Re: [find_orb] Delta-T data used by Find_Orb

Bill Gray Jul 4, 2018

Hi Andrew,

I think Francisco and Dave have answered this nicely. Find_Orb does
use "for-real" Delta-T = TD - UT1 = difference between atomic time and the
earth's irregular rotation. But it does so to figure out how much the earth
has rotated and where an observer really is relative to the geocenter. Actual
times, as Dave describes, are either on an atomic time system (TDB or
TDT... pretty much identical for our purposes, only about 1.6 milliseconds
apart at most), or they're on the horror that is UTC (*). And UTC will have
a fixed offset from TDB or TDT, changing only when the Time Lords in Paris
insert (or, theoretically, delete) a second.

As Dave says, unless you're doing something unusual, you may be able to
say that TD-UTC=69.184 seconds, exactly, for dates since 2017 January 1.
On 2018 July 1 or maybe 2019 January 1, TD-UTC will suddenly be 70.184
seconds.

-- Bill

(*) (Getting on my soapbox here; picture the following delivered as a
red-faced harangue, with spittle being sprayed over the first few rows.)

The claim is made that leap seconds are needed to provide a civil time
system that will stay in synch with the sun, to within +/- 0.9 seconds.
That claim is made, but then actual local civil time (the sort we use when
telling someone we'll meet them for a beer at 5:30 PM) is divided into
time zones, introducing a casually accepted half-hour error; then further
mangled with Daylight "Saving" Time by another hour. You never really
know when a local government will change zones or DST rules.

I see no need for _two_ levels of unpredictable, arbitrary time
mangling. Drop leap seconds, and leave UTC fixed relative to atomic
time. As the earth's rotation requires it, governments may decide,
a millennium or two from now, that clocks need not "spring ahead"
in their jurisdiction, in order to stay at least somewhat in synch
with the sun, to the same minimal degree that they currently do.

(OK, I'm getting off my soapbox. Thank you; I feel better now.)

On 2018-07-03 16:04, Andrew Lowe andrewlowe@... [find_orb] wrote:
>
>
> Hi Bill,
>
> In your Find_Orb documentation, you mention that you use a table with two-year increments to compute the difference between Terrestrial Time and Universal Time (Delta-T).  Could you please let me know these values for the past and future, to be compatible with my ephemerides that I use for input into Find_Orb.
>
> Thanks.
>
> Andrew
>
>
>