What are leap seconds?

Scientists sporadically need to add or deduct seconds from our clocks but if our timekeeping is so accurate, why do we need to do it?
30 June 2015

Interview with 

Peter Whibberley, The National Physical Laboratory


The first atomic clock revolutionised how we tell the time and even today our clocks digital clockare modelled on Louis Essen's creation. But since then, they've dramatically improved in accuracy; now only every 160 million years, will these clocks gain or lose a second. So, if they're so accurate, why do we need to keep adding and deducting leap seconds to our atomic clocks? Peter Whibberley from the National Physical Laboratory gave Chris Smith a run down...

Peter - Well, the clocks that we use today, the primary standards that we still use to measure the second are still based on caesium atoms but they are rather different to the type of clock that Essen built. In Essen's clock, we had an oven forming a beam of caesium atoms ,which shoot through with the microwave. In today's clocks, we use lasers to slow down caesium from a vapour, form a cloud of slowly moving caesium atoms. And then we use the lasers to throw that cloud upwards through the microwave enclosure. The atoms then carry on rising and fall back under gravity through the cavity a second time. That gives a much longer measurement of time of which longer interaction between the atoms under microwaves of around  half a second. This longer interaction gives us a much more precise measurement of the frequency and therefore, a more stable clock.

Chris - Is that opportunity to engage with them for longer - meaning that we make a more accurate and more precise measurement of their state.

Peter - That's right. The longer time you have to make a measurement then the more precisely you can measure an energy or a frequency in this case.

Chris - Why do we need amazingly good clocks like this because one in 160 million years, as in one second being added or taken away every 160 million years, do we really need that level of accuracy and precision?

Peter - They're certainly applications that would benefit hugely from these new types of clock. There's a lot of work going on around the world to develop these optical clocks which will surpass the caesium clocks in their performance.

Chris - If these clocks are so accurate and precise, why do we have all these problem with leap seconds then? Why do we need to add these seconds periodically? This is the 26th leap second now this year?

Peter - That's right, 26th since 1972 when they were first introduced. The reason we need them is that in fact, the Earth is not a very good timekeeper. It speeds up and slows down unpredictably. Over time, we find that our very precise time based on atomic clocks slowly diverges from the time based on the Earth's rotation from the mean solar time. And to keep the two in step, we have to occasionally put a 1 second step into the atomic timescale just to let the Earth catch up with the atomic clocks. This is called the leap second.

Chris - Right. So, there's nothing wrong with our timekeeping apparatus, it's that our planet can't keep time properly.

Peter - The Earth in fact is not a very good timekeeper. It's a million times poorer than caesium fountains that are keeping time.

Chris - What do people think internationally about the concept of the leap second because does it really make a very big difference?

Peter - That's an interesting question and there's been a big debate going on now for around 15 years, as to whether we really do need leap seconds. Should we just base our time purely on atomic clocks and just break the link with the Earth's rotation?

There's no agreement internationally. Some countries favour ending leap second because they do cause problems. Some software in particular has great difficulty handling leap seconds. The simplest solution is simply to end them.

But other countries say it's important to maintain the traditional link between timekeeping and the Earth's rotation and arguing we should keep leap seconds until at least we understand much better the long term consequences of ending them.

Chris - So, where do you sit in the argument?

Peter - Well, I'm part of the UK's delegation at this grand meeting that would discuss the issue in November of this year. So, it's my job to argue the UK's position but from a personal point of view, I don't have any stronger sympathies one way or the other. They're both very good arguments and the problem is, no compromise position. You had to keep leap seconds or you end them. Whatever happens, it's going to be very interesting.

Chris - And the UK is for keeping it?

Peter - The UK government has considered the issue and it's theory is we should maintain this traditional link between our timekeeping and the Earth's rotation.

Chris - If we do decide to go with the counter argument and we ditch the leap second, what might be the consequence?

Peter - In the short term, not a great deal. The divergence between our atomic time and the Earth's rotation will slowly accumulate. It would take maybe 100 years to reach 1 minute difference and maybe a thousand years to build-up to 1 hour difference. So, within our normal lifetime, we wouldn't notice a huge difference. But over the longer term, the rate of divergence increases more and more. And so, we have to do something to correct this difference between this lifetime and Earth time. And there's no real agreement at the moment as to how that should be done. The UK government is certainly arguing that we should understand the long term consequences better before we make such a fundamental change.


Add a comment