Lunar recession/asteroid occn's

Bill J Gray Oct 31 6:05 PM

Paul, the lunar recession is indeed built in to the various
theories of lunar motion. For dates a millennia or more ago, the
main cause of error is the fact that the rate of recession is
known only to middling accuracy.

This is linked to the Delta-T problem. Basically, angular
momentum has to be conserved; in this case, the Earth's spin
decreases, and the angular momentum goes to the moon. (I vaguely
remember a freshman physics problem along the lines of, "If the
moon is receding at three centimeters/year, at what rate is
the earth's spin decreasing?")

Shawn, Guide _can_ display asteroid occultation paths with
wonderful accuracy. But you do have to go to a lot of effort
to do it.

Just using MPCORB will fail on several counts. You really
need an orbit based on highly accurate astrometry in the Hipparcos
reference frame. The FASTT program at the Naval Observatory in
Flagstaff, Arizona produces most of this data; I occasionally
get some from folks at Table Mountain as well. You use this
highly-accurate data, plus occasionally some older, less
accurate observations to increase the baseline, and compute
an orbit with an epoch close to that of the event. (MPCORB
can be as much as 100 days away from "the present", because
its epoch changes at 200-day intervals. You can be ignoring
up to 100 days of perturbations.)

Some people have taken FASTT/Table Mountain astrometry,
run it through Find_Orb, gotten elements, and fed them into
Guide. At that point, the asteroid is OK. I just did it,
and got the following orbit:

Orbital elements:
(431)
Perihelion 2001 Nov 2.801867 TT
Epoch 2002 Nov 3.0 TT = JDT 2452581.5
M 64.98617 (2000.0) P Q
n 0.17794772 Peri. 212.38111 0.86420774 0.50233804
a 3.1304571 Node 117.43880 -0.01567355 -0.02471558
e 0.1783031 Incl. 1.82796 0.54544401 1.80365084
P 5.54 H 9.0 G 0.15 q 2.5722868
From 87 observations 1998 Feb. 24-2002 Oct. 30; RMS error 0.148 arcseconds

Save the above elements to a file, and use "Extras... Add MPC
Comets/Asteroids", and the elements will be imported and the
revised object shown. (Click for 'more info' on it, and you'll
see that the epoch is 3 November.)

Your next problem is usually the star, but in this case, it's
a HIP star with a nice, solid position. You can right-click on
the star, then on the asteroid, then on "Extras... Show
Eclipse", and get a "right" path. Add markings at 15-second
intervals, and you get...

http://www.projectpluto.com/nephele.png

This one runs about 1/3 path north of Steve Preston's result
(mentioned in your post) and about 15 seconds "early" relative to
his result. Steve has probably included some older observations
and done some better filtering of bad observations, and you should
take the path on his site in preference to the one I just hastily
computed. (Steve ran some tests comparing OCCULT, his
program of choice, with Guide, using identical data. The programs
are in decent agreement, though there are differences of up to a
few kilometers that we've not puzzled out yet. We can't even say
which program is wrong.)

It looks as if both you (in Nashville) and me (in Maine) have
some chance of observing this event.

-- Bill