An Eclipse in the Trapezium

On the night of Wednesday, February 26th, 2003, one of the stars in the Trapezium cluster had an eclipse. I observed this event and took some images. The minimum of the eclipse seems to have been shortly after 22:00 (02/27/03, 08:00 UT); this is somewhat later than expected.

An eclipsing binary is a pair of stars orbiting each other in such a way that one star periodically hides the other. Most bright eclipsing binary stars were discovered long ago, but theta¹ Orionis A is an exception; its variability was first reported in 1973. I observed the latest eclipse and tried to determine the time of minimum brightness.

The observations were made from my home in Palolo valley on the night of February 26th, 2003. The sky toward Orion was generally clear, but occasional showers from clouds to the northeast forced me to move the telescope several times. Some images show motion due to winds shaking the scope. I used an Orion 8 inch dobsonian telescope with a focal length of 1200 mm. For visual observations I used a 14 mm eyepiece. For photography I used a Nikon coolpix 4500 digital camera coupled to a Scopetronix 18 mm eyepiece; further technical details are given below.

My first visual observation was made shortly after 20:00; at that time it was already quite obvious that theta¹ Orionis A was much fainter than usual. I alternated visual and photographic observations until rain forced me to stop just before midnight. By then, Orion was quite low in the west.

Images of the Trapezium. Star A is the one on the upper left; proceeding clockwise, the other stars are C, D, and B. The first frame was taken 9 days earlier, while the remaining frames were taken on the night of the eclipse. Local times are given in each frame.

Shown above are the images I obtained of the Trapezium. In the image taken 9 days before (02/18/03, 06:05 UT), star A appears at its usual brightness. The images taken on the night of the eclipse show that star A was much fainter than usual.

Both visual and photographic observations indicate that minimum brightness of theta¹ Orionis A occurred on or shortly after 22:00 (02/27/03, 08:00 UT). Visually, star A looked slightly fainter than star B at 22:00; by 22:45 it had brightened enough to be comparable to star B. In the photographs, star A appears about as bright as star B at 22:13, but generally brighter in images taken at other times. The minimum brightness seems to have lasted for about 1 hour.

According to an on-line article by Roger W. Sinnott at Sky and Telescope, the minimum of theta¹ Orionis A was expected to occur around 21:20 (07:20 UT). My data indicate that star A was still brighter than star B at that time, with the actual minimum occuring roughly one hour later. This conclusion is not airtight - star B is not an ideal comparison star since it too is an eclipsing binary. At face value, however, the data imply that the period of theta¹ Orionis A is slightly longer than the quoted value of 65.4 days.


The raw images have a resolution of 1024×768 pixels. I used the FINE image quality option to record the frames. All images were taken with the camera lens set to its minimum focal length (7.85 mm). To minimize vibration, the camera's self-timer was used to trip the shutter. The exposure was 0.25 seconds at f 2.6. Effective `film speed' was set to ISO 400. A similar set-up was used to record a comparison image 9 days earlier; however, this earlier image was taken using ISO 800.

Image processing was performed on a UNIX workstation. I examined the images using xv and manipulated them using convert. Raw images were cropped to 256×256 pixels centered on star C, and individually rotated so that all images have the same orientation. After rotation, the central 64×64 pixels of each image were extracted, interpolated to a resolution of 128×128, and lightly smoothed.


In retrospect, there are several things which would have improved the quality of the observations:

  1. Use separate telescopes for visual and photographic work to improve coverage in both modes and reduce variations in image quality due to refocusing, etc.
  2. Reduce the exposure time to minimize effects of vibration and avoid saturating the digital camera's CCD detector.
  3. Circumvent in-camera processing by increasing resolution to 2272×1520, using HI image quality, and turning off any contrast-adjustment and image-sharpening options.
  4. Take multiple frames in quick succession and average them later (however, only 12 high-resolution images can be recorded on a 128 Mbyte memory card).
  5. Use a remote control to trip the shutter without vibration.
  6. Attach a monitor to the camera to adjust the focus more accurately.
  7. Use an AC power supply for the camera to avoid problems with limited battery life.

Joshua E. Barnes (

Last modified: March 8, 2003