Re: RPS2 plan version 2.0

From: Rachel G. (gibbo@panisse.lbl.gov)
Date: Wed Oct 29 2003 - 09:33:56 PST

  • Next message: Rachel G.: "Re: ACS/Grism data"

    Hi Greg,

            I'll be able to answer question #1. I have had
    a quick look at this, but should do so in more detail before
    quoting expected errors. I don't recall there being a
    significant difference in stretch or color errors. I think
    stretch is being mostly determined by the good z-band
    coverage when additional search points can be used.
            
            I'll also be looking more at the grism question.
    I don't believe it's feasible to go to z=1.5 in only 10ish
    orbits. I also have grism data for the published z=1.3 SN.
    I didn't grab the data from the Blakesley paper, but may do
    that too at some point.

    On Wed, 29 Oct 2003, Greg Aldering wrote:

    >
    > Hi Rob,
    >
    > In your plan there is a big jump in the number of orbits in going from
    > cases 1 & 2 to cases 3 and 4; I would suggest an intermediate pair of
    > cases, with 6 orbits in z (1,1,2,2), with the option of 9 orbits of grism.
    > This would handle SNe of intermediate brightness/redshift, or be an easy
    > means to adjust the S/N we wish to reach at a given brightness/redshift.
    >
    > I'd also like to rise the question of whether there should be additional
    > cases offering more J and/or H in the place of grism. We can study this in
    > detail later, but there could be benefit to having cases #2'and #2'' with
    > 0 grism and one more epoch of J & H (say 3 orbits of J and 3 orbits of H)
    > or two more epochs of J (3 orbits each). Likewise, cases #4' and #4'',
    > with 0 grism and one more epoch of J & H (say 4 orbits of J and 4 orbits
    > of H), or two more epochs of J (say 4 orbits each and then 1 more ref
    > orbit). One motivation would be that we could implement C-magic if we have
    > the additional color points, which would add a little novelity to our
    > program wrt Adam's; and we may be able to distinguish the SN type (albeit
    > with somewhat less opportunity, I presume, to abort later orbits than the
    > grism might afford) using the color curves.
    >
    > As it should not be hard to add these additional cases once the other
    > cases are entered, I would recommend adding these just so we don't have to
    > worry about them later (and then perhaps find that they can't be done). In
    > particular, as we must require that NICMOS orbits avoid the SAA, we should
    > lock-in as many such orbits we might possible want. Such orbits are very
    > valuable and so might be hard to get later on without a big ugly fuss.
    >
    > Two other issues that we can study after the RPS2 is submitted:
    >
    > 1) Should the z, J, H observations be at the same epoch - which makes the
    > color calculation easier - or at slightly staggered epochs - which
    > helps fill in the lightcurve and possibly make the fit more robust?
    >
    > 2) What is the appropriate number of NICMOS references? If we use
    > a model (continuous or pixel-by-pixel), then the observations
    > which contain the SN provide information and we wouldn't get all of
    > the sqrt(2) hit that pixel-by-pixel subtraction introduces. Thus,
    > where I have added J or H visits of more than 3 and 2 orbits,
    > respectively, we may not need to increase the number of references.
    > (This decision is very likely to come down to how well the PSF
    > NICMOS is expected to be known; Vitaliy is exploring this type
    > of question in relation to Albinoni.)
    >
    > - Greg
    >
    >
    > On Tue, 28 Oct 2003, Robert A. Knop Jr. wrote:
    >
    > > For each search, we should have for different follow-up possibilities
    > > available, to be chosen via TOO. Note that in all cases, the orbit
    > > count includes IR final references. Estimates below for R-I are based
    > > on the z-J color, *not* on the J-H color. Note that both z and J are
    > > contributing to the stretch here.
    > >
    > > #1 : Low-redshift, no grism needed.
    > >
    > > 4 orbits in z, spaced as Rachel did
    > > 3 orbits in J, spaced as Rachel did
    > > 2 orbits in H at the first epoch
    > > (save 3 J, 2 H orbits for final ref)
    > >
    > > TOTAL = 14 orbits
    > >
    > > WORST CASE z=1.3 (usually we do better!)
    > > ds = 0.09
    > > dt = 2.5
    > > dmbcorr = 0.15
    > > d(R-I) = 0.038
    > >
    > > ==> For a galaxy with a very good photo-z
    > >
    > > #2 : Low-redshift, grism needed
    > >
    > > 4 orbits in z, spaced as Rachel did
    > > 3 orbits in J, spaced as Rachel did
    > > 2 orbits in H at the first epoch
    > > 7 orbits grism
    > > (save 3 J, 2 H for final reference)
    > >
    > > TOTAL = 21 orbits
    > >
    > > Worst case as #1
    > >
    > > #3 : High-redshift, no grism needed
    > >
    > > 8 orbits in z, spaced as Rachel did
    > > 3 orbits in J, spaced as Rachel did
    > > 2 orbits in H at the first epoch
    > > (save 3 J, 2 H for final referenced)
    > >
    > > TOTAL = 18 orbits
    > >
    > > WORST CASE z=1.5
    > > ds = 0.08
    > > dt = 2.2
    > > dmbcorr = 0.14
    > > d(R-I) = 0.046
    > >
    > > ===> For a galaxy with a very good photo-z
    > >
    > > #4 : High-redshift, grism needed
    > >
    > > 8 orbits in z, spaced yadda yadda
    > > 3 orbits in J
    > > 2 orbits in H
    > > (save 3 J, 2 H for final reference)
    > > 11 orbits grism
    > >
    > > TOTAL = 29 orbits
    > >
    > > Worst case as in #3.
    > >
    > > -Rob
    > >
    > > --
    > > --Prof. Robert Knop
    > > Department of Physics & Astronomy, Vanderbilt University
    > > robert.a.knop@vanderbilt.edu
    > >
    >



    This archive was generated by hypermail 2.1.4 : Wed Oct 29 2003 - 09:34:09 PST