From: Rollin Thomas (rthomas@panisse.lbl.gov)
Date: Mon Jan 19 2004 - 11:27:12 PST
[Gabriele and I have been having some back-and-forth off the list, so I
am merging that discussion with this one.]
On Mon, 19 Jan 2004, Gabriele Garavini wrote:
> the data you downloaded are the same I show in the plot. Do you have
> any comment on the plot itself? At the phone conference you wanted to get
> to see that plot to draw some conclusion if I don't remember wrong.
You comment in this thread that the wavelength error can't account
for the overall low velocities measured, but they do account for a
substantial fraction of the drop from day +8 to day +11. Your re-analysis
should help clear this up.
The velocity drop from day +8 to day +11 of 1200 km/s seemed very
steep. Usually around this time, the Si II velocity starts to level
off. I was interested in seeing the entire feature in velocity space,
so I was really glad you included the part of the plot going to the
red side of the emission peak. In a single unblended line, if the
photosphere is receding, the line profile should stay peaked at the
rest wavelength, but the overall shape will change.
However, notice that from day +16 to later, the peak of the ostensible
Si II feature has shifted out to -8000 km/s. There is clearly a lot
of blending going on here, starting somewhere before day +16. It is
bad enough that it stretches well into the Si II emission feature, and
takes a big bite out of the blue side of the emission feature. What
is bad is that one of the new peaks visible from day +16 (most
prominent at +24) overlaps a bit with the Si II minimum. This will
tend to cover up the position of the minimum, and the notch you
measure will more likely not be attributable to Si II. In fact, it
will move left more than it should. Have you thought deeply about
this effect? If so, have you got any conclusions. Synow is
explicitly intended to handle line blending. Have you at least given
it a shot?
If this bothers people on the paper too much, you might consider going
with what everyone can agree upon. The value of v10 appears to be
something like 8700. BTW, did you use the original method of
measuring v10 or or something else like interpolating between +8 and
+11? If you use Weidong's value of dm15 of 1.3, then this puts 99ac
on a line connecting 1980N and 1989B. This part of the space is not
so inhabited, but it isn't as far out as 02cx.
Rollin
This archive was generated by hypermail 2.1.4 : Mon Jan 19 2004 - 11:27:21 PST