NSFCAM2 Plate Scale Measurement
Abstract:
The plate scale of NSFCam2 was measured by monitoring the motion of the asteroid (2) Pallas for 1.5 hours on the night of April 12, 2006 (UT). The plate scale was measured to be 0.039706 ± 0.000017. The image orientation is north to the right and east down as seen with DV, with the detector aligned to the cardinal points within a few arcminutes.
Arguably the most accurate way to measure the plate scale of a new camera is to monitor the motion of a well observed asteroid relative to background stars long enough for the asteroid to move across a significant fraction of the camera’s field of view. This approach is generally more accurate that observing a binary star since the separation measurement in binary catalogs can be a century or more old, whereas an asteroid ephemeris can be easily generated for the asteroid in question for the time of observation.
This approach was used on the night of first light for NSFCam2. The asteroid (2) Pallas was used as it was well placed in the sky, was bright enough to see in a short exposure, had a number of reasonably bright background stars in the same field at the time, and had enough motion that a good measurement could be carried out in less than two hours. The data were taken on April 12, 2006 (UT) at J band, with 60s exposures and using a 9 point dither pattern. The data on Pallas is summarized below:
Image# x y peak UT Time RA Dec
0305 695.65 795.91 21589. 13:09:27.20 18:56:55.409 15:21:46.270
0306 708.76 802.73 18399. 13:10:36.62 18:56:55.429 15:21:46.772
0307 720.69 809.20 18137. 13:11:45.62 18:56:55.449 15:21:47.274
0308 731.51 817.11 17799. 13:12:54.47 18:56:55.468 15:21:47.770
0309 744.44 825.30 16044. 13:14:03.43 18:56:55.488 15:21:48.275
0310 758.70 831.57 18529. 13:15:12.17 18:56:55.508 15:21:48.777
0311 769.38 840.14 15982. 13:16:20.93 18:56:55.527 15:21:49.273
0312 782.85 847.33 10688. 13:17:30.74 18:56:55.546 15:21:49.785
0313 796.70 854.72 4753. 13:18:39.61 18:56:55.565 15:21:50.284
0325 1066.52 1005.99 24470. 13:43:18.25 18:56:55.981 15:22:01.041
0327 1103.71 1025.28 28172. 13:46:38.99 18:56:56.038 15:22:02.499
0328 1115.54 1032.05 26985. 13:47:47.72 18:56:56.058 15:22:03.000
0329 1128.73 1040.03 24458. 13:48:56.55 18:56:56.076 15:22:03.495
0330 1140.70 1046.54 30694. 13:50:05.40 18:56:56.095 15:22:04.000
0331 1153.78 1054.72 30304. 13:51:14.24 18:56:56.115 15:22:04.502
0332 1166.14 1061.38 34724. 13:52:22.99 18:56:56.134 15:22:04.999
0333 1178.44 1067.43 29340. 13:53:31.99 18:56:56.154 15:22:05.499
0334 1191.30 1075.88 33371. 13:54:41.07 18:56:56.173 15:22:06.001
0335 1204.00 1083.14 27323. 13:55:49.90 18:56:56.192 15:22:06.506
0336 1492.09 1242.53 32599. 14:22:05.99 18:56:56.633 15:22:17.954
0337 1505.46 1250.42 34876. 14:23:14.72 18:56:56.652 15:22:18.456
0338 1516.66 1256.95 31914. 14:24:23.42 18:56:56.672 15:22:18.952
0339 1531.12 1263.89 38613. 14:25:32.36 18:56:56.691 15:22:19.452
0340 1542.22 1270.62 36726. 14:26:41.10 18:56:56.710 15:22:19.951
0341 1553.48 1277.08 27244. 14:27:49.86 18:56:56.729 15:22:20.447
0342 1570.44 1285.26 31292. 14:28:58.41 18:56:56.749 15:22:20.948
0343 1579.30 1293.29 28823. 14:30:07.98 18:56:56.767 15:22:21.457
0344 1593.05 1299.76 27327. 14:31:16.75 18:56:56.787 15:22:21.953
Image#: The image number for which the following numbers apply
x: x-axis coordinate of Pallas in the aligned image
y: y-axis coordinate of Pallas in the aligned image
peak: Peak counts of the image of Pallas in the image
UT Time: UT time stamp from the image’s fits header
RA: Ephemeris RA for Pallas
Dec: Ephemeris declination for Pallas
The UT time from the fits header was used to generate the ephemeris location. It is likely that this time refers to the time at the start of the image. Since the exposures were 60s, and Pallas showed significant motion from image to image, the x and y coordimates represent the location of Pallas 30 seconds after the time stamp. However, only the relative motion of the asteroid is important, not its absolute position on the sky, for the purpose of this measurement. Thus, I do not believe this effect affects the plate scale result.
The ephemeris locations were taken from the Astroid Dynamics web site (http://hamilton.dm.unipi.it/cgi-bin/astdys/astibo). The maximum time resolution of the ephemeris is 1 minute. The RA and dec position used were derived from linear interpolation of the given ephemeris location to the exact UT time given in the fits header.
The x and y coordinates are from centroiding with imexam in IRAF. There may be a small dependence of this position on the size of the centroiding radius. These x and y positions of Pallas were measured only after the images were aligned on the background stars using the imalign IRAF routine. Depending on the image, 2 to 4 background stars were used for the image alignment. The cumulative errors in centroiding and alignment are reflected in the deviation of the observed motion from a linear path.
Figure 1 shows the motion of 2 Pallas across the sky versus the motion of the image across the detector. When plotting the motion of the sky on the Y-axis and the motion across the detector in pixels on the X-axis, the slope of the linear fit to the line gives the plate scale of the detector. The LINFIT IDL tool was used to derive the best fit line to the
data, yielding the measured plate scale of 0.039706 ± 0.000017 arcseconds per pixel. Figure 1 also shows the deviation of the data points in arcseconds from the linear regression line along the right side of the plot. The maximum deviation is just less than 0.1”, whereas the standard deviation is 0.0335”. This is rather good, given that the asteroid moved over
1000 pixels, and the measurements deviating from the ideal by an average of one pixel at a time when the FWHM of the image is ~18 pixels.
Figure1 : platescale.pdf
Figure 2 shows the motion in RA and dec across the sky and across the detector to measure the position angle of the orientation of the detector relative to the sky. Again, the LINFIT IDL tool was used to find the best fit line to the data. The position angle of the motion of (2) Pallas (29.1674) differs from the position angle of the motion of the image of Pallas (29.2425) by 0.075 degrees (4.5’) in the clockwise sense (i.e. the northern part of the image is too far east). This alignment is, of course, subject to the position angle of the rotator. Thus measurement represents the position of the detector at the time of the observation. It must also be noted that the images are seen in DV, the image orientation is north to the right and east down.
Figure2 : positionangle.ps
We have not observed any significant difference in focus between filters, or in plate scale between filters. We achieved good image quality at M' after focusing at K. Also, differencing J and K-band images of the NGC 5466 field did not show any signigicant difference in the plate scale between these two filters.