Observations and Data Reduction
Report WHIRCAM 09/10 Feb, 1996

Observers : Peletier, Beckman, Lourenso, Varela

Report run WHIRCAM 9/10 Feb 1996
================================

This is a report of the observing run on the WHT on Feb 9 and 10
of 2 half nights, as a start of the International Time Project on BARS.
Present were Sergio Lourenso, To~ni Varela, Reynier Peletier
and (the second night) John Beckman. At the same time Jordi Cepa and 
Mercedes Prieto observed on the INT. The report will start talking about
the weather conditions during the observations, will then mention the
galaxies that were observed, and finally elaborate about the reduction of 
the data. The reduced frames can be obtained from the BARS homepage:
http://kapteyn.astro.rug.nl/~peletier/bars/BARS.html, together with other
useful things like observing logs. Also available are frames with 
intermediate reduction results. Note that some of these files are very
large, so they better be transported at night.

Reynier Peletier, Mar 06, 1996



The observations
----------------

The run, and so the ITP, started off very badly, since 5 minutes
after changeover we had to close because of low clouds. We could open again
later, but we had to close and open again about 4 times more this night.
The second night the weather cleared up around 11.30 PM, leaving
2 good hours for the other ITP and the rest of the night
for us. This night was photometric, although high winds caused the
seeing to be mediocre (1.2-1.4''). The first night was maybe photometric,
but we repeated the objects anyway, to be sure, and to get some higher
signal-to-noise data.

Galaxies observed:
------------------


Feb 09:
-------

NGC 3359 (6x90s + 4x90s sky) K short
NGC 4123 (6x90s + 4x90s sky) K short
NGC 4314 (5x90s + 4x90s sky) K short


Feb 10:
-------

NGC 4151 (6x90s + 4x90s sky) K short
NGC 4340 (6x90s + 4x90s sky) K short
NGC 4596 (6x90s + 4x90s sky) K short
NGC 5383 (6x90s + 4x90s sky) K short
NGC 5850 (6x90s + 4x90s sky) K short

NGC 3359 (6x90s + 4x90s sky) J
NGC 4123 (6x90s + 4x90s sky) J
NGC 4151 (6x90s + 4x90s sky) J
NGC 4314 (6x90s + 4x90s sky) J
NGC 4340 (6x90s + 4x90s sky) J
NGC 4596 (6x90s + 4x90s sky) J
NGC 5383 (6x90s + 4x90s sky) J
NGC 5850 (6x90s + 4x90s sky) J

NGC 4314 (6x90s + 4x90s sky) K short
NGC 3359 (6x90s + 4x90s sky) K short
NGC 4123 (4x90s + 4x90s sky) K short ****  bad due to twilight


This means that we have J and K data for the whole core sample.
The data of the first night has been reduced. To get an idea of
the quality of the data, you can download three fits-images here.
The pixelsize is 0.24''. We decided to use the same integration
times in J as in K', so that the signal to noise in J will be larger.

Data reduction
==============

The data were reduced in a routine way, which means that they were 
brought into a state that they are usable for everybody. It is possible
to do a better job, when one applies the full linearity correction,
does a more careful job on the photometry of the stars, or when one
removes by hand all remaining negative stars as a result of subtracting
stars in background field. However, I have made mosaics that consist
of all the data of a galaxy in a certain band, and all of them have been
calibrated photometrically, although some cirrus was present during the 
first night.

The steps that were taken are:

1. Making a mask of bad pixels. This was done by taking all pixels that
had efficiencies more than 20% different from surrounding values. Bad pixels
are not unambiguously defined. In any case, the frame contains many
bad pixels. However, they generally don't affect neighbouring pixels,
so in general the frames can be corrected quite easily for bad pixels by
taking a mask.
2. Making flatfields from dome frames, with frames with dome lights
off subtracted from frames with dome lights on. The flatfields were made taking
into account the bad pixel mask. On every night a separate flatfield per
filter was made.
3. Then I subtracted offset sky frames, that were somehow cleaned from
stars using median filtering, from each galaxy frame, after which 
a mosaic was made of the galaxy frames. The final mosaic was then divided
by the integration time (in seconds) of its individual frames.
A description can be found in Peletier (1993, A&A).
4. Then the standard stars were measured, and compared with the 
literature. Since very few standard stars had been observed, no 
extinction correction could be applied, but on the basis of the standard
stars we could convert counts to mag/sq. arcsec. 
5. Then I looked at frames of the dome interior of different integration
times to measure the linearity correction that had to be applied. 
Since the frames of all galaxies, except NGC 4151, had less than 5000 CTS
this linearity correction is less than 2%, so it was in the end not applied.


The final data are presented in FITS files in the tarfile whircam.feb.tar
The file contains also flatfields and the mask. The seeing as measured
on the frames is somewhere between 1.2 and 1.4''.

The observation logs are given in the files 960209.wht and 960210.wht. 
These are standard observing logs from the La Palma observatory.

Many useful files with intermediate steps can be found in int09.tar
and int10.tar. Here you can find individual reduced frames, files
with the frames that are used for a certain  mosaic, etc. All this
in iraf format.

Now about the data:
===================

the frame of NGC 4151 in J is saturated. Similarly, the total 
integration time for NGC 4151 in K' is less than for the other
galaxies by a factor of 3, because of the short integration time
per frame. Integration times can be found in the headers.
None of the other galaxies is saturated. Total integration time
per mosaic usually is 9 minutes + 6 minutes on offset sky, although
this number is sometimes slightly different. WHIRCAM suffers from
persistence effects: you will see a very bright object a few minutes
after integration has been stopped. In our case only the mosaics 
of NGC 4151 suffer; for example the small source to the side of the nucleus
is a persistence effect.

Linearity correction:
---------------------

We give here the integration time of exposures of the dome,
the number of counts (median in [76:115,36:75], the median/s, and
this value normalized:

t		median	median/s	norm.

0.2		789.7	3948.5		1
0.5		1977	3954.0		1.0014
1.0		3921	3921.0		0.9930
1.5		5856	3904.0		0.9887
2.0		7753	3876.5		0.9818
2.5		9619	3847.6		0.7744
3.0		11362	3787.3		0.9592
3.5		13097	3742.0		0.9477
4.0		14824	3706.0		0.9386
4.5		16498	3666.2		0.9286
5.0		18105	3621.0		0.9171

We note that in practise the linearity correction should never
be larger than 2%, except for NGC 4151.


Standard stars:
---------------

Here we and Rebolo observed on Feb. 9 the star FS12
(UKIRT Faint standard list) on 3 different occasions, every
time 5 times in Ksh (Ksh = K short is almost the same as K' or K).
They measured for instrumental mag (=26-2.5*log10(total_counts)):
(integration times always 10s):

A (frames 311-315):  15.55  15.60  15.60  15.61 15.57
B (frames 437-441):  15.56  15.55  15.56  15.54 15.56
C (frames 465-470):  15.62  15.59  15.64  15.57 

This star has a literature K-mag of 13.898 and J-K=-0.217


On Feb. 10 Rebolo etal observed FS17 at the beginning of the night,
while we observed FS25. Both in J and Ksh. We also measured FS33 and
FS27, but during the reduction the integration times of individual
frames proved to be too small to get reliable magnitudes.

Instrumental magnitudes for J and K are:

FS17 (frames 166-170,180-184):
   J: 14.63 14.63 14.55 14.61 14.65 (integration time 2s)
   Ksh: 15.61 15.55 15.62 15.62 15.64 (integration time 2s)

FS25 (frames 400-411)
     J: 12.89 12.87 12.89 12.86 12.85 (integration time 1s)
   Ksh: 13.81 13.80 13.84 13.80 13.80 (integration time 1s)


Lit: FS25 - K=9.756, J=10.231
     FS17 - K=12.270, J=12.681

This gives the following average photometric constants:
(phot. const = instr. mag(1s) - lit. mag)

K (FS17) = 4.094
K (FS25) = 4.054
J (FS17) = 2.689
J (FS25) = 2.639

and for Feb. 9 we get an average phot. const. in K:

K (FS12) = 4.197 which is probably an indication of some clouds.

We then finally come to the formula to calculate surface 
brightness from the pixel intensities. Here we take for the 
pixelsize 0.246''. I have not checked this number yet. If it changes
the formula changes slightly:

SB = 26 - pcon - 2.5 log10(Counts) - 3.061 

and pcon is about 4.07 for K and 2.66 for J.

Note that the FITS-files contain intensities. Here the final constant
of the residual sky background still has to be subtracted!


To conclude: 
============

We have obtained images of the central say 100x100'' of 8 barred galaxies.
They are not very deep, and the seeing is not great. However they can
be used to study the potential in the regions and possibly other things.
NGC 4151 has to be redone. I will try this in my own time at the end of
April. The rest has more or less the same quality. It will be nice 
to have higher S/N images with a larger FOV using ARNICA. However
in September most of the galaxies described here will not be visible.

HOME