The most recent leap second was introduced at the end of December 1998. At the time leap seconds were introduced, the excess length of day (LOD) was around 3 ms per day and it gradually dropped to around 2 ms/d by the mid-1990s. See the plot at <http://hpiers.obspm.fr/eoppc/eop/eopc04/eopc04-lod.gif>. For an average LOD of ~2 ms/d, it takes ~500 d (roughly every 1.5 years) to accumulate 1 s and hence produce a need for a leap second. That was the situation until the late 1990s. Since then, LOD has continued to drop to an average of <1 ms (current average is ~0.5 ms). So lately the interval between leap seconds has grown greatly. UT1-UTC is predicted to be -0.45 s one year from now (see ftp://maia.usno.navy.mil/ser7/finals.data) and therefore will still not warrant a new leap second. There is at this moment no likelihood of another leap second within the horizon that can be reasonably well predicted (which is not really very long).
While the secular tidal breaking of the Earth's rotation almost certainly dominates over geologic timescales, the dominant LOD forcing over human timescales is the so-called "decadal LOD variations". You can clearly see those effects in the eopc04-lod.gif cited above (at periods longer than the seasonal LOD variations). It is really not fair to say that the cause of the decadal variations is truly unknown, although one cannot state that the mechanisms have been conclusively demonstrated. Direct and indirect evidence point to the Earth's core as the driver of the longer term LOD changes -- see the informative IERS website at <http://www.astro.oma.be/SBC/main.html>. Because of the very poor observational constraints of the Earth's core, it has been difficult to develop high-fidelity core models. But much progress has been made in recent years and some results for correlations between LOD and core models back to the early 1800s are shown at the website above. Geophysically, it is rather clear that the core effects dominate over secular braking for UT1 variations on timescales relevant for individual people. Only if one concerns themselves with events in the very distant future does the issue of secular braking become important. It is for this reason that I find this entire discussion, and the proposals to change current leap second practice, to be so utterly senseless. If somebody (Bill K?) feels that a continuous timescale is required for some particular application, then by all means use one of those already available (TAI, GPS time, future Galileo time, ...). There is absolutely no justification for making UTC into yet another such flavor while denying a timescale closely linked to UT1. And attempting to use the predicted evolution of leap seconds as justification for this senseless act is to ignore the actual observational evidence -- an act approaching scientific fraud. --Jim Ray Ed Davies wrote: > > Meanwhile, this article also reminded me of a question that > has been lurking in my mind for a while. We haven't had a leap > second for a while even though: > > 1. The Earth has already slowed down enough, on average, to > need a leap second every year or two. > > 2. The usual tidal effects, etc, have presumably still been > busy slowing it down further. > > So, something has been speeding the Earth up for a while. As > this article says - it's not really known what this effect is. > My question: does anybody have a feeling as to whether this > effect is likely to "unwind" at some point resulting in a > faster than normal string of leap seconds? Could we get to > leap seconds more often than once every six months any time > soon? > > Ed. >
