Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/docs/documents/Appendix/omphot_jan01/tn.meet0101.min Дата изменения: Wed Jan 17 19:23:57 2001 Дата индексирования: Sat Dec 22 14:47:03 2007 Кодировка: Поисковые слова: ori

Minutes of OM Photometry Meeting Jan 10, 2000 at MSSL
====================================================

Attendees: Alice Breeveld, Igor Antokhine, Christopher Brindle, Jon Mittaz,
Simon Rosen, Antonio Talavera, Rudi Much

List of Actions:
================
a090101.1 Rudi: Redo colour transformation splitting the full range into
colour intervals. At the same time the spectra should be
sorted according to metalicity.

a090101.2 Igor: Igor will provide the colour intervals for which the colour
equations have to be calculated.

a090101.3 Rudi: investigate the impact of changing the fudge factor at the
red end.

a090101.4 Jon: Jon will estimate the size of the coincidence loss in the
background annullus around a bright source and will
distribute this information.

a090101.5 Igor: Igor will parametrized the empirical linearity curve

a090101.6 Rudi: specify changes to handle parametrized linearity curve and
update the documentation accordingly. The changes will affect
CCF and CAL.

aa090101.7 Igor: provide Rudi with the G158-100 photon rates

aa090101.8 Rudi: compare the measured with the simulated G158-100 photon rates

a090101.9 Simon: Run the SAS on the observation used by igor in the photometric
photometric calibration.

a090101.10 Simon: Implement the revised logic into the SAS photometry.

a090101.11 Simon: Change OMDETECT so that it write the source extraction radius
to the source list output.

a090101.12 Alice: Characterize the PSF as function of CFRR, especially idenify
whether there is a radius which contains the same fraction of counts
irrespective of CFRR.

a090101.13 Antonio: Antonio will talk to the experienced IUE personnel
how to improve the UV filter calibration.


Current photometric calibration (Rudi):
=======================================
The short cuts of the current colour transfromations are, that there is only
one transformation over the full colour range defined, no distinction between
metal poor and metal rich objects is made (branching) and the colour equations
are defined with the standard magnitudes on the right hand side.

a090101.1 Rudi:
Redo colour transformation splitting the full range into colour intervals.
At the same time the spectra should be sorted according to metalicity.

a090101.2 Igor:
Igor will provide the colour intervals for which the colour equations have
to be calculated.

The colour equations will be changed, so that the standard magnitudes do only
occur on the left hand side of the equation and branching of the colour
transformation will be made possible (for both items see also below).

Current calibration is largely based on the results of the simulator written
in IDL. An empirical fudge factor was introduced in order to match the measured
values of BPM16274 with the simulations.

a090101.3 Rudi: investigate the impact of changing the fudge factor at the
red end.



Photometric Calibration based on ground based observations (Igor):
=====================================================================
The transformation from the WFI instrumental into the standard system
are not yet available. Current analysis is done in the WFI instrumental
system.


The rates were extracted from the 2x2 subpixel binned data in the areas:
soure radii 0-13 subpixel, background radii 14-18 subpixel. The background
annulii adjecent to the area of bright sources may suffer coincidence losses.
The instrument electronic simulator at MSSL may provide an estimate of this
effect.
a090101.4 Jon: Jon will estimate the size of the coincidence loss in the
background annullus around a bright source and will
distribute this information.

a090101.5 Igor: Igor will parametrized the empirical linearity curve
a090101.6 Rudi: specify changes to handle parametrized linearity curve and
update the documentation accordingly. The changes will affect
CCF and CAL.


There are 5 observations of the Exo field available.
Igor compared the source brightnesses measured in different exposures
and observations. The uncertainty (sigma) of the brightnesses increases
for fainter sources as expected due to worsening of the photon statistics
(upper diagram).
However unexpectedly the scatter of the brightnesses between different
observations increases more than expected towards fainter sources (lower
diagram). The scatter is of the same size if relative photometry is applied,
i.e. the brightnesses are expressed relative to a reference star. Ideally a
reduction of the scatter is expected if the data of more exposures are analysed,
because more measurements should reduce the statistical error.
The origin of the large scatter is not understood at the moment, however a
possible explanation are deviations from a uniform flatfield.
Hopefully the planned raster pointing (again Exo field) will shed some light
on this issue. This is ongoing work, no action.


Now as the linearity correction is improved, brighter spectrophotometric
standards can be used to verify IDLsim. The observation of G158-100 is an
example.
a090101.7 Igor: provide Rudi with the G158-100 photon rates
a090101.8 Rudi: compare the measured with the simulated G158-100 photon rates

Future Photometric calibration (Igor):
======================================
The colour transformation of the ground based data is limited to red stars.
It is needed to extend the colour transformation towards blue colours.
The search for fields with blue stars is continued at Liege.
Sofar two fields were identified. The globular cluster around LMC X-3,
which expands the colour range slightly and NGC346, which is very crowded
(cf. Massey et al. AJ 1989). Ongoing work.

Stars of different metalicity require different colour transformation, e.g.
a branching of the equations has to be supported. Addressing of the branches
is done via an optional input parameter. The colour transfromation of hte
main sequence are executed by default. The transformation will differ in
the order of 0.1 mag. It was agreed that the required s/w changes will be
implemented and the CCF files populated as needed.


Current implementation of photomertry in SAS (Simon/Jon):
=========================================================
The main open issues concerning photometry are
- how to scale the extracted counts for the aperture area
- how to handle extended sources
- what to do with confusion/blended sources

The change on the linearity correction is already implemented and it uses
now the raw counts, rather the background subtracted counts.


OMDETECT calculates the source flux in the following way:
the extracted flux is the sum of all connected pixels having a significance of
more than 1.5 sigma above background.

The need to analyse the photometric OM observation with the SAS was identified.
This offers the possibility of a scientific verification of the SAS flux
extraction/photometry against Igor's results.
a090101.9 Simon: Run the SAS on the observation used by igor in the photometric
photometric calibration.


OMDETECT provides an intial source characterization. The flux calculation
requires further work. The following logic shall be implemented:

if point source then
if not confused then use default aperture
if blended then deblend, use smaller aperture and scale
if multiple objects use smaller aperture and scale
endif else (extended source) then
if not blended (isolated extended source) use one out of
* n-sigma aperture
* curve growth
if extended and confused: leave as is now
endelse

a090101.10 Simon: Implement the revised logic into the SAS photometry.

Changes required for OMDETECT:
The extraction radius should be added to the source list.
The background count rate underneath the source is already included.

a090101.11 Simon: Change OMDETECT so that it write the source extraction radius
to the source list output.

The use of non-default aperture radii requires scaling of the fluxes. As we are
using a fixed aperture in the calibration data analaysis, the PSF will be
set to 1.0 at the default aperture (12/35 subpixel for the visible/UV filters).
A good knowledge of the PSF is required if the fluxes have to be scaled.
The shape of the PSF is count rate dependant. There might be an optimum
radius (beside the default one), where PSFs at different CFRR (count to frame
ratios) intersect. Such a intersection would define a preferred radius.

a090101.12 Alice: Characterize the PSF as function of CFRR, especially idenify
whether there is a radius which contains the same fraction of counts
irrespective of CFRR.


Once the PSF was derived and the CCF was updated, the source count will be
calculated by scaling the source counts with the PSF, add the background
counts within the aperture area and then apply the coincidence correction.

Changes to CCF/CAL (Rudi):
==========================
Items identified which are not included in the VGs are:
- the aperture extraction keyword used during calibration is filter dependant.
Therefore a filter dependant keyword in the CCF is required (CCF colortrafo).

- a new CAL call is needed to return the aperture radius in the calibration.

UV filter:
==========
UV filter calibration is rather flacky at the moment. Magnitudes set arbitrarily,
non fluxes are not calibrated absolutely. IUE calibration was based on a few
white dwarfs. The relevant people are still around at Vilspa.

a090101.13 Antonio: Antonio will talk to the experienced IUE personnel
how to improve the UV filter calibration.