When the Gales of November Came Early: The Sinking of the Edmund Fitzgerald

On the 10th of November 1975 at 7:10 PM, Captain McSorley radioed “We are holding our own” to the Arthur M Anderson in what he had earlier reported were the worst seas he had ever seen in his 37 years of service. At 7:20, the Anderson noticed McSorley’s vessel had disappeared off radar and attempted to raise McSorley on the radio. At 8:32, the Anderson called the Coast Guard to report the wreck of Captain McSorley’s vessel, the Edmund Fitzgerald. After the search, no sign was found of the men of the Fitzgerald. All twenty-nine men had gone down with her. The subsequent United States Coast Guard and National Transportation Safety Board (NTSB) investigations, while not coming to an absolute conclusion about the sinking, came up with several recommendations that would hopefully prevent a disaster of similar magnitude from happening on the lakes again.

 

Weather Prediction on the Lakes and the Fitzgerald

To understand this incident, we first have to discuss the system of weather reporting on the lakes and a quick description of the Fitzgerald herself, as the Fitzgerald was one of the longest and most well known of the Lakers (term used for bulk cargo carriers that solely sail on the Great Lakes, as opposed to Salties, which travel on both the Ocean and the Lakes) and is the largest vessel to have sunk on the lakes (1)

 

In 1975, the year the Fitzgerald went down, there were ten United States Coast Guard weather stations, two Canadian weather stations, and two automatic stations surrounding Lake Superior. The Coast Guard would take their observations every two hours, and then broadcast their weather reports every 6 hours (1). The automatic stations, by virtue of being unmanned, would sometimes go down during severe weather and wouldn’t be able to be relied upon (1) The National Weather Service, of course, also monitored what was happening across the United States to make predictions about Lake weather to give gale and storm warnings and use that data to predict wave heights.

 

But the fixed weather stations were not the only ones taking readings. Forty Lakers were equipped with NWS anemometers, barometers, and thermometers who would report when weather conditions were more severe than predicted (1). Over the 42 hour duration of the storm on Lake Superior, there were 31 weather observations from these freighters (2). Some weather observations are less scientific however, wave heights were estimated based on a mate looking over at several waves and estimating the average height of them with no equipment.

 

The Fitzgerald herself was known as being one of the most reliable weather reporters on the lake (1) and had all of the NWS equipment. She did not, however, have a fathometer, which would measure the depth to the lake floor, as these were not common at the time (4). At the time of sinking, the Fitzgerald’s master Captain McSorley was considered the best captain in Oglebay Norton’s fleet with the most time in service of any of their captains. The Fitzgerald herself had passed all recent inspections, including a joint Coast Guard and American Bureau of Shipping inspection that took place on October 31st, 1975, less than two weeks from the loss of the Fitzgerald(1). The Fitzgerald also had never had a report of water in her hold after unloading which would imply previous integrity of the cargo hatches, but James B. King of the NTSB noted that Taconite can absorb 4 to 6.7 percent of its weight in water, meaning around 950-1600 metric tons of water could have previously infiltrated the hold without any trace of water left in the hold after the ore is removed. Finally, in terms of electronic equipment, the Fitzgerald was equipped with five radiotelephones,  two of which were battery-powered, as well as two surface scan radar sets. In all though, the Fitzgerald was seen as a seaworthy, dependable boat with an experienced crew, not the kind to be expected to go down with all hands.

 

The Final Voyage and Sinking

On November 9th, 1975 at 2:15 PM EST, the Fitzgerald set out from Superior, Wisconsin with a full load of taconite pellets headed for Detroit, Michigan (4). They passed Two Harbors, Minnesota about two hours later, when the Arthur M. Anderson, a similarly sized laker was leaving the harbor there for Gary, Indiana with a similar cargo. (4) Due to reports of a large storm coming in with winds expected to be coming from the Northwest, the Fitzgerald and Anderson chose to deviate from their planned route and follow the Canadian shoreline, which would protect them from the strongest winds and highest waves. On the morning of November 10th, the Fitzgerald reported to Oglebay Norton offices that time of arrival to the locks at Soo Locks (the locks that separate Superior from the other Great Lakes) was indefinite due to the storm. At 3:30 PM on November 10, while passing between Michipicoten and Caibou islands, the Fitzgerald reported “I have a fence rail down, have lost a couple vents, and have a list.” (3) Many, including Captain Cooper of the Anderson and Philip Hogue, dissenting member of the NTSB report, believed that the Fitzgerald’s hull broke here by hitting ‘Six Fathom Shoal’ a relatively shallow part of the lake whereas the concurring members of the NTSB and Coast Guard claimed it was due to the Fitzgerald striking something on the surface. It was at this point the Fitzgerald started pumping water out of her hold. Soon after, both radars on the Fitzgerald were reported inoperable, and the Fitzgerald started to rely on the Anderson for navigation. The two boats started to turn south into Whitefish Bay and encountered a salty called the Avafors at 5:30 PM, who received a message from Captain McSorley reporting a bad list and that the seas were the worst he had ever encountered. At 7:10 PM, the Fitzgerald sent her last message, not reporting any imminent threat, and was gone within the next ten minutes. At the time of sinking, the Anderson was reporting winds from the West-Northwest at 40 knots with gusts at 65 knots, and wave heights 5.5 to 7.5 meters.

 

After the sinking, and before the Coast Guard and NTSB reports were published, a submersible was sent to investigate the wreck, found the Fitzgerald broken in two on the seafloor with several hatches completely missing or buckled from external load. The Fitzgerald also had sunk in water shallower than she was long. There are two main theories about how the Fitzgerald sank. The first is what was mentioned earlier. The hull was broken on Six Fathom Shoal causing water to infiltrate the hold, leading the Fitzgerald breaking up on the surface and sinking, with the hatches buckling or breaking free from the impact with the lake floor. The official theory from the NTSB and Coast Guard is that the Fitzgerald lost hatches leading to water coming in from the top and eventual buckling of other hatches under the load of the high waves on the increasingly laden down boat. Since there were no bulkheads in the cargo hold, once one hatch broke, the entire hold was flooded quickly, driving the Fitzgerald under, and she broke up when she hit the lake bottom before there was any time to abandon ship or radio for help.

 

Storm Analysis

Now, what was this storm that brought down the Fitzgerald? There was of course the monitoring of the storm during the actual event, and due to the sinking of the Fitzgerald, there was analysis in the NTSB and Coast Guard reports, but due to the high profile nature of the event, there has been analysis with modern techniques.

 

On November 8th, the storm originated over the Oklahoma Panhandle, with a minimum pressure of 1000 hPa, and was moving to the northeast. By November 9th at 7:00 AM Eastern, the storm was over south-central Kansas and moving to the Northeast and the National Weather Service put out its first forecasts of the weather over Lake Superior.

 

 As it moved Northeast over Iowa, the pressure dropped and was predicted to travel over Lake Superior, then traveling North over Ontario, reaching James Bay at 7:00 PM on November 10th and the NWS issued Gale warnings for the lake. While the storm was over central Wisconsin, the Gale warning was upgraded to a storm warning. The pressure continued to drop, and by the time the storm center was over Marquette, Michigan, at 7:00 AM November 10th, the pressure was at 982 hPa. (3) At this time, the surface low corresponded with the 850 hPa low, with a trough to the west at 500 hPa and 300 hPa with the low over Lake Winnipeg going to Eastern Illinois. Since much of Lake Superior is to the upstream of the trough, there was a zone of divergence favorable to enhanced cyclonic activity. (2) The National Weather Surface was warning of waves from 2.5 to 4.9 meters and Northwest winds at 35 to 52 knots with gusts up to 60 knots(4). During the storm, there were also repeated snow squalls. The last time that the Anderson saw the Fitzgerald on radar, the Fitzgerald was blurred due to heavy snow and clutter from the waves. (2) This snow then cleared shortly after, increasing visibility, and the crew of the Anderson were able to see the Salties and the shore, but not the Fitzgerald, indicating that it had gone down.

 

One of the most interesting discrepancies in the wave heights that were forecasted by the National Weather Service against those recorded by the Anderson. The NWS did have a meteorologist go over the calculations but still came up with wave heights of 4.9 meters. And while the method of recording wave heights on a ship underway was less precise, the margin of error was too high for it to be a simple mistake. So the NTSB and Coast Guard made sure that part of their plan for the future was to spend more time researching wave height prediction, so they would not be caught unawares again. The modern analysis by Hultquist, Dutter, and Schwab helps explain this. While wind speed and fetch are still a primary driving factor of wave height, atmospheric instability can lead to larger waves due to increased vertical movement. DJ Schwab would develop a model for wave height that would account for stability as well based on previous work by PC Liu and DB Ross. 

 

So, in addition to the fact that the Fitzgerald was at the easternmost part of the highest wind zone traveling west-northwest along almost the longest fetch possible on Lake Superior, there was the added instability of cold air over the warmer lake. When the Fitzgerald set off from Superior, the lake temperature was less than 1 C warmer than the 925 hPa temperature, but after the cold air mass from the storm moved in, was 8 C warmer than the 925 hPa temperature (Hultquist et al. also point out that this instability also helped increase the wind speeds on the lakes, driving the wave heights in the more traditional sense). When this is applied to the updated wave model, wave heights of 5.5 to 7.5 meters are confirmed, with winds gusting at 70-75 knots. (2) If the NTSB was right, and the deck was almost at the surface of the lake at this point, there would have been an additional 370 hPa pressure on hatches, which would be more than enough for them to fail. It is also worth noting that the frequency of waves in excess of 7m is less than .1% on Lake Superior and these mostly travel from North to South, likely confirming Captain McSorley’s comment on these being the worst seas he had ever encountered. Models are important for the wave height data, as, while NOAA does have weather buoys on Lake Superior, they are removed from November to April to avoid damage from ice, making the stormiest parts of the year go unrecorded. (2)

 

The Lasting Impacts

As mentioned prior, the Edmund Fitzgerald sinking had a large impact. Not only did it impact those living around the Great Lakes at the time and have a song by Gordon Lightfoot written about, but it affected how shipping on the Great Lakes was done. At the end of its report, the NTSB went on to outline their suggestions to make sure an incident such as this would not happen again.

 

In addition to new regulations on ship design and crew regulation, they also suggested changes in how ships behaved in weather, and how the government responded to it. To NOAA, they directed that they find a new way to accurately determine wave heights on the Great Lakes, as the method at the time greatly underestimated wave heights. As mentioned in the Storm Analysis section, the new methods devised for wave height prediction added in surface stability into the calculations, which has been able to more accurately predict wave height on the Great Lakes. To the Coast Guard, they sent out regulations that would prohibit sailing in weather conditions that exceeded vessel designs, requiring them to drop anchor and wait out the storm. They also instructed the Coast Guard to increase their search and rescue capabilities on the lakes, as the bad weather prevented them from sending out search and rescue boats or aircraft immediately after the incident was reported, and had to rely on the Anderson and a few Salties to search for the Fitzgerald. 

 

Since 1975, there hasn’t been a disaster that comes close on the Great Lakes, and the memory of the Fitzgerald lives on. Today the site of the Fitzgerald’s wreckage is a registered gravesite by the government of Ontario (As the Fitzgerald sunk on the Canadian side of Superior, it was under their jurisdiction).

 

Now, to play myself out, of course, Gordon Lightfoot:

 

Postscript Note

This is lightly edited from an essay I wrote for a meteorology class, hence the emphasis on meteorology.

 

References

Hultquist, T. R., Dutter, M. R., &; Schwab, D. J. (2006). Reexamination of the 9–10 November 1975 “Edmund Fitzgerald” Storm Using Today’s Technology. Bulletin of the American Meteorological Society, 87(5), 607-622. doi:10.1175/BAMS-87-5-607

King, J.B., McAdams, F.H., Driver, E.T., & Hogue, P.A. (1978). Marine Accident Report SS EDMUND FITZGERALD Sinking in Lake Superior on November 10, 1975 (Report No. NTSB-MAR-78-3). Washington, DC: National Transportation Safety Board.

Oolman, L. (n.d.). Atmospheric Soundings. Retrieved December 14, 2020, from http://weather.uwyo.edu/upperair/sounding.html

Schumacher, M. (Ed.). (2019). The trial of the Edmund Fitzgerald eyewitness accounts from the U.S. Coast Guard hearings. Minneapolis, MN: University of Minnesota Press.

U.S. Coast Guard. (1977). Marine Board Casualty Report SS EDMUND FITZGERALD Sinking in Lake Superior on 10 November 1975 With Loss of Life (Report No. USCG 16732/64216). Washington, DC: U.S. Coast Guard.