November 1994 (Chile) Expedition Final

The global eclipse-track maps published by the U.S. Naval Observatory and by NASA indicated that the November 3, 1994 eclipse would pass over the Pacific Ocean, across the South American continent from Peru to Argentina, then across the southern Atlantic Ocean. The best land-based viewing locations seemed to lie near the east coast of South America in southern Brazil, where the Sun would be high in the sky and the eclipse would last longest.

Monthly rainfall charts for southern Brazil suggested that November might be acceptable in terms of precipitation probability at eclipse time. So we began detailed planning for the 1994 expedition in August, 1991. While en route to a meeting of the International Astronomical Union meeting in Buenos Aires, Tom Van Flandern (hereafter, TVF) stopped off in Rio de Janeiro and at Florianopolis, Brazil to scout for suitable sites and points of interest for an expedition to South America. Both observing sites and points of interest were located in abundance.

However, upon return to the United States, we checked satellite cloud cover information. This showed unacceptably low cloud cover probabilities for the time of year and time of day when the eclipse would happen. Unlike in the United States, where rainfall often happens in short, heavy bursts, much of Brazil has a fine mist falling from clouds much of the time. This is what keeps Brazil's jungles, and indeed rain forests, lush and green. So some places may experience rain on, say, 80% of the days of a month, yet have comparable total rainfall to a U.S. city where it rains on, say, just 30% of the same days.

Needless to say, this finding was very disappointing, since trips to South America are too expensive to justify a less than 50% probability of seeing the total eclipse that is the prime goal of the travel. Indeed, we generally set 2/3 as the minimum probability of clear skies at the prime site for all Eclipse Edge expeditions.

So TVF made a second scouting trip in December, 1992, with the intent of checking potential sites in Bolivia, Peru, and Chile. Sites near Iguazu Falls in Paraguay had to be excluded because of even poorer weather prospects. And it quickly became apparent that Peru presented a unique set of problems that called for its exclusion. A cholera epidemic was still raging at that time; and a traveler's advisory was in effect for U.S. citizens because of terrorist activity directed at them in Peru. And the best sites, near Ariquipa, very also at very high altitude, causing problems similar to the Bolivian sites (see below).

This narrowed the possibilities to the Bolivian Altiplano, where clear sky prospects were best; and to the Atacama desert in northern Chile, which had the next highest clear sky probability. But the Altiplano is a plateau at an altitude of about 4000 meters (13,000 feet) above sea level. It is not well known by people who have never traveled to such altitudes how unpleasant, and even dangerous, such travel can be. The probability of sea level dwellers getting "altitude sickness" rises from a few percent at 3000 meters to about 30% at 4000 meters, and nearly all of them have some symptoms for their first few days at 5000 meters. Soft tissues swell while adjusting to the reduced air pressure; and this includes the brain, which can then produce headaches and nausea. There are few things in life so miserable as to make one forget about a total solar eclipse; but a bad case of altitude sickness is one of them.

Since too many participants in a large group would be likely to find adjusting to the altitude of the Altiplano an unpleasant experience, sites in northern Chile were adopted in preference to those in Bolivia. The Atacama desert, while amazingly barren, offered its own attractions, not the least of which was the opportunity to get above most coastal cloudiness, and an over 80% probability of a visible Sun near eclipse time.

TVF made a third scouting trip in November, 1993, exactly one year in advance of the eclipse; and Paula Malcolm made a similar trip to secure local arrangements in April, 1994. We contracted with Destination Management, Chile (DMC) of Santiago to be our agent and local representative in Chile. Suitable primary and backup sites in the Atacama desert were selected, night-sky viewing sites were planned, other points of interest for those visiting South America were checked out, and preliminary choices of airlines and hotels were made. The stage was set.

Logistics

We contracted with Ladeco Chilean Airlines to carry not less than 100 passengers from Miami to Santiago to Arica (or Iquique) and return. In addition, we made backup arrangements with Aerolineas Argentina for Miami-Buenos Aires-Santiago flights. While the latter were more expensive, they provided an opportunity for participants to explore South America more deeply at modest additional cost.

As experienced eclipse expedition planners, we had some idea how heavy demand for accommodations would become as the eclipse neared. Many service-providers in Chile did not seem to realize this, despite our warnings. In a continuing struggle to adapt to relentless demand, Ladeco Airlines first divided our group (then still of unknown final size) between two flights on successive days, and later added a third Miami-Santiago flight-day. The domestic Santiago-Iquique and Santiago-Arica flights underwent numerous modifications, with schedules changing even at the last minute, and new flight crews pressed into service. In the end, we were unable to persuade the airline to keep its original, more favorable scheduling as it continued to shift around planes and flight crews in its fleet to accommodate the maximum number of persons and groups.

For hotels, we chose base sites outside the path of totality to avoid the usual chaos that sets in six months before an eclipse. It has become a typical situation that some hotels will succumb to the overwhelming demand for accommodations, and throw out reservations previously made, making their rooms available to the highest bidder. It is also typical that tour operators will overbook spaces, then be unable to sell all of them at the high prices they must charge to cover unpredictable hotel price increases. So about 4-8 weeks before the eclipse, overbookings get canceled, prices collapse, and in the end, some hotel space remains empty -- to the dismay of all. Moreover, local service providers in the path of totality often have "eclipse specials" where they offer their normal fare at up to ten times normal prices for eclipse visitors.

We chose Santiago, 1800 km south of the eclipse path, and Iquique, on the coast 300 km south of the path, for our bases of operations. We might have used Iquique as a base for the entire group; but even 300 km away from the eclipse, all hotel accommodations were booked by a year in advance, prices were unstable, and reservations unreliable. Santiago, however, appeared to be too far south to be vulnerable to such pressures.

Unfortunately, the week nearest All Saints Day (Nov. 1) is festival time in Santiago, a heavy period for the travel industry in any normal year. Although no official figures are available yet, it appears that more people entered South America during the week immediately preceding the eclipse than in any previous week in history. The number of people worldwide who will travel anywhere to see another total eclipse just keeps increasing as knowledge of the beauty and wonder of eclipses continues to spread.

Because of this, Eclipse Edge Expeditions was not exempted from "eclipse hotel stress" for this expedition, despite our careful advance planning. Our original choice of hotel was unwilling to make the price guarantees and other accommodations we needed. The Tupahue Hotel was eager to welcome our group, but much less eager to welcome the Iquique contingent for one night following the eclipse because they could already sell all their rooms for the remainder of the week.

So originally (as described in Bulletin #4), we contracted with the Manquehue Hotel in Santiago for expedition members the Tupahue could not accommodate. The "guaranteed" contract was secured with prepayment in full (not just deposit). Nonetheless, seven weeks before the eclipse, the Manquehue tore up our contract and canceled all our reservations there. This followed shortly after a dispute had arisen with the Tupahue management that threatened to disrupt our arrangements there as well. By this point no other hotel in Santiago had enough space left to accommodate more than very small groups, and all were charging "eclipse rates".

Most expedition members were unaware of these stresses to the organizers. In all, over 100 faxes passed back and forth between the U.S. and Chile during the year before the eclipse regarding details of arrangements. During the peak crisis, when the hotel reservations for the entire group seemed in jeopardy, we spent one worried day and night not knowing if the expedition would have hotels at all, or if our group might have to spend the week camping in the desert.

The resolution of the crisis found us dividing the group up among five different hotels in Santiago for the two nights after the eclipse. DMC picked up the extra cost of doing so, for both hotel charges and transportation costs to the Tupahue and to the airport. Eclipse Edge Expeditions covered all other cost increases. We knew how unhappy expedition members would have been if we had notified you of a "price increase" just weeks before the eclipse! But as Chilean tour guide and host Fabio Gaete explained to passengers on his buses, this assignment (hosting the Eclipse Edge Expedition and getting to see the eclipse with the astronomers) was considered the most sought-after and envied job in the entire travel industry in Chile.

The Expedition Begins

On Friday, October 28, Eclipse Edge Hosts and about two dozen early travelers flew to Santiago on the first of three Ladeco flights carrying most of our group. Some slept on the overnight flight, while others were too excited by their adventure to sleep. As dawn approached, spectacular views of the Andes mountains, occasionally snow-capped, could be seen from the plane's windows. A colorful sunrise brought daylight again, making the terrain below appear even more striking.

Our scheduled Saturday morning arrival in Santiago was for 7:25 am. However, due to heavy (and unusual) fog at the airport, all international flights were diverted to the city of Mendoza in Argentina until weather conditions improved. Our Ladeco flight was then one of the first to actually land at Santiago that morning, shortly before 11 am.

As each of the expedition's Ladeco flights arrived on Saturday, Sunday, and Monday mornings, the new arrivals were met at the airport, then given a quick tour of the city of Santiago. This included an overview of the city from atop one of its many hills that emerge from otherwise flat terrain. In all, 144 people joined our expedition.

At this juncture, the Sunday morning arrivals divided into two further subgroups. One stayed in Santiago at the Tupahue Hotel. The other, consisting of about 60 people, returned to the airport for the journey north to Iquique, a coastal resort city nearer to the eclipse path.

The Santiago daytime program consisted of optional side trips to many tourist attractions in the area. These included Vina Concha y Toro (a vineyard), the famous Maipo Canyon, and the cities of Valpariso, Vina Del Mar, Cartagena, Isla Negra, and El Quisco. Each contained sights of both beauty and historic interest. Other expeditioners chose to relax at the hotel's poolside, or to tour the city’s sights on their own. A pleasant day was spent by many at the Santiago zoo, then taking the cable car to the mountain top for a magnificent view. Those based at Iquique had the option of touring the desert cities of Pica, Pozo Almonte, La Tirana, Santa Laura, and Cerro Unitas; or the well-known Geoglifos of Pintados (large Indian drawings on the mountainside).

But the astronomy program is what most had come for; and few were disappointed. TVF gave his multimedia presentation, "The State of the Universe", to the early Santiago arrivals on Sunday evening, and to the later arrivals on Friday evening; and to the Iquique group on Tuesday evening. The topics included the Comet Shoemaker-Levy-9 comet impact on Jupiter, and the long-term threat to the Earth's environment from similar impacts; the 25th anniversary of the Apollo 13 lunar landing; the discovery of a satellite of Asteroid Ida, and its implications for the exploded planet hypothesis; and some of the latest findings from the Hubble Space Telescope, and their implications for cosmology. These were illustrated with computer-controlled multimedia sound, graphics, and animations, plus a video of the much-praised short film "Powers of Ten" to illustrate the scale dimension of the universe -- the principal way the author's own cosmology differs from the mainstream models.

Co-host astronomer Wayne Warren spoke to the Iquique group on Sunday evening, and to the combined Santiago groups on Tuesday evening. His topics included "The Stellar Content in the Constellation Orion", the subject of his current research; tips on southern sky viewing; plus beautiful slides from the Hubble Space Telescope and from the previous Eclipse Edge expedition to Puerto Vallarta in 1991.

On Monday evening, the combined Santiago groups took buses to the Ferallones ski resort atop a mountain in the general area, to get a good dark-sky view of the southern stars. Overcast and even a bit of light rain threatened to dampen the hopes of the participants. But just as we prepared to abandon the effort, a break appeared in the clouds, followed shortly by several more. Within half an hour, overcast had turned into magnificent, cloudless skies. All the southern jewels were visible, including the first magnitude stars Canopus, Alpha and Beta Centauri, Achernar, and Acrux in the Southern Cross (just over the horizon). The telescopes in the group provided for many a first-ever glimpse of the globular cluster 47 Tucanae, the enigmatic nebular object Eta Carinae, and a host of southern-sky-only wonders. But the famous Magellanic Clouds, the nearest extra-galactic objects to our own Milky Way, stole the show. Many were caught by surprise at what large and bright wonders these objects were, lacking as they do any truly comparable counterparts in the northern sky. After two hours of intense observing the participants consisted of those too frozen in the nighttime mountaintop temperatures to continue, and those who would have stayed until dawn if transportation arrangements had permitted. So the buses made their way back to Santiago.

On the same evening, the Iquique group drove up the precipice separating their coastal city from the high-altitude Atacama desert. The clouds over the city were soon penetrated during the ascent, and higher, distant cirrus clouds were easily avoided, resulting in a spectacular desert-sky view of the same southern sky wonders seen at Ferallones. Some expressed the view that nothing the eclipse could offer would top the wonders seen that evening -- although that remark was not heard again following the eclipse!

The Day of the Eclipse

On Wednesday, the day preceding the eclipse, the first group from Santiago boarded an early morning flight north The remainder of the Santiago contingent took the evening flight. Both flights landed in Arica, the northernmost major coastal city in Chile. From there, buses transported expeditioners about 100 km into the desert to our selected campsite. In the meantime, buses drove the Iquique contingent the 300 km to the same campsite at the small desert village of Codpa.

Codpa is an amazing oasis in the middle of a truly barren desert. The Atacama Desert is the driest in the world, containing some places where no precipitation has ever been recorded. This dryness makes it unlike many other deserts in that there is no cactus or sagebrush, no snakes or scorpions, just rock and sand.

Yet in the middle of all this barrenness lies the valley of Codpa, surrounded on all sides by higher altitude desert. Because the run-off from a high peak in the nearby Andes mountains is channeled underground into this valley, it has trees and grass, and is generally habitable -- a stark contrast with its surroundings. The village has perhaps 100 inhabitants; and the campsite for the Eclipse Edge group had 16 regular cabins for tourists and vacationers from other Chilean cities. It even had a swimming pool!

Of course, 16 cabins was far from adequate for our expedition. So by pre-arrangement, the Chilean army pitched tents with thin mattresses and blankets for overnight sleeping accommodations. But the desert skies were so spectacular, and so filled with unaccustomed wonders; and the fast-approaching eclipse generated such excitement, that few got much sleep that night. Indeed, many rose at 4 am instead of 5 am just to have extra time to enjoy the morning pageant of the southern stars.

Finally, after breakfast, the buses set out for our selected viewing site near the edge of the eclipse path, where the most eclipse-related phenomena would be visible. But dawn brought with it some high cirrus cloudiness in the east, where the Sun would rise. These were not thick enough because of their extreme altitude to prevent eclipse-viewing; but they threatened to dull some of the phenomena.

The buses arrived, and all expeditioners poured out and ran over to the edge of the arroyo for a spectacular view toward the east of the pre-sunrise skies and the shadowed canyon below. Video cameras began whirring, and those with equipment selected their spots and began setups. Some police from the Chilean army and some Codpa residents accompanied our group. Wishing to be helpful, they kept hot beverages available to the group throughout our time at the viewing site. That was especially nice in the late partial eclipse stages, when the loss of sunlight combined with the altitude to make the air rather chilly.

One of the new items of equipment brought to this eclipse was a Global Positioning System (GPS) receiver. We used an evaluation unit of a new-series Motorola land-survey-type receiver, which provides time to a precision of 100 nanoseconds, 3-coordinate position anywhere on Earth to a precision of 100 meters (30 meters after post-processing to remove some of the deliberate degradation of system accuracy for national security reasons), plus speed and heading (if in a moving vehicle) and much other information. Using the network of 24 orbiting satellites with atomic clocks on board, eventually GPS receivers will allow planes to land safely in zero visibility, automobiles to keep track of their own locations on video maps, and many other applications. Hand-held navigation units are already available for under $500; but land-survey-type units generally are more elaborate, and cost $3000-$4000. The new evaluation units we used cost under $1000, but must be attached to a PC for the display and data processing. A second identical unit was used by David and Joan Dunham’s group at the northern edge of the eclipse path in Bolivia.

A beautiful sunrise brought new photo opportunities. It also revealed layers of thicker clouds in the Arroyo below, extending out to the coast. Indeed, as we later learned, the sea-level coastal city of Arica was generally clouded out for the eclipse, as was our original prime observing site near the coast at an altitude of 700 meters. But we always had the option to gain altitude by traveling deeper into the desert and toward the Andes mountains, since we had enough buses to keep the group as mobile as needed. And as our plans evolved, it became clear that our overnight at Codpa gave us easy access to our actual desert viewing site at an altitude of 2200 meters. This was not so high as to cause altitude discomfort, yet high enough to be above all but the highest tier of cirrus clouds.

Totality

The partial phases of the eclipse provided their usual interest. One could notice changes in the landscape and the air as more and more of the Sun’s disk was covered during the 90 minutes before totality. Some expeditioners used pinholes to project a crescent image of the partial eclipse onto the ground, while others criss-crossed their fingers to get a similar effect. Binoculars and small telescopes revealed sunspots and a magnified view of the proceedings.

Beginning about two minutes before totality, shadow bands are often seen on the ground. I heard one report that a few were seen faintly; but generally, the elusive bands did not put in an appearance at this eclipse. That is common from centerline sites; but in six previous expeditions to the path edges, the bands had always been visible, often spectacularly so. It appears that the bands may have been lost at this eclipse because of the thin cirrus clouds present. But perhaps we do not yet understand all the relevant factors in predicting their appearance. In any event, reports from all other sites I have heard from in Chile, Bolivia, and Peru, including northern-path-edge sites, also had high cirrus clouds and no shadow bands.

Shortly before totality, the Moon’s dark shadow could be seen sweeping across distant mountain ranges to the northwest of our site, bearing down on us at a speed of 100 kilometers per minute.

The diamond ring formed about 20 seconds before totality, and made the last large, bright bead on the edge of the rapidly appearing coronal ring into a true gem in the sky. This was repeated in reverse just after the end of totality. Other expeditions that went to centerline sites also reported spectacular diamond rings. Theirs, of course, lasted just seconds. When comparing notes with our expeditioners, and told that our diamond rings lasted 20 seconds each, one of the centerline travelers grumbled “That must be a subjective impression.” Wait until they see the video!

Bailey’s beads formed. Two were reported seen by some keen-eyed observers without aid. As many as ten beads, traveling along the limb as the partial stage ended, could be seen projected onto a paper screen from the eyepiece of the largest telescope at the site.

Then the eclipse became total. Usually, the corona is regular and featureless near sunspot minimum, as in 1994. However, on this occasion, two large coronal streamers were easily seen on the north side of the disc. Planet Venus was brilliant, close to the Sun. Mercury and Spica could be seen above the Sun by some, with Jupiter below. Mars was noticed in the west by the few who turned to look in that direction. But most were mesmerized by the spectacle in the east.

The beautiful red chromosphere, the only part of the phenomenon that can add color to totality, remained visible throughout the total eclipse from our site. If seen at all, it peeks out for mere seconds for centerline observers. Unfortunately, it is also difficult to photograph, even from edge sites. So most videos from our expedition do not show it. However, some of the video and some stills that had exposures specially adjusted to bring out the faint colored hues do show it well.

An eclipse of any length is always over before viewers are ready to let it go. However, a single snapshot of totality remains in their memories forever. The most frequently heard comment immediately after totality ends and the applause and whoops subside is “When is the next one?”

Eclipse Duration & Science Results

We have learned from experience that the best trade-off of length of totality and enhancement of other phenomena seems to occur about 1/18 of the way from the true path edge to the centerline, at which place the length of totality (for all total eclipses) is 1/3 of the centerline duration. Near our site in Chile, the total path width was 174 km, and centerline duration was 179 seconds. So in our advanced planning, we intended to locate ourselves 4.8 km inside the eclipse path's southern edge where the length of totality should have been 60 second

When we scouted that spot along the desert road, the best place to set up observing instruments was about 0.5 km further south (closer to the edge), where the duration of totality was nominally about 56 seconds. So that was the site we actually chose. However, video recordings of the eclipse showed that totality actually lasted between 40 and 43 seconds at our site, apparently placing us just 2.4 kilometers from the path edge, about two kilometers closer than we intended to be.

Initially, we had no idea what caused this discrepancy. Since this was the first use of GPS receivers for positioning, there was initial concern that something had gone wrong in using the new technology. We also worried that the corrections for our 2.2 km altitude above mean sea level had not been done correctly. An investigation was begun by several parties interested in the precise prediction of eclipse circumstances as soon as TVF returned to Washington, DC. David Dunham at the Applied Physics Lab, Alan Fiala at the U.S. Naval Observatory, and Fred Espenak at NASA, all participated in this post-mortem analysis.

The GPS site position data, recorded on portable-computer hard disk, was checked to redetermine the actual site location, which was then fed to the U.S. Naval Observatory and NASA computers for a new prediction of circumstances at our site. Using an average lunar profile, the predicted duration came out close to the nominal value of 56 seconds. But when the true lunar profile for our specific site was used instead of an average profile, the predicted duration dropped to about 40 seconds.

Such a large drop in totality duration had not been previously encountered, and seemed to surprise everyone involved in the analysis. Indeed, inspection of the true lunar profile of craters and valleys at the Moon’s limb revealed that we had hit something of a worst case scenario for discrepancy with the predictions. We actually were about 4.3 kilometers away from the true path edge, as expected. But just at our location, eclipse second contact (commencement of totality) occurred at a spot along the lunar limb where the deepest crater on the entire profile was located, delaying the onset of totality by about 12 seconds. That was coupled with encountering another deep crater at third contact (end of totality), costing us another 3-4 seconds.

The NASA eclipse circular for the 1995 eclipse across India is already in proof, and it is too late to change it. But for all future eclipses, the predictions will include a table giving the actual durations of totality as a function of distance from the predicted path edges. This turns out to be a more interesting and complicated function than had been suspected. Even centerline observers are affected to a degree. Fred Espenak reports that his new table shows that, for the 1995 Oct. 24 eclipse in India, the actual maximum duration of totality occurs at 7 km north of the predicted centerline! Since the eclipse is a short one, that maximum duration, 49 seconds, will be comparable to the totality our edge-group had in 1994 near the southern path limit in Chile.

Meanwhile, on the other side of the eclipse path, preliminary data from Dunham's site (intended to be about one kilometer inside the northern limit of the path) indicated that the predicted duration was very close to the observed duration near the northern edge. Dunham was using a GPS receiver identical to ours at the southern edge, and remarked, "it's a good thing we had it, since the 1:50,000‑scale map we had was over a kilometer in error, and if we had set up just using it, my site might actually have been just outside the true northern limit." TVF had noted similar problems with the Chilean topographic maps near the southern edge, finding several ground features in inconsistent locations as judged by the map. So the GPS receivers can apparently be relied upon to give more dependable positions of observing sites than topographic maps, at least in many parts of the world.

A combination of timing results from the two path edges will enable the path width to be determined with great precision. This will translate into small variations in the Sun's diameter that have been detected from past expeditions to the path edges. When studied further, these variations have important implications for both our understanding of ongoing processes in the Sun, and for climate variations on the Earth. This is just one of several scientific projects that participants in our Eclipse Edge expedition contributed to.