Hi Alessandro,
> I just realized that the dT difference depends heavily on weather the asteroid perturbation
> effect is taken into account or not...
AHA! You are absolutely correct about this. My apologies for misleading you about the
"timing error".
(70401) Davidbishop passed within about 400000 km of Pallas on 1982 Feb 27. It's really
rare to see cases where asteroid perturbations matter very much, but they certainly do here.
If you toss out the five worst observations and include asteroid perturbations, you get
an RMS error of 0.663". Shut off the perturbations and do a full step, and that bounces
back up to 1.124, with the 1950 observations having 14.5" residuals in RA.
You could actually use this to get a decent estimate for Pallas' mass. I've occasionally
thought about revising Find_Orb to do that sort of thing. It would basically just have to
solve for seven parameters: the usual six orbital elements it normally solves for, plus
the asteroid mass. Not a big deal, and possibly useful to people doing asteroid mass
determinations (I don't know what they currently use.)
-- Bill