Paul Maley’s Eclipse Edge Observation

For most people who travel to see solar eclipses, the intent is either to watch the event or to document it in some way. However, there is more to it than that and our activities focus on observations of both annular and total solar eclipses from the edge of the path of annularity/totality. For total eclipses this is contrary to normal thinking but for reasons you will see below, there is a useful science component to our expeditions. Get ready for a huge paradigm shift.

You can participate in science goals of RING OF FIRE EXPEDITIONS, the International Occultation Timing Association, and the NASA Johnson Space Center Astronomical Society’s public outreach initiatives at solar eclipses by observing at the edges (i.e. the north or south limits) of the eclipse path. There are two regions, one at each edge of the path that are also referred to as “graze zones”. The true edges of the path of totality are the dividing lines between seeing and not seeing a total eclipse. The grazes zones are several kilometers inside each edge and from here you can see phenomena whose durations are many times over what you can observe at the centerline.

Our first eclipse limit observation was made from Acapulco, Mexico in 1973. Since that time, our methods have changed due to the advent of technology; but the overall objective is to attempt to collect data that could be useful in ascertaining expansion and contraction of the sun between eclipses. Because the sun is not a solid object, it “breathes” with periodicities of varying nature. It is possible for amateur astronomers to use video cameras to record a phenomenon at solar eclipses that is both rare, beautiful and dynamic. This is called Baily’s Beads (documented by Sir Francis Baily after an annular eclipse on May 15, 1836 even though prior observations were made by Edmund Halley on April 22, 1715 at an eclipse in the UK. They were also seen by Colin Maclaurin from Edinburgh, Scotland during the annular eclipse of February 18, 1737, James Ferguson who observed an annular eclipse from Liverpool UK on April 1, 1764, and by Harvard College professor Rev. Samuel Williams from Penobscot Bay, Maine (east side of Long island)  on  October 27, 1780  from just outside the path of totality.  This latter expedition was even more fortunate in that this area was held by the British during the American Revolutionary War and special permission was needed just to enter the eclipse zone.

ecllipse edge observation

The observer location was the red marker on the left side, the left blue line is the southern edge and the red line is the centerline.

The map above (source:  is a reconstruction of the eclipse path. From this it appears that the Bead observations were made far outside the path of annularity.


eclipse edge observation

Instruments used by the Harvard College team at the eclipse of 1780 to watch Baily’s Beads

On observing the annular eclipse of May 15, 1836, at Inch Bonney near Jedburgh in Scotland, Baily described the appearance of the broken ring of sunlight as ‘a row of lucid points, like a string of bright beads, irregular in size and distance from each other’ which ‘suddenly formed around that part of the circumference of the moon that was about to enter, or which might be considered as having just entered, on the Sun’s disc’.

The beads are best observable at the limits of the eclipse path, not at the center line. For example, a total eclipse can be visible over a long track of thousands of miles as the progression of the moon’s shadow onto the earth sweeps across the earth’s surface. The long track has a diameter of e.g., 150 km. At the north and south edges of the path totality lasts just an instant. Just outside the edge the eclipse becomes 99.9% total and less as one moves farther away from the edge.


The Baily’s Beads are numerous, tiny points of light that pop up as the irregular shape of the moon moves tangentially across the solar disc. They were probably first noticed in 1715 when Sir Edmund Halley saw them at a solar eclipse on April 22 of that year. At the time of central eclipse, bits of sunlight peek in and out of the darkened lunar valleys at the polar regions. For those stationed near the projection of both lunar poles onto the earth, a prolonged enjoyment of the beads can be had. Baily’s Beads are only seen for a few seconds at the centerline; but from inside the graze zone the beads can be seen for perhaps a full minute or more depending on the length of the total eclipse.

If you are ‘in the zone’ you can also see an extended view of the prominences. The reddish ejections of solar material are generally situated away from the polar regions. Therefore you see prominences on one side of the sun from the center line during the first half of totality, while others appear on the opposing side during the second half of totality. From the edge, because the tangential motion of the moon slows everything down, the view of the prominences is prolonged. This slowed motion also increases the duration of the shadow bands that are seen as faint rapidly moving bands on the ground prior to and following totality. Watching for these bands does not take away from any totality viewing. In addition, the remarkable diamond rings that are typically seen at 2nd and 3rd contacts are also visible far longer from the edge.

But it is the Baily’s Beads that grab the majority of attention as they form, merge, dissipate and shift along the edge of one limb of the moon. The detail can be incredible and depend on whether you are at the south edge where the projection of the deep valleys tends to show large beads, or at the north edge which results in much finer, tinier beads that reflect the flat regions at that lunar pole.

Totality is much reduced in the graze zone, by as much as 70%, but still totality is visible and is just as mind-blowing as you can imagine. Those who want to contribute to the science and still see all these amazing phenomena can be witness to the interactive light show that is a solar eclipse.

Not to be outdone, annular eclipses, which do not result in the sky getting dark are just as productive in the Baily’s Beads department. At the north graze zone, large beads are seen (just the opposite from a total solar eclipse) and from the south graze zone, tinier beads will be detected. If you have ever tried to capture 2nd and 3rd contacts of an annular eclipse from the centerline, you probably have been very disappointed. Those contacts are ill defined and while beads can be seen, they are generally so short-lived and hard to isolate that photos do not do them justice. But just move up into the graze zone and you will have an entirely wonderful surprise and still be contributing to the solar radius definition program of IOTA. The beads are just aas prolonged as during a total eclipse and even though the central eclipse does not take on a Bull’s Eye effect as seen at the center, central eclipse is still just as impressive!

The following image sequence was taken from the total eclipse centerline at Batman, Turkey in August 1999; you can see a series of quick images in succession showing the short period in which the beads are picked up at the centerline. Only the second from the right image clearly captures some of the bead action. They were gone within just a mere 2 seconds! These images are courtesy of Dan McGlaun (copyright by Dan McGlaun 1999). Furthermore, they provide adequate rationale for not going to the centerline to record data on the Baily’s Beads.

ecllipse edge observation

As we just mentioned, the moon appears to move the slowest at the eclipse limits, and so the appearance of the beads can occur over a minute or more as opposed to a couple of seconds at the center line. At the edge of the eclipse path, the sizes of the beads are proportional to the depth of the lunar valleys. The larger beads correspond areas as deep as 1-2 km in places while the tinier beads are from features perhaps between 50 and 200 m deep. The idea is to link the recorded times of formation and dissipation of each principal bead with the lunar feature corresponding to it.

To see and capture the beads is one of the ambitious educational and scientific aims of the International Occultation Timing Association (IOTA) which collects and analyzes the data. The application is described by D. Dunham, A. Fiala, and S. Sofia in their work on assessing solar luminosity variations affecting global climate from the analysis of solar eclipse observations. They believe that there is a link between brightness variations of the solar disc and terrestrial climate changes. To establish the variation in brightness, the idea is to model the exact size of the sun at as many eclipses as possible; then determine to what extent global temperatures might be driven by variations in the sun’s size.

But other results presented in 2005 are available and those are recent solar radius determinations from solar eclipse observations made by IOTA expedition participants. These can be found at


What equipment works best to gather this valuable data? As of this writing, a telescope such as a Celestron 5, Meade ETX-type or similar scopes mounted onto a motor-driven portable mount which can be used in either southern or northern hemisphere is employed to image the sun. A small 12-volt CCD camera available inexpensively from an outlet such as  is then coupled to the telescope; we presently recommend the PC-23C. Different cameras/camcorers have different responses as do solar filter materials. One of the challenges as technology advances is to determine the effects of system components on the resulting data.

The output of the camera is fed into the video input of a camcorder such as the Sony TRV-35 which has a fold out viewing monitor. One needs to ensure that if you buy a camcorder, that it has the capability to record video input from an external source. Many DVD camcorders do not permit such capability. A neutral density number 5 solar filter is placed in front of the telescope aperture to protect the camera from being damaged by the intense solar energy. Even as the time of total eclipse draws near, the small elements of sun peeking through lunar valleys are easily recorded. A “speaking clock” (battery powered short wave receiver which can receive international time signals) is typically used to establish a uniform source of time which must be simultaneously recorded on an audio channel along with the video of the eclipse.

Here is an example of Richard Nugent’s portable equipment setup in which a Meade 2045D telescope with a CCD camera attached, a short wave radio, and a camcorder are all interconnected to produce an image of the sun. Such a setup is typical of small Schmidt-Cassegrain telescopes that can be obtained currently.

eclipse edge observation

The telescope is polar aligned and then aimed at the sun. The tracking motor automatically follows the sun and keeps it in the center of the field of view. It is not advisable to use a telescope and video system mounted on a tripod. Auto-tracking of the sun is a must! It is desired to keep only about 1/3 the sun’s diameter in the field of view— not the entire solar disc. So, any telescope or telephoto combination that either has too large or too small a field of view would not be very usable. Wind is another factor that can vibrate the telescope. Care must be taken to 1) protect the equipment from wind and wind-borne particles at the site; 2) to be sure the solar filter remains in place during the entire eclipse process; 3) that you are aware of the migration of the beads along the cusp from 2nd to 3rd contacts so that you do not lose the beads between contacts.

During the eclipse process, the moon slowly begins to cover the sun. When about 90% covered, the point where the lunar limb makes contact with the sun will begin to develop a small bulge on the tip. This is the first sign of a Baily’s Bead. One or more may slowly develop or recede as long as 20 minutes before or after central eclipse. The real action happens within several minutes before 2nd contact when one or more beads begin are clearly visible. Our sites are typically located 1-3km inside the path of the eclipse from the edge. As totality begins (or central annularity for an annular eclipse), the beads become quite prominent and are the most dominant and amazing spectacle in the VCR monitor. But for annular eclipses, the sun is not completely eclipsed and the corona is not visible. Hence, you do not give up much at an annular eclipse to watch from the limit line. To gather data for a complete solar diameter analysis, observers must watch from both the northern and southern edges of the eclipse path.

eclipse edge observation

In this series of still video frames captured by R. Nugent from Curacao on February 26, 1998 at the southern limit of that total solar eclipse, you can see large Baily’s Bead features caused by deep valleys in the polar region. Examine the evolution of beads as they move from panels 1-6 (from top to bottom). In panel 1, a small bright bead (bead 1) emerges on the left side and maintains the same level of brightness through panel 6. Each panel is separated from the next by from 4 to 10 seconds. This means that just in the space of these 7 images span a period of about 25 seconds, and that was only part of the eclipse light show! In panel 2, 4 more beads (beads 2-5) appear from left to right of bead 1. In panel 3 just to the right of bead 5 are two dim beginnings of beads 6 and 7 that are distinct in panel 4. Then in panel 5 another bead appears in between beads 5 and 6. Beads 5 and 7 have begun to expand laterally. Finally in panel 6 there is a complete merging of many beads. Note that the dark areas in between the beads are caused by lunar mountains blocking the sun’s light.

eclipse edge observation

In the above image captured from annular eclipse video taken in April 1995 by P. Maley from Puinahua, Peru, you can see the very tiny bead features at the north lunar pole caused by the flat terrain there. At this annular eclipse, we observed from the southern edge. At annular eclipses north lunar pole features are observed at the south limit and vice versa. This image was taken through a Celestron 5 and Thousand Oaks Type II solar filter.

Data is usually recorded on standard 8mm or Hi-8 video tape beginning at least 5 minutes before 2nd contact and ending 5 minutes or more after 2nd contact. While the VCR monitor should always be watched in case of movement of the beads out of the field of view, if there is no wind and the telescope and video camera are tracking properly, one might take time to watch the corona and other eclipse features during totality.

Once the video recording is terminated, the video should be copied, and the copy sent to IOTA for analysis while you retain the original tape. At present, the analysis can take years to complete since there are many eclipses for which data has been collected that are still in the queue waiting for reduction. Because IOTA is a volunteer organization, the reduction process is tedious and slow and awaits funding to accelerate the results. Eclipse processes in the distant past are also being investigated for certain eclipses in the 19th century at which timings were made in the United States.

The photo below shows an example of Baily’s Beads during an annular-total solar eclipse in 1987 led by P. Maley and taken in Gabon (courtesy of Huguette Guertin). Note the fine detail visible all around the solar disc. This was a one- second total eclipse and hence the beads were actually visible all about the sun migrating from one limb to the other. The image below is a black and white negative enhanced to show detail.

eclipse edge observation

In order to assess the observed beads versus the predicted beads, Dr. Alan Fiala of the US Naval Observatory developed a software program which generates simulations of the lunar terrain at central eclipse. Notice the similarity between the shape of the Guertin photo above the output of the Bead simulator program which models the appearance over time. The simulator utput can be then used to predict the correlation of video images with actual lunar features.

eclipse edge observation

These use data that originate from photographic measurements compiled by C.B.Watts. Watts developed the “Marginal Zone of the Moon” which is a compilation of 1800 small charts based on measures of photographs from 1927-1956. From this, a reference datum was developed that is used as a standard for defining the topography of the mountains at the polar regions of the moon. Limb corrections have been enhanced by thousands of timings (made by IOTA members) of stars as they pass behind these same mountains over the intervening decades since the Watts charts were published in 1963.

eclipse edge observation

The figure above shows a plot of the moon’s mean limb (smooth curve) and two jagged curves illustrating an exaggerated view of the south pole where the two components of a double star (separated north-south) are projected to occult these features. Note the large depth of the features. The plots of predicted lunar features are for those at the south pole. The smooth curve is the mean limb of the moon if it were a perfect sphere.

eclipse edge observation

In this figure, the smooth curve once again is the mean limb of the moon and the variable curve almost overlaying the smooth one, shows the predicted heights of lunar features at the North Pole. Note the difference in the low terrain height at the North Pole from that at the South Pole.

Expeditions we have organized often take us to remote places. This photo shows our site near Kwikila, Papua New Guinea in November 1984 when we chased after a total solar eclipse. Though there were only four of us at the site, we were quickly joined by locals who saw our equipment and were curious as to what we were trying to do.

eclipse edge observation


Normally for most inexperienced eclipse watchers, the safest way to view an eclipse is by projection. Valuable information can be obtained by this method also since it is useful to coordinate different methods at the same eclipse viewing site. The idea is to have one observer project the sun’s image onto a stable white surface (e.g. a white piece of cardboard) using a telescope or other lens that is mounted and preferably tracking on the sun. The system must be protected from the effects of wind which can create unwanted vibrations and result in out of focus Baily’s Bead images. A camcorder can then be used to image the projected sun’s disc with emphasis on the edge with enough magnification to clearly show the individual Baily’s Beads as they form. This can be a bit challenging since the camcorder must be securely mounted and stable, focused on the projected image so that the Baily’s Beads can be clearly recorded. The observer must pay close attention to insure that the focus is maintained throughout the recording process; also the observer must record the GMT time signals on the audio channel of the camcorder or preferably have a time inserter that shows a running digital clock in the background of the video tape.

The above image is an example of early rear screen projection using a elevation azimuth mount which was manually tracked. We discourage manual tracking as it is too labor-intensive. Chuck Herold with Don Haverstock behind him teamed to attempt this effort in 1984 from Papua New Guinea during the total solar eclipse there. However, the photo does show a good example of a video camera pointing in close proximity to a very enlarged projected image of the sun. The frame holding the transparent screen is quite stable and is shaded by two cardboard cutouts. Sitting behind the base of the telescope is a short wave time signal receiver. The biggest drawbacks in this setup are that the single observer must concentrate on the following:

a. moving the scope in elevation and azimuth

b. ensuring the relationship between the output of the sun image and its position on the screen

c. keeping the camera perpendicular to the projected image

d. listening to the time signals to be sure they do not drift

e. recording comments and time signals at the proper level

In addition to the above there is also the difficulty in transporting an 8-inch telescope half way around the world. With two or more people and a tracking mount, all of these problems can be handled better; but one observer employing a direct imaging system can handle all aspects associated with this process. We present this illustration to provide guidance on how you might plan to organize your own equipment.

The next image shows Gary Nealis with a Celestron 5 using a forward projection system. While he used a tracking mount, he also found that being so close to the equator, it was not possible to align the mount to the pole and fell back to an alt-az orientation which also required manual tracking. In the image you can also see the attempt to co-align a 35mm camera which was never perpendicular to the screen. Again, this was a problem with the projection system in trying to do too many things at the same time with too much hardware in close proximity.

It is important that any observer who would like such projection data to be used by IOTA should contact us well ahead of time so that we can coordinate his/her equipment ideas and capabilities along with others who might be planning to observe the same eclipse at the edge. IOTA can provide advice on the proper techniques and give source material on types of equipment as technology changes.

The following images show the progression of beads around the periphery of the sun using the projection method. They were taken by Paul Maley from a cemetery in Atlanta, Georgia in 1984 during an annular eclipse. The telescope used was a Celestron 8 and the projection material was a white screen. This was not an ordinary annular eclipse. It was annular-total meaning that the diameter of the moon was nearly the same as the diameter of the sun. It was so close that Baily’s Beads was seen to completely surround the sun at central eclipse time.

 eclipse edge observation

eclipse edge observation


Time signals can be received with a shortwave battery powered radio, preferably one that is digitally tuned. Favored frequencies to choose are 5.0, 10.0, 15.0 and 20.0 MHz though there are some others less likely to be received. The key thing to listen for is a persistent tick or tone every second. Some stations may have announcements at the start of each minute, while others do not but may generate a unique tone. Still others may have voice announcements in English or another language. Popular time stations are WWV and WWVH which can also be received outside the USA. Reception of good signals is strongly dependent on solar activity and how well the time signals propagate through the earth’s atmosphere. Signals may be received better at night than the day or vice versa. Timing precision required by IOTA is 0.5 seconds in time or better.

Observers should consider investing in a digital quartz watch which can be preset to a local shortwave time signal station before departing for the eclipse expedition. This is not a substitute for shortwave time signals but should only be considered a backup. If time signals are not received at the eclipse site and there is a way to annotate an audio tone onto a videotape of Baily’s Beads, one can insert verbal markers calling out ‘mark’ at the 0th second of each minute and also noting the GMT in hours and minutes on the tape. The watch can then be rechecked upon returning home to determine how much it has drifted per day. This drift can be applied to the data which is reported after the expedition. Time inserters are now a popular method of directly recording an accurate digital time signal superimposed on the Baily’s Beads video. The Kiwi OSD model produced by PFD systems ( are small compact and lightweight devices that are driven from either a GPS source as part of the Kiwi product lne or can display a simple running clock that can be calibrated with a short wave time signal reference.

The Kiwi OSD with a GPS receiver (small round disk) and associated cabling is shown here.

eclipse edge observation

The following is a frame courtesy of Brian Loader showing a transit of Mercury on the sun with a Kiwi OSD display in the video frame. The left-most set of numbers are GMT hours, minutes, seconds.

eclipse edge observation


Locating the exact sites is vital and is accomplished by our use of Global Positioning System receivers. These are now available at comparatively low prices (under USD$100 from sporting goods or on-line stores), are simple to operate and provide very accurate information on latitude, longitude, and elevation. These receivers will pick up as many as 12 satellites at a time. It only takes the reception of 4 GPS satellites to yield a valid 3-dimensional position. The observer turns on the receiver, and after some period of time (generally 3-5 minutes) the output coordinates will stabilize. It is then that the observer can log the site coordinates. The target goal is to obtain a position that is accurate to within 30m in latitude/longitude and 20m in elevation. Always set the datum to WGS84 and navigate using kilometers as a standard. In the absence of GPS, accurate topographic maps are required. Those obtained from a national mapping agency are generally on par with those issued, e.g. by the U.S Geodetic Survey. Outside the USA, such maps may be hard to find or severely outdated. Web sites like may also have good references for specific site locations especially in the USA. The web tool Google Earth is also useful for obtaining latitude, longitude and altitude in the USA, but as of 2006 this tool should not be relied on for site measurements in other countries. Outside the USA, a GPS is the main positional reference tool that is recommended.

The NASA solar eclipse bulletin (superseded by the RASC Handbook) provides a detailed listing of latitude and longitude coordinates. The observer can take the centerline coordinates and then by using the ‘off course’ or ‘cross track’ feature inherent in the GPS receiver, navigate from the centerline to the proper distance away that defines the graze zone. Alternatively, the Handbook also provides the coordinates of those graze zones and the same method can be used to navigate directly.



  • 1.telescope 3-inches (75mm) or larger in aperture on a motor driven platform
  • 2.finder scope boresighted to the main telescope
  • 3.portable shortwave time signal receiver
  • 4.GPS receiver
  • 5.camcorder with flip screen
  • 6.batteries for all hardware fully charged
  • 7.required connectors and cables
  • 8.time inserter (optional)
  • 9.ND5 solar filter
  • watch calibrated to time signals
  • 11.CCD video camera that attaches to the prime focus of the telescope


  • 1.items A1, A3, A4, A5, A6, A7, A10
  • 2.projection screen
  • 3.tape recorder
  • 4.eyepieces for projection of sun’s image
  • 5.shading material to keep direct sun off the screen

Note: item A9 is not required since you want to be able to fully project the sun’s image unfiltered. 


In 1987, Pasachoff and Nelson (Jay M. Pasachoff and Brant O. Nelson, “Timing of the 1984 Total Solar Eclipse and the Size of the Sun”, Solar Physics, Vol. 108, p. 191-194, 1987) cast doubt on solar radius values determined from solar eclipse observations based on the size of prediction errors. They missed the point that the prediction errors are removed by the analysis process, which solves for corrections to the lunar ephemeris relative to the solar ephemeris in addition to the solar radius. For the eclipses before 1991, including the 1984 eclipse observed by the authors, Newcomb’s theory of the motion of the Sun was used for the predictions. That theory, over a century old, is known to be in error by nearly 1″ at current epochs. Recent observations show that the DE200 ephemeris that was used for predicting eclipses after 1985 is in error by less than 0.2″, the determination being limited by the errors of the Watts lunar profile data. But the authors did not recognize that both ephemeris and profile errors are largely removed by obtaining observations at both the northern and southern limits of total and annular eclipses. The difference of the northern and southern limit residuals removes the ephemeris error, leaving the solar radius residual. Then, the profile error can be removed for the solar radius residual from one eclipse to another by using contacts (or Baily bead timings) defined by the same lunar valley bottoms, which can often be arranged due to the fact that the latitude libration is always near zero during an eclipse, so that the same lunar features define the profile in the lunar polar regions for each eclipse.


eclipse edge observation

 Lunar eclipse May 15, 2003 (by Loyd Overcash) with star about to be occulted

Amateur astronomers can also participate in expeditions to refine the polar diameter of the moon. Such opportunities take place when a star is eclipsed as seen from earth during a total eclipse of the moon. During the total phase of a lunar eclipse, it takes a star brighter than 6th magnitude to be bright enough to be easily observed so as to clearly time it as it passes behind lunar peaks. We first attempted this at Dagupan City, Philippines in 1982 and timed the eclipse of a bright star at the northern edge of the moon from the beach exactly 37 years to the day after Gen. Douglas MacArthur landed there to liberate the Philippines in 1945. A complementary expedition in Australia by others at the southern limit of the eclipse failed due to weather problems. Three years later, at a total lunar eclipse in 1985 a star named Alpha Libra 2 was observed to be eclipsed by the north polar features of the moon from El Geteina and Hag Abdullah, Sudan (coordinated by P. Maley and D. Dunham) and simultaneously by mountains at the south pole of the moon by observers in South Africa. The figures below show preliminary plots of the predicted and observed features at both limits based on the eclipse timings.

eclipse edge observation

 eclipse edge observation

Another IOTA expedition in November 1993 recorded similar occultations of a star during a total lunar eclipse; this time, northern limit data was collected by a team led by Doug Hube in Canada under very cold weather conditions. Southern limit video was achieved by P. Maley from a desert site north of La Paz, Baja California, Mexico. The video in Baja from one station was completely successful and the site was resurveyed in May 2000 using GPS. It is expected that the results of that expedition will be folded into an updated polar diameter of the moon–the first such measurement by amateur astronomers.

Ring of Fire Expeditions (ROFE) is the longest consecutive astronomical tour organization in the United States. ROFE specializes in astro-tourism since 1970 with expeditions organized and led by Paul D. Maley of the NASA Johnson Space Center Astronomical Society and arranged by Future Travel in Houston, Texas USA. These include tours to observe such events as Halley’s Comet, the Leonid meteor shower, transit of Venus, spacecraft reentries, solar eclipses, grazing occultations, and occultation’s of stars by minor planets.

We are a public outreach effort of the NASA Johnson Space Center Astronomical Society led by amateur astronomers and welcome all persons who are interested in astronomy and the natural sciences. You do not need to have a science background or any prior experience to join us! Contact us to set up your perfect astronomical tour and/or cruise today!


Back to Top