The Plan to Lose Leap Seconds Would Throw Our Clocks Out of Sync With the Earth

Our clocks are more accurate than the Earth's rotation, and we need to decide whether to keep them in step.

|
Jun 6 2014, 11:20am
Image: Flickr/Arjan Richter

How do you tell the time—by studying a super-accurate atomic clock? Looking at the clock on your computer? Or gazing at the sky to calculate the position of the sun? 

By our current methods of counting the passage of time, those three should all match up (so long as you’re being accurate with your measurements). But that’s only because we currently fiddle around with the first two so that they match the great time-teller of the ages: the rotation of the Earth. Current discussions to stop this meddling could instead push us to rely solely on the atomic clock—and throw us off kilter with the Earth’s natural hourglass.

The UK’s National Physical Laboratory (NPL), which has its own share of atomic clocks, is turning to the public for their views on whether we should abolish the “leap second,” the device we currently use to keep solar time and atomic time in sync. There’ll be a summit in London next month to cement the UK’s position ahead of an international meeting next year that will decide the future of the leap second.

Here’s the problem: The rotation of the earth isn’t perfectly regular; it’s very gradually slowing down, as a result of the long-term movement of the tides, tectonic activity, and polar ice build-up. But while the Earth is spinning happily away to its changeable natural rhythm, our atomic clocks—which form the basis of Coordinated Universal Time (UTC)—keep ticking with perfect regularity.

“We now have clocks that are much better at the Earth than keeping time—atomic clocks” Peter Whibberley, who leads work on the operation of UTC at NPL, summarised for me. “We still need to keep the two in step, because otherwise UTC, which is the basis of clock time around the world, would slowly drift apart from the time based on the Earth.”

"Without leap seconds we will eventually lose the link between time and people's everyday experience of day and night."

Since 1972, we’ve been ironing over this discrepancy by inserting the occasional “leap second”, an extra second added to a minute at the end of a month to give the Earth chance to catch up with our  modern timekeeping devices. The last was in 2012: You probably didn’t notice, but the very last minute in June had 61 seconds. You might have noticed that Reddit went a bit glitchy.

And therein lies the main concern with leap seconds: They’re screwing with our tech. Last leap second, the time jump caused a bit of disruption in some computer systems because of glitches in the coding to deal with leap seconds. Part of the problem there is that you only get a few months’ warning if there’s going to be a leap second, so it’s more difficult to plan than, say, a leap year. The worry is that, in the future, a leap second could cause a bigger glitch—after all, having reliable, universal time is pretty important for things like navigation and communication systems.

Countries are split on the issue, but we need to come to a global decision. Whibberley explained to me that, despite different time zones, we all rely on UTC and so it needs to remain universal. “UTC is the single underlying reference time-scale that’s used worldwide for all precise time measurement, and so you might offset your country’s local time by a whole number of hours from UTC, but using UTC as a reference,” he said. Additionally, time signals often cross boundaries. Many people worldwide, for instance, get time from GPS satellite signals, and they’re controlled by the US Naval Observatory.

The US has supported dropping the leap second for quite a while, and had the backing of countries including Japan, Italy, Mexico and France at an international meeting in 2012, according to a BBC report at the time. The UK, on the other hand (along with Canada and Germany if the last meeting is anything to go by), is pretty set against the change.

Because while a second once or twice a year doesn’t sound like much, they’ll add up to throw our measured days out of sync with solar days—which we’ve been using long before fancy atomic clocks as an indicator of time. When opening up the public dialogue on the matter, Minister for Science David Willetts made his view pretty clear. “My view is that without leap seconds we will eventually lose the link between time and people's everyday experience of day and night,” he said.

The NPL does the math: In 800 years, the sun would be at the highest point at 1pm instead of noon. Michael Brooks, who’s a member of the group overseeing the public dialogue, wrote in the New Statesman that, “Without leap seconds, or some other adjustment of time, noon in the year 4000 will occur in total darkness.” 

Which brings us to the pivotal question that needs deciding: “Can we justify dropping the leap second—and maybe redefining 'noon'—just because of computer programming problems?”