Small Publications in Historical Ceophysies No.9 An Investigation of the Tidal Conditions at the Loss of the World's Most Impressive Sailing Ship ‘Martin Ekman Summor Institute lor Historical Geophysics Aland Islands 2002 Small Publications in Historical Geophysics No.9 An Investigation of the Tidal Conditions at the Loss of the World's Most Impressive Sailing Ship Mautin Ekman Contents 1. Alma sailing shipping and the Llerzogin Cecilie 2. ‘Lhe loss of the Lerzogin Cecilie 3, Character of the tides at the place of loss 4. ‘The tides at the event of the loss 5. Conetusion References Summer Institute for | istorical Geophysics Aland Islands 2002 1, Aland sailing shipping and the Ilerzogin Cecilie Scafaring has Inng traditions on the Mand Idands, the autonomous grap of islands in the central Baltic Sea. This tradition extends not only within the Tallic, bul also internationally. Since 1865 Alundic merchant ships have crossed the workl's oceuns. During the first hulf of the 20th century Gustaf Lirikson at Matieharan, the capital of Aland, built up a lect of greal deep-sca saiEng ships, mainly for the wheat trade between Europe and Australia, This put Aland in the remarkable position of having, Ihe world’s largest fleet of inerchanl sailing ships; see furthec KShre (1948), Cireenbill & blackman (1986) and Kahre & Kéhre (1988), ‘this was, according to Geeenbill & Jackman, “made possible by the nature and history uf the pecaliar community in which they were owned, « Sweuish-spealsing archipelago society which averaged 23,000 Inhabitants spreacl over some fitty islands, a distinctive minority group [inside Finland}! The sailing vessely not only brought money into the Aland economy but also were a part of the Alandic culture. In audition, these ships attracteat a lat of international attention because of their impressive beauty. ‘The most inypressive and hewutitul of them all was the Llerzogin Cecilie (‘Duchess Cecilie", the flagship of the Aland sailing fleet The Llerzogin Cecilie was bought by Gustaf Eriksen in 1921 from the Krench state, France had, after the ond of the frst world war, ablained the vessel ay a war Teperalions payment [om Germany. Germany had originally built the vessel in 1902 as an imperial ship, named afler the empress-in-be, the Duchess Cecilic of Mocklenbng, Tt-was a fomr-mastad barque, 102 m long and 60 m high, sith 35 salle; sce Figure 4. The totsl area of the sails amounted to 4050 m?. Gusla? Frikeon had this magnificent ship repaired and put itin on the ‘wheat trade lo Australia; see Dethy (1937) and Greenbill & Hackman (1991) for details. ‘The ship usually travelled, as the other vawsels om the wheat trace, from. Aland across the Atlantic Ocean and the Indian Ocean lo Australia, Tacre wheat was loaded. Than the ship croseed the Pacific Ocean and the AHantic Ocean to go to England for discharge, The different routes were chosen to make use of the dominating winds aul wea currents as z1uubt as pussibie. After that there was often a possibility for a brief visit home to Aland, until it ‘was time for the ext freight. The whole voyage around the globe normally took clase to ome year Figure 2, Herogin Cecilie (Irom Villiers, 1972) 2, The loss of the Herzogin Cecilie In 1986 the Herzogin Cecilie was lost by stranding itt the English Channel. Lhe vessel had almost completed that year's voyage tv and from Australia, ‘Lhe load consisted of nearly 5000 tons of bagzed wheat Mester an board was sea captain Sven Frilesson, and officer on watch was chiaf male Flis, Karlsson. The crew consisted of 0 man. The lass haa been exlensively reported and discussed by Derby (1937) and Greenhill é Hackman (1994), 2n the latter book with maritime inquities included. The captain's wife, 20 was ont board during the whole voyage, as well as the mate have written books about the voyage (Lriksson, 1958: Karlsson, 1964), The ship, arrivirg from Australia, had first reached Falmouth at. the sonth coast af Fngland. These arders were given to go ta Tpswich op the cast coast. Cansoquentiy one continued castwarels in the English Channel, now in Joggy weather. “finr some eight hours, at (8 $5 on Saturday night the 25th of Aptil, the shig went ashore int thick fog. ‘Lhis huppetted at the cock of Hausa Stone west Gf the peninoula of Bolt Lend, neur Scleombe southeast of Plymouth; see Ligure 2. Lalf an hour later the ship yeas taken afloat by the rising tide and Urifted further towards fara, going ashore un the souks there instend, Distresy rockets were fired, and che crew was taken ashore by a fifebvat. ‘The vessel was gradually filled by water, and later on, in spite of attempts to bring aboutsulving, it was broken in a storm: The event caused considerable excitement, especially of course on Aiand and in Fngland. Fngiish people came in enormous crowds to see the famous sailing ship im ils Gragic position. Trom che ship one suceeded in salving, among other things, Ihe captain's saloon including furniture, which was fransparied to Marichamn and nowadays form a part of she Aland Masitime ‘Musoum. The captain and the mate ceased sailing afer fe accident and later con elt Aland to settle in Attica, Lhe cause of the loss has not been fully understood. ‘lowever, it can be said to he a combination of overoptimistic passage planning, some unknown navigational error, and thick fog. Cme relied in this case entiraly on dead reckoning, wilhou: con(gol possibilities. a method containing risks for unfavourable osror propagation in the positianing, This bad the consequence ‘that the ship, instead of going far enough out in the Linglish Channel, went ov war the coast. While the ship actuafly should have passed far south of the Peninsula of Bolt Head, il wenl ashore approximately 1/2 raulical mile north ‘west of the southern lip of the peninsula. Tic crew dic not know Fron where they were. % {gure 2, The place oF ase in the Taiglish Channel (at the arrow) Loose speculations on the tides having contributed to the loss were put forware. alteady at the maritime inquiries in Plymouth a few weckes afterwarus, buth by the caplain and the mate. 4 vessel that gocs close to land in a tidal region like the Fingl'sh Channel tates the risk of being influenced by coastal tidal streams differing from the general lidal streame further out. ‘Computations that may shed Light on the role af tho coastal tides at the loss of he ship never scem io have been made, We will here try to perform such computations. 3 Character of the tides at the place of loss the tides in the Fnglish Channel are, as in dhe North Sea region in general, pronounced semi-diurnal. This means that the tides are dominated by the Lumar prinicipal tide Mz with the period 12.4 hours and the selae principal (ide $2 with the period 12.0 huss; in addition the Junar elliptic tide Ne is impartan|. The dinrnal main components, Ki, Ovand (1, are vey sina, To the Ling Channel, as partly also in the Irish Sea, there occurs a well-known resonance phenomenon in the tides, producing considerable eigenosuillations of the water mass (see e.g. Pugh, 1987). ‘The resonance period is dependent cm the length and the depth of the water basin; in the Channel these quantities yield a resonance peried close to the periods of the somi- diurnal tides, ‘hws there ere large semi-diurnal tidal variations al lhe onds of the Channel. At Plymouth near the weslern end, close fo the place of the loss, the spring tidal renge may amount ta 6 meters. ‘The tides ave mathematically characterized by their harmonic constants, i.e. by the amplitudes ar and phase lags af the tidal components. The bay of toss is, in this kespoct, probably best represented by an average of Plymusth (Pevonport) and River Yealnn Intrance, the principal harmonic constants of which then became: Mz c=165m, g= 135° & 0.60.1m, XN Ki 123° a Qorm, ge 2 PL @4003m, Llamnenic constanls for Saleambe, en the alher side of Boll Head, have been, ued for check calculations, Ih shallow waters nvar the coast of Ghe Boglish Channel - where the Doight of the Lidal wave is nv longer negligible jn refation in the depth of the ‘water -a spacial tidal effect ucurs, Ht pruduces quarter-diternal ticles, the saint ones of which ate Knewr: as My, MSi and MIN, ‘these have periods uf a little more thatt 6 hours, half of thet of the semi-tiurnal pexiod, Lhe ¢pearterdiemmat tides are nott-linear "overtides’, which are more ditfica‘t to deal with than the ordinary tidal componenis. They were closely investigated by haber! dHitres & Le Provost (1970, 197%), using a scaled Aydrauite madel of the Frglish Channel; this is aowadays replaced by numerical modelling, their amplitides and phase lage may be given as (frum the POI. data bank) Ma n= 018m, g- 14 MS 6=009m, y= 197° MING ar andm, = 112° Their norclineat character makes them sumetimes nore important then & quick glance at the amplitudes may reveal; at spring tides the cjuatterdimmal tidal variation in the vicinily of the place of loss may amount to half a meter. Linis group of tidal components lurns out to play @ noticeable role in our computations 4, The tides at the event of the loss In comection with the tidal oscillations of the water mass in the Fingtish Channel there occurs @ horizonlal transport of water. ‘Lhis results in the welle known tidal stream going allernalely westwards and castwards, wilh some modifications depending on the shape af the coast. The specd of the lidel stream is in general a few knots. Already Derby (1937) has shown tha this lidat stream could not to any considerable dogee have contributed to the sbip Teaving ils planned caurse, even if some driit towards land must Rave taken place, Greenhill & Haclattan (1991) do not add anything new in that respect, ln any case, on boar the éhip one should have had knuwledge enoagh fo correct for such a tidal deviation of the course, When the ship for various reasons nevertheless had approached Bolt Head, the tidal si(astion becomes different, In the vicinity of Une coast the tidal stream lends lo follow the direution of the coast: we will rete to that later, using a numerical model, Still closer to the coast the tidal stream tends tu be directed move towards land when the tile is rising, and away from land when the tule is falling, expecially 4 the coastal area is shallows and irregular. We will treal this coastal lide now. There is no detailed information on coastat tidal stream in the area of the Juss, In utder tv obtain some knovrledge of the veriations of the tel streains there we will have to resort to a ascful simplification: With good approximation the speed of the tidal stream #s proporlional lo the rate of Ihe rise of the Ede, Thus we compute the rain af rise of the lide and use this as a Kid of moasure of the sirength all the tia’ stream, The computation has been performed in the following way. ‘Ihe height UE the tide has been calctfated for a series of times duzing the 24 howrs from. Apel 2th, 16 00, to April 25th, 16 00, in the year 1986. This time span is symmetric with respect io the haur af Cae loss, April 25th 04 00. The tidal computations have been posformed using harmonic constants as described abore, although mare constants than those listed there have heen applied, Tor every time interval between the times mentioned the rate of rise uf the tide has then beer calotlated. ‘The vesult is presented in Ligure 3, in the form of a continuous curve showing the rate of rise, in cm per minute, as a function of time, Positive rates imply a vising tide and an in-going tidal stream, negative rates imply a falling tide and an out-going tidal stream, [t can be mentivned tht in the entrances to Plymouth and Selcombe: a rate of rise of 1 em per minute corresponds lo a slam speed of abou: 1 kn. We immediately nola that the carve is somewhat asymmetric: il has a pointed maximum which also is shifted to the left. The reason for this is the group of quarler-diurnat overlides descrihed in Section 3. The maximum rate 8 rise amounis io 2.2. em per minute, which oreurs al around D410, Toro interesting questions now present themselves: How lagge la the variation during, this day compared to normally? (Where on the curve does the oss occur? The first question can he answered by comparing the con:imous cmve in the figure with the dashed one, showing the cakulaled raie of rise in the normal case (mean variation). As can be seen, the variation during, the day af the loss is considerably larger than normal. This is parlly due lo the tide being, feiiJy close to spring (Combination of Mz and §), but alse lo the moon being very close tir the perigee of its orbit (combination uf Me atte No), he snercionian is extra large, because of the influence of the quenter-diumal overtides (combination of Me MSs and MING). Tae loss occurred, as earlier mentioned, in twa steps, a ferat steonding at 08 85 and a second ane at about 04 28, "Chey ace matked with arrows in the figuee. The second queésion whove can, thereby, be answered: ‘The loss 26 ane ie 3, The vale of vise of the lide, fm re per minuke, during the day of fas (continuous curve) aad ia the normal cose (dashed curve). The arraws show when the strandings uccurrel