Executive summary: The TASS Mark IV cameras measured Var Cas 2006 on seven occasions in Oct-Nov, 2006: one before max light, one shortly after max light, two about 10 days after max light, and three more about 20 days after max light. On each occasion, the Mark IV units made V-band and I-band measurements. The I-band measurements are much better than the V-band measurements in this region, and the main result of this work is to present I-band data on this star before, during, and after the event. The combined V and I measurements suggest that the variable was slightly redder at max light than before and after, not by a significant amount (given the large uncertainties in the V-band magnitudes).
In late October, 2006, Japanese astronomer Akihiko Tago reported that a star in Cassiopeia was growing brighter by over two magnitudes:
From CBET 711, distributed Oct 31, 2006:
S. Nakano, Sumoto, Japan, reports the discovery by Akihiko Tago (Ayabe, Tsuyama, Okayama-ken, Japan) of a brightening star on 30-s CCD frames taken with a 70-mm f/3.2 lens and a Canon EOS 20Da digital camera (limiting magnitude 12), with the following magnitudes available: Oct. 25.538 UT, 10.7; 27.409, 10.5; 30.411, 8.8; 31.469, 7.5. Tago adds that a star of mag 11.8 (which did not vary on past images) is located very close to the new variable's position, for which he identifies GSC 3656.1328 with R.A.= 0h09m21s.81, Decl.= +54o39'43".8, equinox 2000.0);
Below is a finding chart of the area around the variable; the field is about 20 arcminutes on a side, with North up and East to the left. The variable is marked by the vertical bar, with a nearby star of similar (red) brightness circled and labelled "Z1". The labelled stars are identified later in this document.
Many people started taking images and spectra of the object. Although the star appeared blue at maximum light -- like several types of outbursting variables -- it did not redden as it faded. Moreover, it faded very rapidly at first, spending no more than a day at peak light. Strange. You can find many discussions of the star's peculiar behavior on newsgroups:
On November 3, 2006, M. Mikolajewski and colleagues suggested in Astronomer's Telegram number 931 that the variable might be an example of gravitational lensing. This would be very unusual, since the source star is much closer to the Sun than all other lensed sources we've detected to date, and in a much less crowded field.
If the gravitational lensing theory is right, then the source star isn't really changing its properties at all. We should see the star remain the same color throughout the lensing period. The I-band measurements presented here might be useful in that regard, when combined with measurements made in other passbands by other observers.
The Amateur Sky Survey (or TASS for short) measures large areas of the sky on every clear night from the roof of Tom Droege's home in Batavia, Illinois. You can read a detailed description of the Mark IV cameras in TASS Mark IV Photometry Survey of the Northern Celestial Hemisphere , a paper which will be published in the December, 2006, issue of Publications of the Astronomical Society of the Pacific. You might also look at TASS Show-and-Tell number 10 , which includes lots of pictures and not so much text.
The Mark IV cameras do cover the field around the variable on a regular basis. Unfortunately, the Mark IV units are not well suited to the crowded star fields in this portion of the sky: they have large pixels, about 7.5 arcseconds on a side, which cause stars to blur into one another. Below is a small portion of a Mark IV image in V-band of the region around the variable.
And now in I-band:
Note that all the stars have much higher signal in the I-band images. The I-band measurements are more reliable for that reason.
Most of the labelled stars were measured by Arne Henden:
# RA Dec V (B-V) I my label # ----------------------------------------------------- 00:08:38.76 +54:35:49.5 10.503 0.447 9.922 A 00:08:48.46 +54:37:26.5 10.537 0.453 9.997 C 00:08:16.15 +54:37:44.3 10.605 0.334 9.941 D 00:08:34.32 +54:36:15.4 10.824 0.995 9.819 E 00:08:11.58 +54:37:12.3 10.924 0.298 10.555 F 00:08:52.43 +54:47:41.1 11.544 0.582 10.829 G 00:08:27.48 +54:41:10.1 11.556 0.659 10.799 H #-------------------------------------------------------
The star labelled "Z1" in the chart above is only 34 arcseconds to the northeast of Var Cas 2006. This is close enough to cause the TASS image reduction pipeline software to detect only a single merged object at their position; the merged object is sufficiently elongated to fail a roundness test. As a result, the merged object is discarded from the pipeline. Therefore, despite over 40 visits to the field during the past five years, the Mark IV engineering database contains only a single measurement which may refer to Var Cas 2006 (but may be the neighboring star Z1 instead).
In addition, some recent Mark IV images have been deleted after they have been processed --- only the properties of detected sources were saved.
Not all is lost. Tom Droege found some recent images of the field which had been saved. He was able to send me seven (V, I) image pairs taken during October and November, 2006. The initial letter "j" in the file names indicates that they were taken by the third Mark IV unit, nicknamed "TOM3", which concentrates on regions between Declination +54 and +88.
mid-exposure
JD-2,450,000 exptime filter name
-----------------------------------------------------------------
4022.67517 100 sec V jvr_40221740.fits
4022.67517 100 sec I jir_40221740.fits
4042.63779 100 sec V jvr_40421373.fits
4042.63779 100 sec I jir_40421373.fits
4048.63701 100 sec V jvr_40481359.fits
4048.63701 100 sec I jir_40481359.fits
4049.62360 100 sec V jvr_40491225.fits
4049.62360 100 sec I jir_40491225.fits
4059.59191 100 sec V jvr_40590908.fits
4059.59191 100 sec I jir_40590908.fits
4060.60326 100 sec V jvr_40601021.fits
4060.60326 100 sec I jir_40601021.fits
4061.59814 100 sec V jvr_40610970.fits
4061.59814 100 sec I jir_40610970.fits
-----------------------------------------------------------------
I subtracted the master dark image for each camera which Tom created as part of the usual Mark IV operations. I also divided by the master flatfield image for each camera, again created as part of the usual Mark IV operations. I also rotated the Mark IV images to match the normal convention of North up, East to the left.
I then analyzed the images in the following manner, using routines from the XVista suite.
At this point, I noticed that several of the stars in Henden's list were so close that my routines were failing to detect them individually; specifically, Var Cas 2006 and Z1, A and E, D and F. Therefore, I went through all the images and manually "removed" one star at a time from each close pair (by pasting a nearby section of blank sky on top of it), then measured the position of the remaining star. I put these revised positions into the list with all the other stars which had been detected without a problem. Continuing on my merry way, I then
I gathered together all the measurements of the variable and the comparison stars from Henden's list, and put them through inhomogeneous ensemble photometry ; in essence, I used all the other stars as local standards to determine the best zeropoint offsets to apply to each image. In the course of this work, I found that the star I called "C" had a much larger scatter than others of similar brightness, so I removed it from the set of local comparisons. This left 6 local comparison stars in V and 7 in I, roughly spanning the range of the variable's brightness.
One of the results of my ensemble photometry package is the shift which must be applied to each image. The V-band adjustments, shown below, indicates that the first and fourth images were taken during the clearest conditions; the penultimate image, on JD 2,454,4060 (= MJD 4060), suffered nearly 0.3 mag of extinction (as a casual inspection of the frames would suggest).
In the I-band solution shown below, the shifts are smaller (as one would expect for atmospheric extinction) and have a different pattern. Once again, the image from MJD 4060 requires the largest adjustment.
The ensemble procedure also provides an estimate of the variation in each star's instrumental magnitude from night to night. We can use these variations to estimate the uncertainty in the measurement of the variable star. Below is the "sigma-vs-mag" plot for the V-band; note that the horizontal axis shows differential magnitude, with the brightest star in the ensemble given a value of 0.00. The outlier with an average magnitude of about 0.40 is Var Cas 2006.
The ensemble solution in I-band yields much smaller residuals from the mean differential magnitudes. One reason is the higher signal-to-noise ratio in the I-band images. Var Cas 2006 is again the outlier.
Here are a couple of during-and-after pairs, showing the field when Var Cas 2006 was bright, a few days after its peak, and then faint again weeks later. You can click on each image to retrieve a larger version (and the larger versions are exactly the same size as each other, too).
First, V-band:
Now, I-band:
You can also look at an animated GIF showing the field in V-band before, during, and after the event. The images making up the animation are linked below.
For background on photometric properties of the Mark IV units, one might read portions of the TASS Mark IV Photometry Survey paper. The bottom line is that one might expect a small color term in the V-band measurements, but no significant color term in the I-band.
To convert the differential instrumental magnitudes from the ensemble into Johnson-Cousins magnitudes, I started with the same set of nearby comparison stars. I used Henden's photometry to compare the instrumental differential magnitudes produced by the ensemble procedure to the standard Johnson-Cousins scale.
For the V-band images, I used Arne Henden's set of V and (B-V) measurements of the comparison stars. The does not appear to be any significant color term in the small set of comparison stars (not a big surprise, since they were selected to match roughly the blue variable star). I find a mean offset of (ensemble V - Henden V) = -10.52 +/- 0.02 mag .
In the first version of this document, I used I-band magnitudes from the Mark IV "patches" catalog as the reference values. The current version uses Henden's much more accurate photometry.
In the I-band, again using Henden's photometry for the comparison stars, I again find a good solution without a color term.
I find a mean offset of -9.80 +/- 0.03 mag .
So, with these zero-point offsets in hand, I can shift the ensemble photometry to the standard Johnson-Cousins scale and make light curves for Var Cas 2006 and the nearby comparison stars. I will first show the results graphically. Here are the V-band light curves: it appears that Var Cas 2006 was brighter than usual by about 0.65 mag on MJD 4042.
And the I-band light curves; they suggest that the variable brightened by a bit more on MJD 4042, about 1.0 mag.
As I mentioned earlier, it is this I-band record which might be the most valuable contribution of TASS to the study of this star.
Is it possible that the apparent slight reddening of the variable star during its bright phase is real? It depends on the uncertainty one attaches to each measurement. There are common uncertainties of 0.02 mag in the V-band measurements, and 0.03 mag in the I-band measurements ... but these systematic uncertainties affect all values, and cancel out when one looks at changes within a single passband. In the table below, I attempt to estimate an uncertainty for each magnitude measurement of Var Cas 2006 by looking at the scatter from the ensemble solution of the comparison stars.
# JD V Verr Vsys I Ierr Isys #----------------------------------------------------- 4022.67517 11.22 0.09 0.02 11.09 0.04 0.03 4042.63779 10.58 0.03 0.02 10.15 0.02 0.03 4048.63701 11.10 0.08 0.02 10.94 0.05 0.03 4049.62360 11.17 0.08 0.02 10.93 0.05 0.03 4059.59191 11.40 0.09 0.02 11.19 0.05 0.03 4060.60326 11.19 0.09 0.02 11.13 0.05 0.03 4061.59814 11.25 0.09 0.02 11.15 0.05 0.03 #------------------------------------------------------------------------------
We can then compute the (V-I) color of Var Cas 2006 for each night, with a formal uncertainty, assuming that the errors in the two passbands are uncorrelated.
# JD-2,450,000 (V-I) random uncert systematic uncert #------------------------------------------------------------------------------ 4022.67517 0.13 +/- 0.13 +/- 0.05 4042.63779 0.43 +/- 0.05 +/- 0.05 4048.63701 0.16 +/- 0.13 +/- 0.05 4049.62360 0.24 +/- 0.13 +/- 0.05 4059.59191 0.21 +/- 0.14 +/- 0.05 4060.60326 0.06 +/- 0.14 +/- 0.05 4061.59814 0.10 +/- 0.14 +/- 0.05 #------------------------------------------------------------------------------
If we look at the colors of all the comparison stars during the 7 nights, as well as the color of the variable, we see that
I conclude that the reddening suggested by the Mark IV measurements is not significant.
Last modified 11/20/2006 by MWR.
