WFC3 Mini-Handbook notes

6/9/06: FrameMaker tutoring from Susan Rose:

maker
click Open
select .book file
double-click the item to be edited

For italics: use the script f at upper right, NOT Format


6/22/06: Jim Younger came up & told me how to print: if error message, open the
file first, then use the Print Book feature.  Also explained how to use "Figure
2" tags.



6/20/06: edited cover page & Ackn sections. Started editing main body.


Things for Susan to do:

1. Italicize Hubble Space Telescope, HST, and Hubble throughout document.



==============================================================================

7/5/06: Silvia came to discuss the table of filters, which led into a discussion
with several people of what info is best to include in the table. Consensus
seems to be:
	o some kind of wavelength
	o some kind of width
	o peak transmission
	o Stefano also suggests some kind of integrated efficiency
What kind of wavelength? There was some agreement that the "pivot wavelength"
may be useful.

Some google searching led to a note by Koekemoer & Brammer at
http://www.stsci.edu/instruments/wfpc2/Wfpc2_memos/wfpc2_filters_archive.html
who endorse pivot wavelength.

http://www.journals.uchicago.edu/ApJ/journal/issues/ApJ/v503n1/37428/sc2.html
notes that "The pivot wavelength is defined to allow exact conversion between
the HST broadband flux densities expressed per unit wavelength and per unit
frequency (Koornneef et al. 1986)."

Tokunaga & Vacca 2005PASP..117..421T : Appendix A gives an excellent discussion
of the various definitions:

"Effective wavelength":

  lam_eff * int[ Flam(lam) S(lam) dlam ] = int[ lam Flam(lam) S(lam) dlam ]

where Flam is the source SED and S(lam) is the system response.

"Mean wavelength": (independent of source)

  lam_0 = int[ lam S(lam) dlam ] / int[ S(lam) dlam ]


The number of photons detected is proportional to:

  Np = int [ (Flam(lam) S(lam)/ h*nu) dlam ]

     = (1/hc) int [ lam Flam(lam) S(lam) dlam ]    (eq. 1)

The mean flux is

  <Flam> = int [ lam Flam(lam) S(lam) dlam ] / int[ lam S(lam) dlam ]


Note that

  <Flam>/c int [ lam Flam(lam) S(lam) dlam ] = <Fnu> int [ S(nu)/nu  dnu ]

					     = <Fnu> int [ S(lam)/lam  dlam ]
  
     since dnu/nu = dlam/lam

So if we set  <Fnu> = [<Flam>/c] * lam(pivot)**2

we find    

  lam(pivot) = sqrt[ int [ lam S(lam) dlam ] / int [ S(lam) dlam/lam ] ]

==============================================================================