On 1 mai, 19:53, "Mike Dworetsky" <platinum...@[EMAIL PROTECTED]
>
wrote:
> "Crown-Horned Snorkack" <chornedsnork...@[EMAIL PROTECTED]
> wrote in message
>
> news:5b40436f-cb79-4966-807b-7b83b9eaacd8@[EMAIL PROTECTED]
>
>
>
> >Well, when was it?
>
> >The actual length of sideric day, and of tropical day, increases
> >slowly, as tidal friction irreversibly slows down the rotation of
> >Earth.
> >The actual length of sideric year can change - the energy of Earth=B4s
> >orbit may change as energy is exchanged between orbital movement of
> >Earth and other planets. But those changes are said to be minor. The
> >changes in tropical year are likewise minor.
>
> >The result is that the number of tropical days in a tropical year
> >decreases.
>
> >Sometime in palaeozoic, tidal rhytmites allegedly show that there were
> >400 days in a year.
>
> >But the rate of tidal slowing is not constant. It changes with changes
> >in the configuration of shelf seas and ocean basins. There would be
> >major changes during ice ages, for example.
>
> >The true duration of tropical year is around 365,2423 to 365,2424
> >tropical days. Gregorian calendar requires 365,2425 days.
>
> >The ***ulative error of Gregorian calendar through recent is thus less
> >than two days.
>
> >But how valid was Gregorian calendar in ice age?
>
> The question is, in the everyday sense, meaningless, because the
Gregorian=
> calendar is not used for years prior to 1582. Not for any sensible
purpos=
e,
> because the main event it was designed to regulate thenceforward--the
date=
> of Easter--had already taken place in each of those previous years.
(Late=
r
> in England--Newton was born on Christmas Day in the Julian calendar, so
we=
> still call that his birthday.) Looking backwards for dates is what you
ge=
t
> using a "proleptic" calendar and usually the Julian calendar is used for
> this.
>
This causes problems. Julian calendar is considerably farther from
true length of tropical year than the Gregorian.
> Aside from the removal of 10 days in 1582, the difference between
Julian
> and Gregorian is in the number of days in an average year: 365.25 or
> 365.2425. So if you tried to decide the date in some very distant past
yea=
r
> (e.g., 10,000 BC) you would have to account for that difference, or
about
> .75 d per century or 7.5 days per millennium.
>
And 75 days over 10 millennia. Which means that a proleptic Julian
calendar is plainly out of alignment with seasons in early recent.
In which calendar month did Laach lake eruption take place?
> Of course we also have a ***ulative error in the sense that the actual
> period length of a tropical year is 365.2422 days (to 4 decimals) so the
> Gregorian calendar itself will gradually need some reforms; a leap year
wi=
ll
> have to be omitted, perhaps the year 4000. I imagine the Catholic
Church
> will give thought to this in about 1700 years or so. (Could be an
> interesting SciFi story theme?). The error ac***ulated is about 1 day
eve=
ry
> 3300 years rather than 4000 years, so in the very far future a further
> reform might be needed.
>
> The effect of changes in the rotation of the Earth would be smaller but
> would ac***ulate more rapidly as you went back in time. The figure for
c.=
> 2000 years ago is of the order of three hours error. See:
>
> http://www.blackwell-synergy.com/doi/pdf/10.1046/j.1468-4004.2003.442...
>
> It is difficult to extrapolate rotational rates back accurately over
many
> millennia because more than tidal effects are involved.
>
Indeed. However, they could be measured directly.
Misalignment over a few weeks should show up in geologic record.
> Astronomers use something called the Julian Day number which is a
running
> count. The Julian period starts on 1 January 4713 BC (Julian calendar)
an=
d
> lasts for 7980 years. It starts on the day when the Roman Indiction,
Gold=
en
> Number, and Solar number all had a value of 1. This allows them to
avoid
> having to use dates in one calendar or another for calculated events in
th=
e
> distant past (like eclipses). For accurate timings they of course have
to=
> use a dynamical time scale independent of variations of Earth rotation.
Indeed. Julian day is bound to rotation of Earth which, in long term,
is more variable than tropical year (bound to orbital movement of
Earth). Neither of which is constant flow of time as bound to inertial
laws.
> When they define somenthing to be measured in years they usually use a
uni=
t
> of Julian years of 365.25 days, e.g., if discussing the period of a
visual=
> binary star, or the orbit of the Sun around the galaxy.
>
Of course, it may be a question what the most im****tant year is:
tropical, sideric or anomalistic...
|
