First Total Eclipse Cloud Out

Tony Crocker

Staff member
:-( Liz and I are on the Paul Gauguin cruise ship, from which I successfully observed the 2010 eclipse:

We were not so lucky this time. It's been continuously cloudy ever since we left Tahiti June 26. We reached Pitcairn Island, where descendants of the Bounty mutineers still live, on June 30. The islanders came out to sell souvenirs on our ship, but unlike the 2005 eclipse cruise, passengers were not allowed to go ashore due to swells up to 3 meters and Bounty Bay's marginal harbor.

While at Pitcairn the weather two days ahead showed a large cloud band and the captain had to make a decision then whether to get in front of the clouds to the NE or behind them to the SW, in either case a 500+ nautical mile trip of over 36 hours. Distance wise it was a tossup, but the NE detour would have taken us so far from Tahiti that we would miss at least one port call. So we went SW and arrived this morning to mostly thick cloud the first 15 degrees above the horizon and broken cloud another 15 degrees above that with overhead mostly clear.

The sun would be 21 degrees at totality so last minute maneuvers would be likely, and the ship did in fact change direction three times. We were at about 27 degrees south latitude, winter water temp 66 degrees and a lot of humidity. Thus the cooling of the eclipse itself condensed more water vapor rapidly during the last 15 minutes before second contact, resulting in 90+% cloud cover until well past totality. And of course the upper sky gradually cleared as it warmed some later in the morning.

In more tropical locations "eclipse cooling" reduces heat convection and cumulus clouds, but where we were the effect was much the opposite.

Liz and I no longer have unblemished records of 11 and 9 eclipses respectively. Several other people lost longer perfect records:
Travelquest (with whom I bought both Paul Gauguin cruises and Liz the 2008 North Pole eclipse flight) tour company owner Aram Kaprelian had 17.

Bill Kramer was also at 17 and will go over 1 hour lifetime totality at his next success. I met Bill on the Mediterranean Celebrity Galaxy cruise that included the 2004 transit of Venus. This is his 4th eclipse cruise on the Paul Gauguin.

Cal Berkeley astronomer Alex Filippenko had 15 successful totalities before this one. He is the astronomy advisor for many travel groups including our Iceland trip in 2015.

David Buchla, who made the elaborate dragon pinhole projection shown in my 2010 TR, also had 100% success in 15 total solar eclipses before this one.

Preliminary reports are that all went well in Chile and Argentina. We would likely have been successful going NE rather than SW, but that's Monday morning quarterbacking. Skiers know better than anyone that you are always at the mercy of the weather and that $#!% can happen.

We are more critical of the failure to land on Pitcairn Island. People must ascend/descend an 18-foot rope ladder from the lowest floor of the ship to the Pitcairn longboat and also time their entry/exit from that boat onto the Pitcairn jetty in the ocean surge. Nonetheless several islanders in our age bracket did this, bringing their crafts to our ship.

This is a 14 -day cruise, and my observation is that the longer the cruise the older the age range of passengers. It's a fascinating group of well traveled people, many of whom started eclipse chasing two decades before we did. Nonetheless some of them are noticeably slowing down, and we don't think the captain wanted to tell people who could and who could not safely make that longboat transfer. It's what Liz and I call "managing to the lowest common denominator," and for such a remote once-in-a-lifetime location, we think those people who were capable should have gone ashore. A rope ladder could have been set up on the pool deck to test people in advance.

About 15 cruise ships call at Pitcairn per year, and most of them only receive the islanders and make no attempt to land passengers, as was the case with my friend Richard on Oceania last year. Anecdotally we heard that only about 10% of cruises land passengers at Pitcairn.

Pitcairn gets a supply ship every 3 months that also carries 12 passengers. Those passengers can stay on Pitcairn for a few days. If you really want to set foot on that island, that's the most likely way to do it.

Tony Crocker

Staff member
MarcC":2wqwvaaq said:
It got really dark for a couple of minutes. Lights came on. People had these pinhole cards that projected the image onto the ground. Some animals freaked. Then it was over. Eh.
I can sadly say now that's a very accurate description of a clouded out eclipse, including the "Eh" part.

No surprise we have many excellent photographers on board, and Linda Cornfield summed it up well.

After 8 days at sea, we are at least now enjoying French Polynesia. The locals say they have rarely seen this much rain during July, which is normally the peak of winter dry season. Today we were at a vanilla plantation on Tahaa. Both last year and this year the flowering of the vanilla plants have been delayed at least two months by abnormal winter rain.

Tony Crocker

Staff member
I documented the eclipse cruise decision making process shortly after the cruise was over based upon extensive conversations with knowledgeable eclipse chasers on board.

While that documentation relates to cruises, I've read or heard about numerous land tours that have been inflexible and clouded out, notably Shanghai 2009, where eclipse climatologist Jay Anderson issued a strong weather warning nearly a week in advance but few tours took note and relocated. We expected better from the Travelquest travel company and the Paul Gauguin based upon past history but the detail analysis indicates that we were too complacent.

Eclipse Cruises and How to Improve Them

Prior Experience
2019 was my fourth eclipse cruise. In 2009 I was on Costa Classica charted by Roy Mayhugh. The ship left Kagoshima about 45 hours before totality and arrived in Kobe about 42 hours after totality. The stated objective was the maximum eclipse point east of Iwo Jima but Roy said with their port schedule that they could work with as much as 1,000 miles of the path. It turned out that the area around Iwo Jima was quite clear even though the two sea days on the way to/from the max point were mostly rainy. Costa Allegra was also at the 2009 eclipse but was concluding its cruise the next morning in Shanghai. Being constrained to a short distance from Shanghai it was clouded out.

In 2010 I was on the Paul Gauguin. The eclipse objective was about 120 miles south of Tahiti. One full sea day was allowed to reach this area from Bora Bora for a ~8AM eclipse. The day after the eclipse we needed to be in Moorea with the cruise ending the morning after that. At the eclipse there were scattered puffy clouds and the captain had to turn the ship just before second contact to outrun them. We had two cloudy periods of 20 and 10 seconds within the 4 minute totality, and everyone heaved a sigh of relief that the last minute maneuvers had been successful.

In 2016 Liz and I were on the Damai II, a scuba liveaboard in Indonesia. The travel agent had never seen an eclipse and planned a custom itinerary for her best customers. Liz and I met her at the Long Beach Scuba Show in 2013 and persuaded her to let us sign up and be the “eclipse experts.” We saw the ~9AM eclipse just north of the equator near the Goraici Islands, where we had scheduled diving after noon. We saw weather maps the prior day which suggested that perhaps we should be farther west and farther from land. The captain responded that “weather maps aren’t very reliable out here” and stuck to the planned itinerary. In the last hour before second contact we were in the clear but sailing toward clouds and only temporarily did the captain slow the progress toward our scheduled afternoon dive sites.

2016 was our first direct realization that eclipse cruises all have a planned itinerary, and the default position is that itinerary will be followed regardless of weather. We had read of some sad tales in 2012 of ships sailing into clouds and the infamous Pacific Jewel failing to reach totality at all.

For 2019 I returned to the Paul Gauguin, this time with Liz who had never been to French Polynesia. Another attraction was the chance to visit Pitcairn Island, which was not achieved due to ship liability concerns with the clientele in challenging weather. We had no real concerns about the capability to chase the eclipse. Paul Gauguin had successfully viewed 4 prior eclipses. All of these cruises had been chartered by Wilderness Travel and Travelquest. Travelquest specializes in astronomy travel and its president had 17 successful eclipses without a miss. Noted climatologist Jay Anderson is retained by Travelquest to assist in planning itineraries and provide weather forecasts leading up to the eclipse.

Paul Gauguin 2019 Eclipse as Experienced by Passengers
The three sea days on the way to Pitcairn and June 30 at Pitcairn were consistently overcast with occasional rain. Though Pitcairn is 23 degrees latitude the ocean temperature was a chilly 66 F vs. 80F in the Society Islands at 17 degrees. The original plan was to stay at Pitcairn through the next morning, then move over the next 20 hours to the eclipse path. In 2017 that point on the Travelquest website was NW of Pitcairn past Oeno Island but on the cruise the map showed a point due west and farther from Pitcairn.

During the afternoon at Pitcairn the captain, the astronomers and trip leaders examined weather forecast maps for the eclipse that showed a wide band of solid overcast centered at Pitcairn’s latitude. The ship needed to leave Pitcairn soon in order to reach the eclipse path beyond the wide band of thick clouds. The ~550 nautical mile distance was about the same to the SW or the NE. A move to the NE would mean skipping the half day port call in Rangiroa on July 5 and probably the July 6 port call in Bora Bora. So at 4:30PM on July 1 the astronomers announced that we were going to the SW. The target for the eclipse was to each 27 degrees 52.7’ S, 140 degrees 00.5’ W by 8AM July 2, and to view totality there at 10:13AM.

We heard that Jay Anderson was on the road in Chile with another Travelquest eclipse group on July 1 and not reachable during the 2-3 hours that the decision was made to go SW rather than NE.

On the morning of July 2 there were westerly winds of 30 knots. Thus at about 7AM the captain slowed the ship to 5 knots to lessen the winds on deck where people were already gathering. At 9AM first contact the sun was in clear view in a hole even though it was only 9 degrees up and in general there was thick cloud up to 15 degrees and broken cloud up to 30 degrees above the horizon. Most of us were not that concerned because we figured last minute maneuvers as in 2010 could find a break in the broken cloud layer. Shortly after first contact the captain turned the boat 180 degrees to sail at 5 knots with the wind to make the viewing deck more comfortable.

What most of us passengers did not think about was the cooling effect of the eclipse itself. About 15 minutes before second contact the clouds practically exploded upward and soon we were in at least 90% cloud cover. Some passengers were observing a bright spot on the horizon but it was then too late to chase it. The heavy overcast lasted 20-30 minutes past third contact, then gradually retreated to a similar level of cloud as around first contact.

Post Mortem of the Paul Gauguin Eclipse Chase and Decision Points
This cruise was well populated with veteran and knowledgeable eclipse chasers. At least two of them were examining weather maps independently. With 8 more days on the cruise there were many discussions and questions raised. When and how might different decisions have been made? I’ve already said I was too complacent from past experience, and I believe that was true for many people including the decision makers. There have been “close calls” before as in 2010, but the decision process was perhaps not reviewed then as we hope it will be now.

Jay Anderson’s July winter weather maps for the Southern Hemisphere are quite consistent. The higher the latitude, the greater the average cloud cover.

Thus I found it puzzling on the cruise that the original plan to view the eclipse NW of Oeno was changed to a higher latitude due west of Pitcairn. What’s not on average cloud charts is the cooling effect of the eclipse itself. The cooling can be a positive effect in tropical climates by reducing heat convection. But with 66 degree water and being just past a heavy cloud layer, perhaps the likelihood of the eclipse cooling generating more clouds should have been anticipated. Surely this is a greater danger at 27 degrees latitude to the SW of Pitcairn than at 20 degrees latitude to the NE with air and sea temperatures 10C warmer.

Meteorology and the NW vs. SE decision at Pitcairn:
The ability to acquire and analyze a lot of data by satellite on board ship is limited. There needs to be land based backup to do that. It’s not reasonable to expect Jay Anderson to do it when he’s on a different trip with intermittent access to South Pacific weather data. And on the ship we had two expert astronomers but they are not meteorologists. A trained meteorologist might have known the dangers of eclipse cooling but most of us never considered it until we saw it happen so dramatically.

Weather Observation: How Soon Before the Eclipse Could Clouds Have Been Avoided?
Weather maps have been analyzed after the fact. The eclipse cooling can be identified in some of that data. There were some holes in the clouds at the time of totality where humidity was lower. The estimate is that a low humidity area needed to be identified 90-100 minutes in advance for the ship to change its heading and make it there on time. This is another situation where the data needs to be collected on land to be timely, with frequent updates being communicated to the ship.

What actually happened was that the Paul Gauguin moved back and forth in close proximity to the 27 degrees 52.7’ S, 140 degrees 00.5’ W target location that had been set 40 hours earlier at Pitcairn. No attempt was made to chase brighter horizons and it seems very likely that the decision makers were as surprised by the eclipse cooling effect upon the clouds as most of us passengers were. It also seems likely that the decision makers had no information in the hours before second contact that there might be better options than the target set 40 hours before.

Later in July Joe Rao published on the Solar Eclipse Mailing List how he, as on-board meteorologist in 1991, consulted at dawn with a land based meteorologist in Honolulu to move two cruise ships into a hole among the widespread clouds near Hawaii's Big Island. Those ships headed for that hole after dawn at full bore speed of 26 knots. This is exactly what should have been done on the Paul Gauguin but we had neither on-board or real-time land based meteorologists available to make the attempt.

We have also learned that the Paul Gauguin 2010 eclipse cruise was a closer call than necessary. The ship was in quite clear skies at first contact but sailed to the exact coordinates planned in advance and in this case printed on commemorative T-shirts. This brought us into clouds, lost 30 seconds of totality and could have been more if not for last minute evasion.

Cruise Mobility, Is Reality Short of Expectations?
I would say first that the communication between trip leaders, astronomers and ship captains needs improvement. The situation of sailing to exact coordinates set in advance has been a recurring problem on many eclipse cruises including three of the four I have been on. Those coordinates should be viewed as a first cut, a suggestion not an order. The absolute priority should be avoiding clouds and a trained meteorologist/weather observer should be making those calls. Analyzing weather maps in the last few hours should be part of the process, probably with land based support having a fast internet connection. Direct observation/chasing holes or chasing the clear area you are already in applies to the last hour or two, not just the final minutes. It’s even OK not to be on the centerline and lose 15 seconds of totality to stay away from the clouds.

An Itinerary Not Optimally Planned for Eclipse Success
When we signed up in 2017 the viewing spot was shown on the TQ website as the shortest distance to the centerline from Pitcairn. That’s about 23d 30’ S, 130d W. That gave the impression we could move either way to chase clear skies. See attached Jay Anderson map of July cloud cover climatology. The only land speck within totality in the Pacific is Oeno Island, which is on the border between the blue (better odds of clear skies) and green (more tossup odds of clear skies). Advance planning should have picked a spot or have been prepared to move to the NE rather than SW of Oeno/Pitcairn.

When we got on the ship the viewing spot was due west of Pitcairn, about 24d 49’ S, 133d 39’ W. Why was it changed? It turns out that point is the easternmost from which we can reach Rangiroa by early morning July 5. It seems Travelquest/Wilderness Travel didn’t figure this out until well after they sold the cruise.

I measured distances from Rangiroa and Bora Bora to various eclipse points, because it’s clear now those were the key constraints how far NE we could go to chase the eclipse. From Tahiti to Pitcairn we took 81 hours averaging 14.58 knots. From Pitcairn to our eclipse viewing spot we took 39 hours averaging 14.41 knots. Accordingly I calculated hypothetical viewing points farther NE on the eclipse centerline assuming a speed of 14.5 knots, with the following results.

1) We can skip Rangiroa but make Bora Bora July 6 from 22d 22’ S, 128d, 13’ W. This is only 185nm from Pitcairn, not nearly far enough to get out of the cloud band we faced.
2) We can make Rangiroa July 6 (meaning Bora Bora port call is July 7) from 21d 00’ s, 124d, 53’ W. This is 377nm from Pitcairn.
3) We can skip Rangiroa but make Bora Bora July 7 from 19d 49’ S, 121d, 42’ W. This is 547nm from Pitcairn, the desired distance based upon information we had June 30.

I believe the cruise itinerary was poorly planned. From the beginning an extra sea day after the eclipse should have been scheduled in order to allow the ship to reach the lower latitudes with better weather prospects and longer totality. This could have been done by:
1) Making the cruise 15 days instead of 14, or
2) Dropping one of the lesser ports Huahine or Tahaa from the itinerary. Paul Gauguin cruises that go to the Marquesas usually have an abbreviated itinerary in the Society Islands.

For the Paul Gauguin 2019 eclipse cruise I believe there were issues of complacency and project management, with the most conspicuous examples cited above. I’m sure the decision makers are motivated to improve the planning and process in the future.

Tony Crocker

Staff member
The fine print in cruise contracts absolutely exempts the cruise line of any liability for missed port calls or cruise activities. As far as a clouded out eclipse is concerned, any skier knows that if they book a destination week of lifts and lodging and the weather/conditions suck, that's just too bad. Our Chile trip in December 2017 when we never set foot on the Futaleufu River due to adverse weather was another example.

My friend Richard has frequent status with Princess and bitched when 3 port calls were cancelled in Iceland and Greenland on a repositioning cruise from England to New York. He did get a fairly decent credit toward his next cruise.

I heard on the Paul Gauguin that some of the passengers on the infamous Pacific Jewel in 2012 sued and got refunds. That was marketed as an eclipse cruise but it left Sydney late and failed to reach the totality zone on time.

In this day and age most well run eclipse trips, land or sea, should have real time access to fresh weather maps in the hours leading up to totality. The exceptions are remote third world places where it's a chore just getting to the proper location, so tough logistics preclude mobility. So those locations need to be chosen far in advance based upon favorable climatology; Liz' trip to the temporary camp at Jalu in the Libyan desert in 2006 was a good example.

But the reality is that some cruises and many land tours in countries with good infrastructure don't plan for last minute mobility either. On land a good rule of thumb is the larger the group, the less likely it is to be mobile. But the bottom line is to ask the hard questions before you sign up for a tour.


Tony Crocker":38nicnx2 said:
Skiers know better than anyone that you are always at the mercy of the weather and that $#!% can happen.
The comparison with a ski vacation is only apt to a certain point. IIRC, eclipse cruises start in the high four-figures per person and more importantly -- once you're on the boat, you're completely at the mercy of those running the operation, i.e. not allowed the sea variant of unilaterally driving to a different ski area or region with better conditions.

Tony Crocker

Staff member
jamesdeluxe":eue2rbno said:
The comparison with a ski vacation is only apt to a certain point.
Consider that most destination ski trips are one week packages including lodging, lift tickets and airport transfers with no rental car. How many of those people pick up and leave their prearranged packages and spend more money to seek out better snow conditions? Very few I suspect. It's the same inertia as with eclipse tours. Most people, especially trip organizers (ski clubs, Epic/PugSki Gatherings) won't deviate from original plan come hell or high water.

James and I prefer the flexible rental car model, but we need to recognize that's a small minority practice.

jamesdeluxe":eue2rbno said:
eclipse cruises start in the high four-figures per person
Paul Gauguin and the Damai II dive boat were premium cruises. Mainstream large cruise line trips like Costa Classica 2009, Celebrity Millennium 2012 and Holland America Volendam 2016 were low four figures, certainly comparable to a ski week in a top tier resort in the US West or the Alps.

jamesdeluxe":eue2rbno said:
once you're on the boat, you're completely at the mercy of those running the operation
Just as you are on a typical foreign country package tour with hotels and bus transport included. Theoretically the cruise should be more flexible than a land itinerary with hotel reservations, though it's not necessarily so in practice.


Tony Crocker":25ltmwmd said:
James and I prefer the flexible rental car model, but we need to recognize that's a small minority practice.
Exactly, which is why I was trying to imagine the frustration when you couldn't do anything about the cloud-out. Congratulations for keeping your cool.

While NCP is always possible during road trips in the northeast, I can only recall two major events during fly-to destination trips where I wasn't able to do anything but ride it out. Out west, the infamous rain to the top of Snowbird in January 2003 (the day after, I broke my leg at iced-over Snowbasin) and the January 2016 gulley washer that covered a large portion of the Alps (followed by one of the most enjoyable ski days ever at Grand Bornand).

Tony Crocker

Staff member
jamesdeluxe":167oirbm said:
Exactly, which is why I was trying to imagine the frustration when you couldn't do anything about the cloud-out. Congratulations for keeping your cool.
At the time of the eclipse, my attitude was "$#!% happens," as with the weather debacles of the 2005 Tropical Punch in Canada and the Futaleufu River in Chile in 2017. Skiers know better than anyone about the vagaries of weather, and 10 out of 11 is still a quite good track record for eclipse success. The fanatics for whom eclipse chasing is the #1 travel priority with 20+ average around 85% success.

It was only in the post-mortem that I found out about the lack of real time weather input and that the eclipse viewing point was so constrained by subsequent port calls. I calculated all of those points on the Google Earth map in Tahiti after the cruise was done.

We made much more of a stink about the failure to land on Pitcairn.