The Elusive Eclipse Shadow Bands
Ever sInce my experience with shadow band phenomena during the 1970 Mexican eclipse I have often wondered what effects it might have had on observers of ancient times. Surely they were not informed of a total solar eclipse approaching which would not only cut off their only light during daytime hours, but would also subject them to a last minute crazy shadow band display. It eIther made quick believers out of them or drove them insane.
What are Shadow Bands and What Causes Them?
First of all, shadow bands are a phenomena which occurs for a few minutes before and after totality. They are visible to the naked eye as fast moving light and dark bands along the ground surface, much like those seen on the bottom of a swimming pool when the surface water has been disturbed. Typically the bands are a few centimeters wide and travel at speeds of a few meters per second. Although they are easy to see, they are quite faint and extremely difficult to record accurately. The contrast difference between the shadow bands and the background is normally less than two percent. However, those seen in Mexico were as prominent as the shadows cast by branches of a nearby tree. The direction of band movement is usually tangential to the non-eclipsed crescent of the sun.
Little is known about the causes of shadow bands, but I have heard many different explanations. I feel that the explanatIon given by Wes Kendall, a scIentist from General Electric, has the most merit. His explanatIon is that during normal sunlight, the sun's energy is absorbed by the ground and re-radiated as heat (F i g. I). Th i s causes the air near the ground to be heated in layers. When a solar eclipse begins, there is less light energy emitted resulting in less warming of air near the ground. Meanwhile, the upper layers of cooler air begin to seek the lower level and flow downward and mix with the warmer air layers. This mixing process takes place in a cyclonic pattern and forms crude lenses of varying air density, which in turn cause the diffraction effects in the light passing through this turbulence.
As totality approaches, the light source becomes a thin crescent in the direction of the air flow and the density fluctuations now focus the light and dark bands at a lower level. The air layers supporting this turbulence are many and there will usually be at least one layer which will be at the correct height to focus the shadow bands at the observer's level.
Because of the fact that shadow bands are mysterious and difficult to record for further study, the challenge has stimulated my interest in this area. Since the Mexican eclipse, many experiments have been performed on this subject including the use of applicable equipment at Dalmo Victor Company. The latter involved checking the feasibility of using a sensitive, low-light level television camera (LLLTV) coupled with a video tape recorder to detect and record the shadow bands. Both Jerry Howard and Dick Arnoldusen contributed many hours after work on these experiments and the results were impressive and worthy of further consideration.
To begin with, a set of worst case conditions were established and simulated to check the feasibility of the system mentioned above. They included:
A) Detection sensitivity capable of recording shadow movement under full moonlight conditions.
B) Light and dark bands whose contrast difference was less than 2%.
C) Band velocity of up to 50 ft./sec.
D) Record faint movement of simulated shadow bands on video tape.
A shadow band wheel was constructed by cutting alternate segments out of a 2 foot poster board disk. The disk, which now resembles a windmill, was then attached to a small variable speed, DC motor. Shadow bands simulation was then obtained by rotating the test wheel in front of a "primary" light source and projecting shadows onto a screen. The percent contrast was controlled by positioning a second "background" light source between the wheel and screen, but facing the screen. The intensity of both the "primary" and "background" light sources were adjusted so that the contrast difference between the projected shadow and the screen illumination was less than 2% when measured with a photometer. Moonlight conditions were simulated by reducing the light source illumination to 10-2 foot candles as measured on the photometer. Shadow band velocity was controlled by increasing or decreasing the RPM of the DC motor.
The results of this set of experiments using a LLLTV system showed promise. It recorded simulated shadow bands under moonlighted conditions, contrasts of less than 2% and band velocity of approximately 10 F.P.S. (This was the maximum speed of the DC motor used).
The movements of the shadow bands would now be recorded on video tape and reviewed during replay. However, the information would still be faint. The next step was to enhance the weak contrasts to a higher level (grays become black). This is done by running the video signals through a special threshold detection circuit and amplifying those signals which exceed a pre-determined threshold. The "souped-up" video is now returned to its synched position.
A replay at this stage would provide a display of shadow band information which would be easy to watch and study.
The availability, support and transportation required to operate the LLLTV system at the Canadian eclipse was beyond the scope of our month long expedition. Instead, we developed a simpler photographic technique involving hi-reflectivity projection screens and sensitive films for our first try at shadow band detection. These are described in the following paragraphs.
Shadow Band Experiments
For the eclipse (July 10, 1972), the author selected two types of 35mm film to be tried alternately on two different shadow band screens. The films selected were required to be fast and sensitive to a blue-green spectral response, while the screens were to provide high reflectivity characteristics and enhance contrast and detail.
Fast Kodak 2475 recording film (ASA 4000) was loaded in a Cannon FT (f/l.8) camera and was to be operated by Barbara Daugherty. At the same time, Vern Daugherty was to operate a Kodak Retina C-III camera (f/2.0) containing blue-green sensitive 2484 film (ASA 3400). Both operators were located within a 6 foot diameter circle midway between the two screens. During shadow band activity, they would face opposite screens and a set of three exposures were to be made using different shutter speeds of 500th, 250th and 125th of a second. They then would rotate their positions and repeat the series on the opposite screens. This same practice was to continue for the duration of shadow band activity both before and after totality.
The screens used consisted of a beaded material for one and frosted mylar for the other. The beaded screen, which is normally used for slide projections, was to provide high reflectivity characteristics when photographed from a position perpendicular to the plane of the screen. The frosted mylar screen was to be used for rear projection photography and was expected to enhance the detail and possibly contrast. The latter screen was used to function much like the ground glass on a camera back.
Shadow Band Screen Construction
Since transportation of bulky equipment to the eclipse site presented no problems, it was feasible to construct two large shadow band screens. Once assembled, the screens measured 52" x 56". The frames were made of wood and bolted to plywood corner braces.
The screen material for each screen was secured to the frame using lacing cord, wood strips and nails. Each screen also contained a set of cross markers calibrated in inches, which were to be used to measure the recorded shadow band characteristics, including band widths, velocity and direction of travel.
Each screen was also outfitted with a Gnomon. A length of wood doweling which protruded outwards at right angles to the screen was used for this purpose. It would cast a shadow if not directly in line with the sun. This provided a neat and quick way of aligning the screens just prior to shadow band activity.
A requirement for the frosted mylar screen is that it be located in an elevated position with its plane perpendicular to a line between the sun and the photographer during shadow band activity. This was accomplished mounting the screen on top of an automobile using four suction cups as footing. The same setup withstood 25 miles per hour winds during pre-expedition tests in San Jose.
Unfortunately cloud interference (at Cap Chat in Quebec, Canada) destroyed any hope for shadow band photographs. However, chances for the African eclipse later this year (1972) are very good..
This article originally appeared in the magazine Celestial Observer, April-June 1973, Vol. 4 No. 2.