Steve Allen sla at ucolick.org wrote on Tue Jan 7 20:15:38 UTC 2020

In the middle of last year the Earth's crust accelerated such that we were rotating faster than 200 years ago.  The value of Delta T decreased during a few weeks.

The Length of Day (LoD), as Dr. Allen knows better than I, is effected by many factors. Having picked a length for the atomic SI second where 86,400 of them add up to a day shorter than the mean Solar day by milliseconds, we have been playing "catch up" since UTC was defined. The tides slowing down the crustal rotation, along with torque applied directly on the Earth oblate spheroid by Sol and other celestial bodies, along with other factors, change the rate at which the LoD changes. I have been fascinated by the fact that Bulletin A projections of UT1-UTC have put the point where it reaches -0.4 (which is typically when IERS inserts a leap-second) further and further into to future.  Recently, every week, the MJD where it gets to -0.3 (currently at -0.29480 on 2021-01-01) gets a week farther away. It has been proposed that the latest oddities are caused by angular momentum being transferred from the Earth's rotating solid core to the liquid surround it, the semi-solid mantle and, finally, to the crust.  Not something instantaneous by any means. My favorite thought on the matter involves wondering about conservation of angular momentum as megatons of ice, melting at altitude, unaffected by friction, flows down hill to (a rising) sea-level.  Much of that ice is miles in the air.... lots of potential energy realized into actual rotation by a change of position.  (Think of it as an spinning ice skater pulling in her/his arms.)   Change of spin would be effected by closeness of the ice to the spin axis/poles as well as height of sea level at the equator. One more thing to worry about....