The equinox marches on

Spring has officially arrived in the Northern Hemisphere.  Not that you'd notice here in the Western US, where it's unseasonably cold.  Of course, in the East it's unseasonably hot, and in the middle the combination is causing tornadoes.

In fact, spring arrived last night, at 10:14:09 pm my time, Pacific Daylight, on the 19th day of March by the Gregorian calendar (56006.21816).  That's the 20th of March for most of the world, but this is still the earliest equinox in over a century.  Part of the reason for this is that there was a leap day last month, which pushes the date back a day.  The next one, in 2013, will come nearly six hours later, and so on until the next leap year in 2016, when it will again jump back a day.  In fact, it will jump back slightly more than a day.  It continue to slip earlier and earlier every leap year, until 2100, when we will skip a leap year.  The reason we have leap years is to keep the equinox from arriving later and later every year.  The reason we skip leap years is to keep the equinox from arriving too early.  The earliest vernal equinox this century will be in 2096 (86686.5848) after which it will advance nearly six hours each year until 2103 (89243.265).  In 2104 the leap year cycle will start again, so the equinox date will again go back a day, or about 18 hours before 2103 (89608.509). 

For the rest of this century, spring will arrive more and more often on the 19the day of March in Pacific Daylight Time, instead of the 20th.  Then it will arrive on the 20th every year from 2099 to 2135.  Of course, it will always arrive earlier in the Eastern Hemisphere, often on the 21st day of March, as it's supposed to.  The Romans, it's said, originally fixed the equinox on the 25th day of March, as Julius Caesar noted on his calendar, but by the fourth century it had slipped to the 21st, and the latter date is the one that Pope Gregory noted on his calendar, by which time it had slipped by between 9 and 10 days, so he dropped 10 days, although it probably would have been more accurate to have dropped only 9, since it usually falls on the 20th in Europe.  At least it's closer in Asia and Australia.  (I know that doesn't quite add up, since the date should have slipped only 3 days in the first four centuries rather than 4, but I don't know the reason for the discrepancy.)

There is a good graph of the March equinox's wanderings on Wolfram.

MJD 56006.974