure of energy delivered by the sun.

You are correct. Insolation:

https://www.sciencedirect.com/topics/earth-and-planetary-sci...

> otherwise the North Pole and Scandinavia, having super long days, should be hot during summer

They are hot during the summer, relative to winter. You’re assuming a particular definition of warmth/heat, but these places do have a summer that is much warmer, the average annual temperature swing of ~25C is similar to many cities in the US and Europe. (Which doesn’t answer your question directly, but does say a little something about the relative importance of angle vs length of daylight.)

*edit, btw check out plots of temperature during the day vs during the year. Angle affects temperature on an hourly basis. [1]

Here’s another article all about sun power vs angle, and mentions atmospheric absorption. Note that while lower surface angles receive less surface energy, it doesn’t change the angle of atmospheric energy absorption, and low angles increase the distance light travels through atmosphere. The atmosphere gets about a third of the sun’s energy absorbed by the earth, the surface gets the other two thirds. [2]

[1] https://weatherspark.com/y/9847/Average-Weather-in-Kansas-Ci...

[2] https://en.wikipedia.org/wiki/Solar_irradiance

The article says that southern hemisphere skies are better than northern? I’m interested in what way?

Better stars? Bluer day sky? Less clouds? Better clouds?

Anyone know what he was referring to?

The southern sky has a better view of the Milky Way and there are more bright stars and constellations visible. Also the fact that the southern hemisphere is less populated means that the skies are darker (on average).

See below for a more detailed explanation: https://www.skyatnightmagazine.com/advice/southern-hemispher...

https://www.lightpollutionmap.info

Oh nice, thanks

I grew up in the northern hemisphere and now live in the southern hemisphere.

The night sky appears brighter and more starry in the southern hemisphere (after correcting for moonlight and light pollution, of course) due to the Milky Way, but it is harder to see constellations due to the low contrast.

The most noticeable individual stars sit quite low in the sky. Here, the "background" behind the nearest stars is black, making them easy to see and seemingly brighter.

I think it is no coincidence that the vast majority of named constellations occur in the northern hemisphere. In the southern hemisphere, people traditionally navigated by the Milky Way rather than by individual patterns of stars.

Both hemispheres have a very beautiful and distinctive night sky. It is impossible to say which is "better" because they are so different from each other.

The thing that always gets me us that when we reach the shortest day (southern UK here) the latest sunrise and earliest sunset are actually several days apart. And this leads to the rabbit hole of how mean time varies from solar time, when considering noon

“a complete cycle of one day and night is called a nychthemeron […] I love odd words like that”

I love them too! Lithuanian language has the similar concept of “para”, sometimes very handy and i don’t think i’ve seen it in other european languages.

By para, you mean a complete day & night cycle?

If so, we have something similar in French, with « jour » being a 24 hour span, whereas a « journée » only designating the daylight part of it.

in English there is 'day' and 'daytime'

In Danish we have ‘døgn’ and ‘dag’.

“I stayed one day at Disneyworld” doesn’t mean i stayed 24 hours there

"I drunk a glass of water" doesn't mean that I filled a glass exactly full of water and I drunk all of it leaving no drops on the inside of the glass.

That’s my point, “para” is a rare concept

I see now. So you are claiming that you use the night exclusive version of the word, when you talk about spending a day somewhere. (Bc the night exclusive and night inclusive days certainly exist in most every languages, even in English you can say that you spend the daytime at disneyworld, when the distinction matters.)

That sounds rare indeed.

That’s where different languages have different concepts which one may or may not find useful from the outside. In other languages it doesn’t really matter whether the person was night inclusive or night exclusive, in Lithuanian you have to pick one word or the other.

Using the verb ‘stayed’ does tend to imply you slept there and were there ~24 hours. If you had ‘stayed two days’ or more, that really would imply N 24 hour cycles approximately, and at least N-1 cycles certainly; any other number besides 1 in your sentence would mess up your example. You already know the word “day” has multiple meanings, and in some contexts (such as time codes and dates and time measurements, using “day” always means 24 hours). An example using one of the meanings is irrelevant, it does not contradict the other meanings.

If someone tells you they are only staying one “jour” at a place you can’t infer it includes the night as well. With “para” it’s clear.

Do you have an example in mind?

It’s 3pm and you buy a “one day” metro ticket, will it still be valid at 4am? The nychthemeron concept makes it clear.

Not really, because the nychthemeron does not carry the concept of an anchoring hour, only a consecutive 24h span.

I'm a little confused by this - how is this different from...a day? Like in the "24-hour day" definition of a "day."

> southern skies are far better than northern ones

yes they are, by far

Crying out for an animation. A recent (5y) BBC series on the seasons had a phenomenally good speeded up view of the equinoctial shifts and visually explained why the North and south don't get an equal deal in this (the orbit is an ellipse so their respective summer and winters take place at different distances from the sun)

Can you remember what it was called?

Probably this one, and 2012 so I was out by 6 years!

https://www.imdb.com/title/tt2286877/

My wife has the equinox watch face on her Apple Watch, and she monitors the progress daily. We all need our own hobbies and interests. I forwarded this article to her.

https://archive.ph/K8VN4

Consider a spherical, no-atmosphere planet with a pointlike sun

TLDR: it is, but there's some interesting details.

If you want a clock that shows this visually: https://sunclock.net/

>day and night would be equally parse

Isn't it just day = light and night = no light, ignoring moon, stars etc.?

tldr the sun is not a point

Is equinox a subsidiary of equifax? /s

Who the heck is SA writing for? Children know this.

Therefore they must be writing for children. Good point!

Relevant xkcd: https://xkcd.com/1053/

It's not.

Reply on news.ycombinator.com