VR Photometry of Open Cluster NGC 6819

Open cluster NGC 6819 was imaged through V and I filters to a limiting magnitude of 15 under mostly photometric conditions. A calibrated color-magnitude diagram was constructed using crowded field photometry tools within IRAF(Image Reduction and Analysis Facility). Correction for line-of-sight reddening was applied and results are compared to Kalirai ET AL 2001.

1. Introduction

The project goal was to create a color magnitude diagram for NGC 6819 in which the main sequence is apparent. Then compare with a 2001 paper by Kalirai ET AL. In preparation, I calculated exposure times based on the published turnoff magnitude of 15. This assumption was incorrect and will be discussed along with the results below.

2. Observations

Three Evening flats were taken in each of the V and R filters with the CCD at a uniform temperature of -15 degrees Celcius. The same was done with the morning flats and all flats had acceptable counts. 5 x 30 s and 5 x 120 s darks were taken without any problems. Then, 7 x .11 s zeros were taken. Following all the calibration frames, a single 120 s exposure in V and R filters of NGC 6819 was taken. Then six standard stars were exposed through V and R filters. Followed by annother set of exposures on NGC 6819 when is was <10 degrees off of our zenith.

Figure 1. A fully reduced, color composite, 18'x11' image from which photometry will be obtained. FWHM is about 7 pixels.

3. Reductions

Now that the basic reductions were complete, I was able to start the photometry. The first order of business was to fix the image headers by correcting the exposure time and adding an airmass value (the GOT cannot do this automatically). Then it was time to choose an aperture size for photometry of the standard stars. It took a few tries to figure out that the values the users guide assumes are quite a bit smaller than the ones we achieve with the GOT. Values which gave me good results were a FWHM of 7, an aperture radius of 30 pixels and an annulus radius of 40 pixels. Next, a considerable amount of time was spent setting up the parameters in phot. It was at this point that I realized I was going to lose a significant amount of my standard stars simply because I could not locate them in the frame with any reasonable amount of confidence. On account that I couldn not locate some of the standards myself, I assumed I should not use the automatic star finder in daofind. After locating three of my seven standards by hand, I used phot to construct a list of the magnitudes of the standards in each frame. This worked well, and I was reasonably sure the ones I have located were actually the ones I was after. The mknobsfile task was the used to create a list of all the standard stars and their information. This file will be used later by the transformation equation solver.

Setting up the transformation equations made me realize another unfortunate occurrence: all my standard stars were at almost exactly the same airmass! It was determined by people of greater knowledge than I that I could still calibrate the magnitudes with these stars. Then I could find a correction factor by comparing my 2 exposures of NGC 6819 that I took at different airmasses. This means my transformation equations will not have an airmass or extinction value. My two equations were mV=V+v1+v2*VR and mR=(V-VR)+r1+r2*VR.

Now it was time to start working with daophot; which is IRAF's crowded field photometry tool. After the initial parameters setup, the first step is using daofind to locate the coordinates of all the stars on the image. These will eventually be sent to phot so the magnitudes can be obtained. The threshold and sky standard deviation required a little tinkering to keep daofind from locating noise as stars or skipping over stars that I wanted to include in the photometry. I used tvmark to mark the located stars until I was happy with the selection.

Figure 2. This is the image displayed by tvmark. The red dots indicate selected stars. Notice only stars in the vicinity of the cluster have been located. This cuts down on field contamination.

Making the point spread function (PSF) was by far the trickiest part of this data analysis. It took many attempts before I was happy with the star subtraction which leads me to believe the PSF has some variance across the field. Below is an image after running substar. The subtraction looks satisfactory, but closer inspection reveals some disturbances where the brighter stars used to be.

Figure 3. Object frame with all marked stars subtracted. Compare with the first image and note the difference.

I then located all the central, distinct stars I could and added them to the coordinate list. Some of the fainter stars can be recognized, but noise and the disturbances from the previous star subtraction prevents daofind from locating them and makes manual location uncertain. Aperture correction was determined automatically using the .psg and .nsg files that allstar created during the star subtraction step. Lastly, transforming the instrumental magnitudes into the standard system makes use of tools mkobsfile and invertfit. These apply the transformations that were solved earlier and output a photometry file with corrected V and V-R magnitudes.

I found the error in V to be +/-0.021737 Mag and a V-R error of +/-0.0118292 Mag. These errors were then applied to the 200 selected stars in the cluster. Figure 4 shows a plot of the V errors. Notice how rapidly the error values increase as the stars get fainter.

Figure 4. Errors in V. Notice they increase with magnitude.

Figure 5. Errors in V-R. Notice they decrease with redness.

4. Results

Figure 6. Color-magnitude diagram before extinction correction.

Figure 7. Callibrated color-magnitude diagram with extinction correction.

Figure 8. C-M diagram from Kalirai ET AL 2001.

5. Discussion

Although my goal of measuring the age of the cluster was not obtained, the red clump was observable with reasonable accuracy. When compared to the published C-M diagram, the visual magnitude is almost the same and the color is also close after it is convered to B band. I do believe this project is achievable with the equipment used, and my errors will undoubtedly prove helpful to others in future observing runs with the GOT.

References