From: Marek Kowalski (MPKowalski@lbl.gov)
Date: Fri Jan 13 2006 - 15:29:43 PST
Hello proposal writers,
I want to share my thoughts on the issue of F814 vs F850 from the SN
perspective in order to to start a discussion and make sure we come to
the optimal solution for both SN and cluster science. In particular I
think that F850 as the primary search band might have some advantageous
over F814 and I have the feeling that the SN argumentation has some
relevance for the clusters.
The two strategies we want to compare are:
1) 4/5 of the orbit is spend observing F850LP (four exposures) and the
rest is spend on a single exposure of F775W.
2) 4/5 of the orbit is spend in F814W and the rest in F850LP.
Here are some thoughts which go beyond the fact that it is not yet well
established that we can fully rely on the U-band for SN cosmology.
The color measurement is severely degraded from the switch to F814.
The color obtained from F815-F850 is about sqrt(2) worse than that
obtained from F775-F850 because of the smaller lever arm. If we search
in F814 the photometric error in F814 will be much smaller than that
obtained in F850 both because of the longer exposure time and the larger
throughput. F850 will completley dominate the error. If we compare
option 1) and 2), searching in F850 will have about a factor of two
lower error on the color measurement for SNe with z<1.2. Hence this
will effect the final measurement of the largest fraction of field and
cluster SNe and we would have to schedule a number of ToO orbits to
compensate for this.
I would think that the galaxy color-magnitude diagram would suffer as
well from the smaller lever arm and the imbalance of depth.
Searching in F814 would provide a significantly higher SNR up to a
redshift of about 1.45-1.5. Above that F850 becomes better. However, the
higher SNR at z<1.45 will not buy us more SNe, as we are finding them
anyhow. I think for galaxies the SNR turning-point might be be reached
at lower redshifts, as they are on average redder than a SN. Hanks
analysis will show what the gain for weak lensing is once the correct
weighting is taken into account, but I can imagine it is smaller than
the simple number count implies.
I hope we might discuss this further on our next telecon
Marek
-- ----------------------------------------------------------------------- Marek Kowalski Lawrence Berkeley National Laboratory mpkowalski@lbl.gov 1 Cyclotron Road, MS 50R5008 (510) 486-4652 Berkeley, CA 94720-8160 -----------------------------------------------------------------------
This archive was generated by hypermail 2.1.4 : Fri Jan 13 2006 - 15:30:55 PST