On Thu, 09 Jul 2015 22:07:57 -0400, Phil Hobbs
wrote:

On 7/9/2015 9:14 PM, Jeff Liebermann wrote:
On Fri, 10 Jul 2015 01:37:30 +1000, Clifford Heath
wrote:

The earth's slowing is also somewhat chaotic, inasmuch as equatorial
weather affects the sea-level heights, which introduces noise into the
earth's angular moment of inertia, and hence its rate of rotation. That
has nothing however to do with how we know we're measuring time accurately.

It might help to mention that we have two types of time accuracy. One
is sidereal time, where 12AM on Jan 1 is astronomically correct and is
used to aim telescopes on earth.

Nope. Sidereal time is different from civil (solar) time. The Earth
rotates 365 and change times per year with respect to the Sun, but 366
and (the same) change with respect to the fixed stars. So the two get
out of phase pretty fast.

Sorry, my mistake.

The problem is that the two systems don't quite coincide. The current
difference between UTC and International Atomic Time (UTC-TAI) is now
36 sec and growing. The recent leap second just made things worse.

Well, worse if you don't think that the Gregorian reform was an advance.
Pretty soon the vernal equinox would have been in February.

The Julian calendar was working just fine for 1500 years as everyone
know how to tweak the date so that it matches the solar calendar. That
was fine for farmers and bankers, but didn't do much for the church,
which had the bad taste to celebrate their holidays by the calendar
month and date. Most everyone else used the signs of the zodiac to
set the beginning of the month. That worked well for the GUM (great
unwashed masses) except that the church equated the zodiac with pagan
religions, alchemy and witchcraft, so that wasn't going to work. A
pope previous to Gregory XIII tried to switch the holidays to the
zodiac months (can't find the name) but gave up before going public.
I suspect that Gregory XIII must have had second throughts when he
allowed the astronomers to fix the calendar. At least they named it
after him, so I guess he was happy.

Incidentally, if you want a really screwed up calendar, try the Hebrew
calendar, which adds an extra month every 2 or 3 years, every 7 of 11
years.
https://en.wikipedia.org/wiki/Hebrew_calendar
There's nothing like a duplicated month (adar) to create confusion.

Personally I think that civil time is more important than atomic time.
Folks who need to know the difference, do.

If we knuckle under to atomic time in civil life, our version of the
Julian problem is that midnight by the clock will soon start occurring
at sundown. The leap second inconvenience principally affects software
developers (and those who trust them).

I don't see a problem. If every political time standards organization
can have it's own time standard, I see no reason why they can't expand
theirs to a calendar standard. You just pick the calendar that is
appropriate for whatever you're doing. It's not much different than
the US before the railroads, where every town had it's own time and
DST standard.

Yeah, there were some hiccups in 2012.
http://www.wired.com/2012/07/leap-second-bug-wreaks-havoc-with-java-linux/
http://www.wired.com/2012/07/leap-second-glitch-explained/
I missed the fire drill, but still managed to get wakened by a
customer wanting to know why their backup failed. Stupid me had set
cron to start the backup exactly at midnight. That worked, but one
second later, it started a 2nd backup during the leap second. Why, I
don't know, but that's what the log files showed. I killed both
processes and started over.

The fun starts when tracking spacecraft in otter space. Not only does
one have to deal with relativistic effects, but one also has to use a
time system that is independent of how the earth spins, wobbles, and
thrashes around. It would be a major disaster if a leap second were
thrown into the timing if you're tracking a spacecraft such as Voyager
1 moving at 17 km/sec (38,000 mph).

You'd be off by 17 km. Is Voyager 1's position known to that accuracy?
Didn't think so.

I was thinking of it in terms of the change in angular error for the
rotation of the earth.
degrees = 17km/40,075km * 360 degrees = 0.15 degrees
Let's see if that works. Voyager 1 and 2 uses the DSN (deep space
network) with 34 or 70 meter dishes at about 8 GHz. That's about 67dB
gain and a -3db beamwidth of about 0.07 degrees for the 34 meter dish,
and 73 db gain and 0.04 degree beamwidth for the 70 meter dish. Since
the DSN tracks the rotation of the earth, a change of 0.15 degrees
would move the main lobe sufficiently to miss the spacecraft.
http://www.satsig.net/pointing/antenna-beamwidth-calculator.htm
http://www.uhf-satcom.com/misc/datasheet/dh2va.pdf

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558