by John D. Reid
As surely as "back to school" shoppers specials, 69 cent apples and shorter days, hurricanes are harbingers of summers end. September, prime hurricane season, can mean the last of the summer reruns preempted by itinerant television reporters breathlessly bringing the latest on the storm from a windswept Florida beach.
September 2000 saw the centenary of the most deadly hurricane, and natural disaster, in North American history. Remembered in song and fable, it caused more US fatalities than the legendary Chicago Fire, San Francisco Earthquake, and Johnstown Flood combined.
This mighty storm, a category four hurricane, battered the Texas coast on September 8th 1900 with 115 mile per hour winds. Thats the official figure, but no anemometer could withstand the hurricane blast. Estimates ranged to 125 miles per hour with gusts to 150. A fifteen to twenty foot storm surge drowned Galveston Island. The next day, a bright blue Sunday, one in six Galveston citizens at least 6,000 lay entombed in collapsed buildings, or bodies dumped at the whim of the receding flood.
The tragedy in Texas was the lead story in newspapers internationally for days, perhaps not surpassed in that aspect until September 11th 2001. Its hardly surprising that the continuation of the storm in Canada has been overlooked; yet it remained a killer. Fatalities north of the border, while only one hundredth those in Texas, still meant more deaths than from hurricane Hazel in 1954, which killed 84. And more than the 1998 Ice Storm, 2000 Pine Lake and 1988 Edmonton tornadoes combined.
History of the Storm
The storm was first noticed on August 27th in the equatorial mid-Atlantic. Still getting organized it drifted westward through the Greater Antilles with moderate winds. Nothing was moderate about the rain. Miles of railway roadbed were washed away in Jamaica. Santiago de Cuba was inundated by 10 inches of rain in just eight hours on Monday, September 3rd; two feet before it was over.
Ships were cast ashore in Florida from Palm Beach south to the Keys on September 5th. Gaining energy from tepid Gulf waters, the storm made a beeline for Galveston and its rendezvous with history.
The story of the hurricane in Galveston is told by Erik Larson in Isaacs Storm: the books title taken from the officer in charge of the Galveston office of the US National Weather Service Isaac Cline. Larson weaves together the stories of Cline, Galveston, the hurricane and the centralized operation of the US weather service. Cline (1900) wrote his own report shortly after the storm. Neither Cline nor Larson bothered much with what happened to the storm after Galveston.
What happened is it beat its way north through Texas, Oklahoma, Kansas and Iowa. On the morning of Tuesday September 11th , its tropical storm characteristics fading, the system was near Des Moines, Iowa, with a pronounced trough to the northeast. Five inches (127mm) of rain fell in Minnesota.
By 8 p.m. on Tuesday evening the MSC surface pressure analysis shows a binary structure with one pole, associated with the continuation of the tropical system, over southwest Michigan, and the other over Georgian Bay. Winds in Chicago, where the storm reaped another three lives, reached 80 miles an hour.
The Storm in Canada
The storm complex headed east. Ontario shook. Winds clocked up to 50 miles an hour pounded Toronto on Tuesday evening. Windows were smashed across the City, in homes, offices, churches and the Parliament building. Trees toppled and telegraph wires lashed about in the gale. Boats battered by boisterous eight to ten foot seas dragged anchors or broke moorings. Skiffs were up-ended and cast ashore.
Greater damage was inflicted further south and west. In the Niagara Peninsula, and along the Lake Erie shore, apples, pears and peaches ready for picking were ripped from the tree. Losses were estimated as half the value of the crop, about $1,000,000.
A flour mill fire in Paris, Ontario, escalated into a local disaster as strong westerly winds blasted the flames. Wooden buildings flared like tinder, all too familiar a situation in those days. A devastating fire had scorched Ottawa just five months earlier. Damage to the Paris Post Office, Customs House, Bank of Commerce and numerous small shops and businesses exceeded $250,000. Three quarters of the business district was in ashes, less than half of it insured.
There was only one fatality in Ontario, a worker at Niagara Falls keeping storm-derived debris clear of sluice intakes overnight was found drowned, and thought to have lost his footing.
Just south of the border it was a different story. Two ships were lost in Lake Erie. Eleven drowned when the steamer John B. Lyon foundered five miles off Conneaut, Ohio. Two additional lives were lost when the schooner Dundee sank 15 miles from Cleveland In Buffalo a women cleaning up debris was electrocuted by fallen wires.
Well north of the storm track there was significant rainfall. Some of the larger storm totals were: 58 mm at Bruce Mines (46 o 18N, 83 o 55W); 53 mm at Parry Sound; 54 mm at Emsdale (45 o 30N, 79 o 13W); 39 mm at Huntsville; 73 mm at Uplands (45 o 48N, 79 o 25W); and 78 mm at Haileybury
Further east, the Ottawa Evening Citizen reported the highest winds in that City as 30 miles per hour at 9am on Wednesday the 12th. Yachts at Britannia and Aylmer were stranded on the shore or dragged their anchors. The steamer tug Albert, in difficulties in gusty winds, had to release a load of 14,000 logs it was towing. They littered the shore of Lac Deschenes.
As if becoming breathless, racing toward Montreal, the storm was losing its punch. The MSC weather map shows the system centred on Montreal at 8 a.m. on the 12th with troughing elongated east-west. The isobars indicated a storm surge into the Bay of Fundy.
Most of Quebec appears to have escaped lightly as the storm scooted east. Rainfall totals were: 36 mm at Ste Agathe; 59 mm at Quebec City; and 28 mm at Chicoutimi.
Passing through the Gaspé and northern New Brunswick the system regained energy. Precipitation totals were 74 mm at Bic (48 o 22N, 68 o 42W) and 100 mm at Perce. To the south amounts declined:, 64 mm at Bathurst (NB), 35 mm at Chatham (NB), and 30 mm at Fredericton.
By 8 p.m. on Wednesday the 12th the storm centre lay offshore of Port aux Basques.
Most damage and loss of life was at sea. From Perce one fishing company wrote to the Meteorological Service of Canada (MSC Archives, 1900) that they lost 30 boats and six men.
In New Brunswick eight small fishing schooners of the Gloucester County fleet, the Anglesea, Emma, Fly, Frances, Garfield, Hibernia, Nellie and Penguin were lost off PEI. The local Courrier des Provinces Maritimes listed 38 Acadian crew members drowned. Two other fishing schooners, from Caraquet, the Japan and Midnight, were believed to have foundered..
Further south an exceptionally high tide drove up the Petticodiac River. A chance alignment of the earth, sun and moon meant that a naturally high tide reinforced the Bay of Fundy storm surge. The moon and sun were on opposite sides of the earth so the gravitational effect on the tide was not as large as possible, but it was, nevertheless, the highest tide since the locally renowned Saxby Gale thirty years earlier (Ruffman, 1999). In that case too a storm surge reinforced a naturally high astronomical tide .
Prince Edward Island was more fortunate than New Brunswick. Only one Island vessel, the 35 ton Reality, was wrecked, at Cascumpec Bay (Alberton). The crew of four drowned. Island newspapers reported other vessels ashore, likely some of those from New Brunswick. The catalogue of losses onshore included: at Seacow Head a lobster factory totally demolished; at McDonald Point 600 lobster traps washed away; at Wilmot a large barn and windmill blown down. Fruit was torn from the trees in Island orchards.
Although the Halifax Morning Chronicle described it as the worst storm in Nova Scotia since the 1873 August Gale, another hurricane, there was only minor damage in Halifax fences downed and slates blown off roofs. The only Nova Scotia vessel lost, with no deaths, was the 237 ton Clyde, at Margaree, Cape Breton.
The storm crossed to Newfoundland, not then part of Canada, from Corner Brook to Gander. News trickled in. Not until October 4th could Newfoundlands Western Star newspaper confirm 82 schooners, with displacements up to 67 tons, ashore or foundered, and another 100 seriously damaged. At least 50 lives were reported lost with hope for 25 others was almost abandoned.
Some of the larger storm total precipitation measured by the sparse network of Newfoundland stations included: 86 mm at Point Riche (50 o 42N, 5 7 o 25W), 70 mm at Point Amour (Labrador, 51 o 28N, 56o 51W), and 43 mm at St. Johns. Channel (Port aux Basques) received 38 mm and Cape Norman (51 o 38N 55 o 54W) 29 mm.
The fishing fleet of St. Pierre et Miquelon was decimated (Sasco, 1970). Nine schooners and 120 men were reported lost leaving 50 children without fathers. One of the capsized schooners, the "Ali Baba", was towed into port at St. Pierre with thirteen bodies inside so bloated that only one was recognizable.
Warnings in Canada
To its credit Canadas Meteorological Service issued warnings for this storm from the Lower Lakes to the Atlantic, although only a few hours in advance. Weather forecasters had very little information to work with twice daily telegraphed weather observations from 34 Canadian and 60 US centres. And they had to work quickly. Useful forecasts could only be made for about a day in advance so time was critical. Within 90 minutes of receiving the observations forecasts were flowing, or rather should have been flowing, back out by telegraph to coastal signal stations, newspapers and community offices.
A vexing problem was how to inform ships at sea. September 1900 was before radio still a year before Marconis first reception of a trans-Atlantic signal. The warning system was a network of coastal stations that hoisted various combinations of cones and drums atop a mast to signal a storm. Orders to hoist the signals were telegraphed from Toronto. It was an effective system too: in a report to Parliament the harbour master at Tignish, PEI, described a case in 1899 where many fishermens lives had been saved by the signal.
That September even the storm signals failed. On September 14th the Acting Director of the Meteorological Service wrote to the Superintendent of the Great North Western Telegraph Company complaining that although warning messages had been sent out at 10:30 p.m., and should have been received no later than 8 a. m., delivery was recorded as: Digby - 10:30 a.m.; Yarmouth - 11:30 a.m.; Liscomb - 4:40 p.m.; Tignish - 3:00 p.m.; St. Andrews - 11:10 a.m. The warning seems never to have reached Northern New Brunswick and Gaspé communities.
Despite these problems it is clear the warning system was generally considered a success. It was after this storm that Newfoundland commercial interests expressed interest in having Canada provide weather service. In a letter to his US counterpart, the Director of the MSC, Frederick Stupart, wrote on 31 December that "our Canadian forecast and storm signal system will very shortly be extended throughout Newfoundland."
A detailed survey of newspapers and government reports identified 86 Canadian (including Newfoundland) fatalities from the storm. Thats a conservative estimate. These were people who were either identified by name, or where the number of crew of a vessel clearly stated to be lost were given.
Many other people likely fell victim. A letter to the head of the Meteorological Service from Perce, mentioned above, lists six fatalities but without additional information one cannot be sure they are not accounted for elsewhere. There are numerous similar ambiguous cases for newspaper reports in the Maritimes.
The Newfoundland reports are particularly troublesome. Newspapers report vessels missing or unaccounted for in one issue, but fail to resolve the question in subsequent editions.
Extra-tropical Storm Transition
Lacking any information on the three-dimensional structure of this storm, this was well before upper air soundings were systematically taken, and having no access to the synoptic data, one can do little more than speculate on the storms apparent redevelopment on the East Coast. The collapse of the bipolar structure evident on the surface map for 8 p.m. on the 11th of September suggests the remnant tropical system gradually came into phase with a wave in the Westerlies proceeding it. Perhaps a more detailed analysis of the type conducted by Abraham et al, (1999) on the 1869 Saxby Gale would show this to be a similar type of transition.
I thank the staff of the National Library of Canada (Ottawa), National Meteorological Library (Downsview), and Meteorological Service of Canada (Climate Services, Downsview) for assistance with contemporary information sources.
Abraham, Jim, George Parkes and Peter Boyer, 1999. The Transition of the "Saxby Gale" into an Extratropical Storm [Extended Abstract]. 23rd Conference on Hurricanes and Tropical Meteorology, Committee on Tropical Meteorology and Tropical Cyclones, American Meteorological Society, 79th Annual Meeting, January 10-15, Dallas, Texas, Preprints, Vol II, Paper 11A.3, Thursday, January 14, pp. 795-798.
Cline, I., 1900. Special Report on the Galveston Hurricane of September 8, 1900, Monthly Weather Review, September 1900. Also at: http://www.history.noaa.gov/tales/weather/cline2.html
Larson. E., 1999. Isaacs Storm, Random House Publishers, 323p, ISBN 0-609-60322-0
MSC Archives. Letter Books of Directors Correspondence for 1900. Located at MSC Downsview Offices.
Ruffman, A., 1999. A Multi-disciplinary and Inter-scientific Study of the Saxby Gale: an October 4-5, 1869 hybrid hurricane and record storm surge. CMOS Bulletin SCMO, Vol 27, No. 3, June 1999, pp 67 - 73
Sasco, E. and Joseph Lehuenen, 1970. Ephémérides des Iles Saint-Pierre et Miquelon, Imprimerie du gouvernement, Saint-Pierre et Miquelon, 1970. Also at: http://www.geocities.com/Athens/Atrium/1385/spm.html
© John D. Reid, 2000 - 2001