On Monday, August 21, 2017 the first total solar eclipse to hit the mainland U.S. in nearly four decades traced a path from Oregon to South Carolina. I witnessed and imaged this event from central Wyoming.
The last total eclipse visible from the continental U.S. occurred on February 26, 1979. I drove up to Walla Walla, Washington, to witness it, my first total solar eclipse. While that eclipse just grazed the northwest, this last one swept coast to coast, within easy travel range for tens of millions of Americans, and occurred in a time when the Internet has made information about such things extremely accessible. The country saw an eclipse frenzy.
I had been aware of this eclipse for many years, but first began serious preparation in mid-2015, which the organizers of the Astronomical League conference contacted me about being a speaker at the eclipse-themed 2017 meeting to be held in Casper, Wyoming, on the eclipse centerline. Since then I've been researching and purchasing equipment, studying weather patterns, designing an imaging strategy, and practicing everything in preparation for just over two minutes of totality.
I decided to shoot two separate streams of data. Telescopically, I used a Canon 7D camera on a Stellarvue 102A refractor riding on an iOptron CEM25P equatorial mount. The telescope was fitted with a 0.8X reducer/flattener, giving an effective focal length of 571 mm and a field-of-view of 2.2° x 1.5° (about 4 x 3 solar diameters). I designed a full aperture solar filter for this scope which I produced on my 3D printer. My imaging plan was to use an intervalometer to make a shot every 30 seconds from just before first contact to totality, at which point I'd remove the solar filter and activate a bracketing sequence in the camera (using the firmware add-on Magic Lantern) ranging from 1/4000 second to 1 second in nine 1.5 EV steps. After totality I'd replace the solar filter and restart the 30-second intervalometer sequence until after final contact.
In addition, my plan was to shoot a wide field sequence, using my old, but still very serviceable Canon 300D. Because the eclipse occurred near noon at my location, the Sun was very high in the sky, making it tricky to capture both it and something of the surrounding landscape. I determined I could manage this using my Samyang 14 mm wide angle lens on the 300D. For this, too, I designed and 3D printed a custom full aperture solar filter.
I made it up to Casper by mid-morning on Friday, August 18, in time to catch Bob Stencel's interesting talk on observatories near longitude 105° W. I presented a talk on meteor science later that afternoon, and then hit the road to scout a good observing location. Nothing was promising within about 50 miles west of Casper, but to the east I found a county road that headed south from Glenrock, and I ended up in an area of state trust lands overlooking a broad river valley with a nice view of mountains to the south. I returned to Casper for more talks and a reception at Astrocon on Saturday, and headed out before dawn on Sunday for the site I had found the previous day. That morning I could just make out the waning crescent Moon before sunrise.
I set up camp on a plateau via a road that was just barely navigable with my Ford Ranger in low gear 4WD. Technically there is no camping on this land, and a Converse County sheriff deputy did come by, but he saw my telescopic setup and the need for being there, and also that I wasn't doing anything problematic, and didn't make an issue of it. The site was only 270 meters from the eclipse centerline. I spent Sunday morning going through a dress rehearsal of my imaging plan, timing everything according to the actual eclipse events for the next day.
Sunday evening presented some wind and rain, but cleared up in time for me to polar align the iOptron mount. This weather was forecast and not too worrisome. Monday morning dawned clear and cloud-free, with just a little haze from distant western wildfires filling the valley below.
I got the telescope tracking on the Sun and the wide angle camera correctly oriented across the valley. There was nobody else around. My company consisted of deer, coyotes, pronghorn, and some ranging cattle. Not long before the eclipse began, a couple from Crestone, Colorado, parked down below and walked up the hill. And a few cars came along, but were out of sight the entire time. The entire eclipse experience was calm and quiet, unlike down in the towns along the path, where there were groups of hundreds of people together in many places.
As first contact approached at 10:22 am, I started both cameras collecting their 30-second sequences. The partial phase advanced, and I was able to casually observe it through a filter as the cameras automatically went through their operations. Over the next hour, the quality of the light gradually changed, with the landscape assuming a low-contrast appearance, and the temperature dropped from the low 70s to the low 60s. These effects became very obvious in the last 20 minutes before totality. In the last few seconds before 11:43 am, it got dark rapidly, and there was another drop in temperature. Unlike other eclipses I've witnessed, there was no unusual response from any nearby birds or animals. Suddenly, it was dark, with a deep blue sky, Venus and a few stars visible, a 360° twilight on the horizons, and the startling coronal ring around the now hidden Sun. Spectacular! And my practice sessions paid off, allowing me to remove the solar filter from the telescope and start the fast camera sequence while observing all of this. I also walked over to the wide angle camera and collected a few unfiltered shots.
This sequence of three shots is aligned on the Sun, so we can see the shift of the Moon over about two minutes, from just after the start of totality until just before its end. Note the significant amount of coronal activity and the large prominences along the upper edge. This was welcome, but rather unexpected given the generally low activity the Sun is currently experiencing. Indeed, there was a large, active sunspot region in the center of the Sun, and a smaller group very near the limb, which was still visible when only a few percent of the Sun remained uneclipsed.
At the moment totality ended, the famous diamond ring effect was observed, just as a very thin high altitude cloud layer formed, creating a slight haze visible only in images.
A Note on Processing
The eye sees far more dynamic range than a camera can capture in a single exposure. In order to produce an image more representative of the visual appearance, it is necessary to take exposures of different lengths, and digitally combine them using high-dynamic-range (HDR) processing techniques. In this case, I have totality sequences that were exposed at 1/4000, 1/1600, 1/500, 1/200, 1/60, 1/25, 1/8, 1/3 and 1 second. In essence, each image contains perfect exposure conditions for part of the corona, while underexposing or overexposing other areas. The magic of digital image processing allows the properly exposed parts of each image to be combined into a single, high dynamic range final view.
The entire three-hour eclipse, captured by the wide angle camera facing south across the valley below. In this composite, images captured every five minutes are combined. The landscape lighting was captured during totality. I also processed time-lapse videos of both this wide-angle sequence and the telescopic sequence.