Total Lunar Eclipse: November 2003

by John Walker

Early on the morning of Sunday, November 9th, 2003 the Moon skimmed through the southern edge of the Earth's shadow in a total lunar eclipse visible (at least in part) from every continent except Australia. At my observing site in western Switzerland, the entire eclipse was visible with the Moon high in the sky throughout. Totality occurred between 02:06 and 02:31 local time (01:06--01:31 UTC). Remarkably for this time of Telescope and Olympus C3040 camers during lunar eclipse the year, the sky was clear from the start of the eclipse until just before the end of totality, when a light high altitude haze began to form. The air was extremely turbulent, however, and "seeing" was marginal--telescopic views of the Moon were blurred by the boiling atmosphere except for fleeting moments of calm when lunar features became sharp.

To photograph the eclipse, I set up my Brandon 80mm apochromatic refractor on its simple wooden tripod and altazimuth non-driven mount with mechanical slow motions. I attached an Olympus C-3040 digital camera to the telescope with a ScopeTronix Digi-T adaptor, which couples a variety of digital cameras to many popular 1¼ inch eyepieces. I used a 15mm ScopeTronix ST Plössl eyepiece for all the eclipse photography--it produces an image of the Moon's disc about as large as possible without vignetting by the camera's optical train, with enough margin to allow for small errors in centring the image in the frame.

The Digi-T adaptor couples the telescope's eyepiece to the camera's zoom lens (which is not removable). While this introduces additional optical components into the light path, it is convenient in that the camera's zoom can be used to obtain image scales intermediate between those of available eyepieces. I locked out the camera's automatic focusing mechanism and set the focus to infinity, then focused with the telescope's focuser, having first zoomed in to maximum optical and digital zoom for fine focusing with the LCD screen on the back of the camera. I then zoomed completely out and maintained that setting for all exposures. The LCD screen was left on continuously throughout the eclipse and used to centre the Moon in the frame. Leaving the monitor screen on eats batteries faster than a bathtub full of electric eels, but since I'd set up the telescope on the top floor balcony at Fourmilab, I was able to run an extension cord from the adjacent conference room and use the mains adaptor to power the camera.

I mounted the camera on a 90° "star diagonal" so the LCD monitor would be easier to see with the telescope aimed high in the sky, and also to better balance the camera and telescope on the tripod head, which is less robust than it might be. Also, having a heavy camera hanging downward from the eyepiece end of the telescope invites trouble from the Brandon telescope's drawtube, which has a tendency to slip when you can least afford it; mounting the camera at a right angle pushes down on the drawtube, holding it in place.

All exposures were made with the infrared remote control included with the Olympus camera, which permits triggering the shutter without touching the camera, thereby avoiding any risk of inducing vibration which might blur the image. The lens was set at f/2.6, near its maximum aperture. While stopping down a lens usually improves image quality (at least until diffraction sets in), with the eyepiece coupled to the camera lens, smaller apertures mask the exit pupil of the eyepiece, vignetting the image field. Exposure was adjusted entirely by setting shutter speed. At the start of the umbral eclipse, an exposure time of 1/500 second was used, increasing to 1/2 second during totality.

No film or digital image sensor has the dynamic range and logarithmic response of the human eye. Observing the Moon visually during the partial eclipse phase, the eclipsed portion of the Moon's disc was clearly visible to the eye from the time the Moon was about one quarter eclipsed all the way to totality, and lunar features on the eclipsed part were clearly visible in binocular views. An exposure long enough to show the eclipsed disc, however, grossly overexposed the uneclipsed portion (which, after all, was just as bright as the penumbrally eclipsed disc before the start of the umbral phase), causing it to bleed into the rest of the image. All of the partial eclipse photos below were exposed correctly for the illuminated portion and do not show the eclipsed part of the disc except for moments just before and after totality.

Most of the 142 pictures I took (aren't 128 megabyte SmartMedia cards great!) were blurred by the bad seeing. I've selected the images below from those I happened to snap during moments of relative atmospheric calm. There's no way to predict when the image will stop boiling--all you can do is take lots of shots and select the best after the fact. At least with a digital camera you don't break the bank on film and processing!

Toward the very end of totality, atmospheric haze began to form. You can see evidence of this in the last few images which show the end of totality--features on the lunar surface are less distinct and in the last shot the haze is evident around the emerging limb of the Moon. After that photo was taken, atmospheric conditions got increasingly worse, with much of the emergence phase obscured by light clouds. I ended the photographic session shortly after the end of totality when it became evident I wasn't going to accomplish anything by persisting.

The telescope and camera optical train produce upright (non-inverted) images but flip left and right. I digitally flipped the images back so they appear as the Moon does to the eye, then rotated each image Path of Moon through Earth's shadow so lunar North is at the top, instead of the rotated disc one sees when observing at the 47° North latitude of my site. This rectification of the images makes it more clear the Moon is crossing the southern edge of the Earth's umbra, as shown in this chart extracted from Fred Espenak's data sheet for the eclipse. Image processing was performed with the Netpbm toolkit, The Gimp, and Jasc Paint Shop Pro.

This was a very bright eclipse. Observers estimate the magnitude of the Moon at mid-eclipse as -3.5, which is about four times brighter that the eclipse of May 16, 2003 (which I didn't observe because it was only partially visible in Europe and I was clouded out to boot). Eclipses where the Moon crosses the edge of the umbra tend to be bright, but this one was bright even for an edge crossing. The "man in the Moon" was clearly visible to the unaided eye throughout totality, and lunar features were easily seen through binoculars. The eclipsed Moon had a coppery colour--the images below appear more red than my recollection of the visual appearance.

Below are 26 of the best images, chosen to show the progress of the eclipse from near the start of the umbral eclipse through the end of totality when deteriorating atmospheric conditions caused me to pack it in. Click on any of these reduced images to view the full resolution image in a separate window.

Click on small images for full resolution display.














Eclipse Animation

Click animation for full-resolution display.
The individual photos taken during the eclipse can be assembled into an animation which shows the progress of the Moon through the Earth's shadow. To avoid jerkiness in the animation, it's necessary to very carefully align the images with one another and insure there is no apparent rotation from image to image. Even if I hadn't decided to rectify the images so North was up, I'd still have had to adjust them to remove the apparent image rotation due to the shifting perspective from the observing site on the rotating Earth.

To avoid distracting frame-to-frame shifts in brightness and colour balance, I have adjusted the intensity of all the partial eclipse images to be comparable to one another and forced them to grey scale. The individual still images above are, however, "what the camera saw"--they have not been colour or brightness adjusted. The animation was assembled, optimised, and output with Jasc Animation Shop.

Lunar Eclipses: 2001 to 2020

Astronomy Resources at Fourmilab

by John Walker
November 2003

All images in this document are in the public domain and may be used in any manner without permission, restriction, attribution, or compensation. Back links to Total Lunar Eclipse: November 2003 are welcome.