From: Rachel G. (gibbo@panisse.lbl.gov)
Date: Thu Sep 18 2003 - 11:38:24 PDT
---------- Forwarded message ----------
Date: Thu, 18 Sep 2003 10:13:55 -0700 (PDT)
From: Rachel G. <gibbo@jimbean.lbl.gov>
Reply-To: Rachel G. <ragibbons@lbl.gov>
To: Saul Perlmutter <saul@LBL.gov>
Cc: Ariel Goobar <ariel@physto.se>, scpexec@LBL.gov,
Gerson Goldhaber <G_Goldhaber@LBL.gov>, Jakob Jönsson <jacke@physto.se>
Subject: Re: lightcurve simulations
Hi Ariel and Saul,
I should be able to have the SNR calculations done
today. However, the ACS ETC was down over the last day+ and
is still not functioning. I've e-mailed to see what's going
on because I need to run a few examples to get the noise levels
right. But I can say at z=1.5 it would be hard to get decent
SNR in the Z-band with only 1 orbit per point - the following
is a lightcurve expected for a Ia at z=1.5 observed with
ACS/F850LP. I suppose I should lop off the last point if we
expect to have a point before max instead.
rest
date ABmag
0 25.20498
7 25.71454
12 26.3822
17 27.04559
On Thu, 18 Sep 2003, Saul Perlmutter wrote:
> Let's see... I think the most important constraint is that we have a
> total of 60 orbits to use for following-up the HST supernovae. We have
> been assuming that we won't end up with more than 3 good SNe to follow,
> but it is possible that we will waste some orbits on a dud, if we don't
> have perfect selection criteria and if we think that there will be lots
> of non-Ia "decoys." So we should aim to use an _average_ of 20 orbits
> for each supernova (and if we can use less that will allow for some
> wasted orbits following a dud). We will want to come up with something
> like six alternative scenarios: e.g. z = 1.2, 1.4, 1.55 with an
> eliptical host, and the same possibilities for a spiral host.
> Hopefully, our ideal strategies for many of these scenarios will end up
> being the same, so we won't have to send in to HST too many alternative
> target-of-opportunity set-ups.
>
> At this point we don't know of any constraints on the number and spacing
> of points on the lightcurve, except that we won't be able to start the
> first observation until at least ~11 observer's days past the discovery
> date. Our first guesstimate was that we would need 4 points along the
> lightcurve to identify the date of maximum and the stretch (and there
> would be some concern that without any redundancy a single bad
> measurement might ruin these fits).
>
> Rachel should be sending you the S/N estimates (how soon can you send
> these, Rachel?).
>
> One extra possibility that we might consider for the lowest-z SNe that
> we want to follow (z ~ 1.2) would be to use some orbits (~7?) for a
> grism spectrum. At higher redshift, this is unlikely to help much.
>
>
>
> Ariel Goobar wrote:
>
> > Hi,
> > with the help of Vallery, I managed to de-dust a version of
> > SNMinuit that fits our simulated lightcurve data. In order
> > to proceed, Jakob and I need the relevant input data for the
> > tests:
> >
> > 0) How many orbits do we have for lightcurve follow-up
> > 1) expected measurement error vs magnitude for Z-band and J (f110w)
> > as a function of number of orbits/point
> > 2) Constrains on the number and spacing of points for each SN
> > 3) Any other limiting factor we should know about.
> >
> > Cheers,
> >
> > Ariel
> >
> > --
> > ___________________________________________________________________
> > Ariel Goobar (www.physto.se/~ariel)
> > Department of Physics, Stockholm University
> > AlbaNova University Center, SE-106 91 Stockholm, SWEDEN
> > tel: +46 8 55378659 fax: +46 8 55378601
>
This archive was generated by hypermail 2.1.4 : Thu Sep 18 2003 - 11:38:24 PDT