Tuesday, March 20, 2012

Equinox: spring usually begins on March 20th

Even though you were probably taught that the spring begins on March 21st Greenwich Winter Time (UTC), the last beginning of a spring on that day was in 2007. At least for 84 following years, the equinox will take place on March 20th.

A dogma of the Christian churches is that spring begins on March 21st, too: so you may see some tension between their teaching and astronomy.

UT date and time of solstices and equinoxes[1]
year Equinox




day time day time day time day time
2007 21 00:07 21 18:06 23 09:51 22 06:08
2008 20 05:48 20 23:59 22 15:44 21 12:04
2009 20 11:44 21 05:45 22 21:18 21 17:47
2010 20 17:32 21 11:28 23 03:09 21 23:38
2011 20 23:21 21 17:16 23 09:04 22 05:30
2012 20 05:14 20 23:09 22 14:49 21 11:11
2013 20 11:02 21 05:04 22 20:44 21 17:11
2014 20 16:57 21 10:51 23 02:29 21 23:03
2015 20 22:45 21 16:38 23 08:20 22 04:48
2016 20 04:30 20 22:34 22 14:21 21 10:44
2017 20 10:28 21 04:24 22 20:02 21 16:28

You may see that every non-leap year, the spring starts about 6 hours (5:53 or so in average) later than the previous year. However, on every leap year, such as 2012, it's 18 hours later and the three previous 6-hour jumps are undone (and spring actually begins about 40 minutes earlier than 4 years earlier). Consequently, the 2012 spring equinox is the earliest equinox since 1896 when it was even earlier because 1900 wasn't a leap year (and because of mathematical reasons you should think about). This record will keep on improving until 2096. The year 2100 won't be a leap year so the date of the equinox will jump by 1 again.

Yes, this blog entry officially appeared at 6:14 am, Prague Winter Time, which is when the spring started.

Someone has created a video on the origin of seasons for Brian Cox: Seasons are not caused by the changing distance from the Sun

A homework exercise for you: Did the folks try to keep the same date for equinox? And if they did, what is the reason that they failed and the date has shifted to another date? Although leap years have periodicity of 400 years, vernal equinox in 1612 occurred 2+ hours later than today. That makes a 12-hour difference in 2000 years. If you've forgotten almost everything, try to remind yourself about some leap year maths.

Jean Michel Jarre, Equinox, 1978

A web technicality: Readers who don't see the motls.blogspot.com but a national one should use the pirate logo below – near the "fast comments, snail comments" label – to participate in the global discussions (hosted by the U.S.). The comments posted via the Echo form that opens up won't be visible to the readers outside your country (if anyone at all).


  1. While in the topic, lets note that from 20 March to 22 Sept we have, hmm, 186 days, 9 hours and 35 minutes, while from 22 Sept to 20 March of 2013 there are 178 days, 20 hours, 13 minutes.

    Obviously even if it were exactly from 21th to 21th, the intervals are diferent, then disproving the idea of a circle with the sun (nor the earth) in the center.

  2. Dear Alejandro, your discovery is known as the elliptical orbit and it was known in some detail to ancient astronomers, including a detailed description by epicycles which are often humiliated by some of the very same people who don't even know that the orbit isn't circular.

    Kepler was able to precisely describe the shape and immediate speed of the planet at every point of the ellipse, aside from noticing it was an ellipse with calculable parameters.

  3. Hmm I am not sure if ancient astronomers even considered to set a solar epycicle to answer to this. Ptolemy just moves the earth out of the center but keeps it fixed and the orbit is still a single circle. But yes, given that we can say that earth and sun move around the common mass center, an epycicle is pretty obvious. Just perhaps not needed within the old precision of the measurements.

    By the way, I think that Kepler was able to measure the "ellipticity" of Mars orbit, but I am not sure if he had the tools to confirm, at five sigma :-), that the orbit of the Earth was not a circle.

  4. Dear Alejandro, you're just wrong. Epicycles knew about all these effects we attribute to ellipses today.

    As a review on epicycles and ellipses states, every ellipse is identical to an epicycle but not every epicycle is an ellipse.

    Epicycles are a systematic method to approximate an ellipse - or any other orbit of the same kind.

    In this sense, Kepler's coining of ellipses was similar to the transition from quantum field theory to string theory; it inherited a more concrete, more constraining concept, but it didn't change the actual predicted and osbervable numbers in any significant way.