Astronomers say that a once-in-26-year eclipse, predicted to occur in January, probably happened months ago without anyone seeing it.
Photometrists around the world have been psyched for months about the once-in-26-year eclipse predicted for the edge-on binary star Alpha Comae Berenices. We wrote it up in the January Sky & Telescope, page 50: "Vigil for a Unique Stellar Eclipse." The eclipse was due to happen just about now.
Well, oops. Lead investigator Matthew Muterspaugh (Tennessee State University) announces that the prediction was wrong. Three measurements of the binary made about a century ago accidentally swapped the primary and secondary stars. As a result, the eclipse (if it happened at all) already happened a couple months ago.
I have some bad news regarding Alpha Com and its predicted eclipse. After many difficulties in obtaining updated separation measurements, it has been discovered that the orbit model which predicted the time of the eclipse was in error due to 3 position measurements from ~100 years ago listing the binary position angle incorrectly by 180 degrees. Using these skewed the orbital model, despite the presence of over 600 additional measurements, in just enough of a manner to cause the timing prediction to be incorrect. The eclipse, if it did happen, was about two months ago (at which point, the system was only barely above the horizon at sunrise, so photometry would have been rather difficult anyways). I am working on writing up a formal report detailing the error, describing the events leading to the discovery and correction, presenting a corrected orbit, and the times of the eclipses. However, I wanted everyone to know this as soon as possible so that efforts might not be wasted.
Here's the full story.
Next chance? Late 2041.