Arctic Tides

001      |      Vol_VII-0613                                                                                                                  
EA-Oceanography

(Jonas Ekman Fjeldstad)

ARCTIC TIDES

        The first attempt to give a survey of the tides in the arctic regions

was made by Rollin A. Harris in his “Arctic Tides” (1911). But the available

tidal information was at that time very meager, and the results naturally

uncertain, since no Observations from the Siberian coast were known. The pat–

tern of cotidal lines which he drew for the Arctic Sea has not been confirmed

by later observations.

        Important tidal information was gathered by the Norwegian North Polar

Expedition with the Maud , 1918-1925. The scientific leader of the expedition,

Dr. H. U. Sverdrup, made tidal observations at Cape Chelyuskin, Aion Island,

Cape Serdts Kamen, and Four Pillar Island, and besides these observations the

expedition also carried out numerous observations of tidal currents, and some

attempts were also made to determine tidal range on the shelf by means of

hourly soundings.

        The observations from Cape Chelyuskin and Aion Island showed conclusively

that the cotidal map of Harris could not be correct. A new system of cotidal

lines was constructed by the present writer in 1923, and H. U. Sverdrup also

drew a system of cotidal lines for the shelf. Still the tidal information

from the arctic regions is very incomplete, and it is not safe to draw cotidal

lines across the Polar Sea. Some details of the propagation, however, may

be given.

002      |      Vol_VII-0614                                                                                                                  
EA-Oc. Fjeldstad: Arctic Tides

       

Semidiurnal Tides in the North Polar Sea

        The semidiurnal tide-producing potential vanishes at the Pole and obtains

only small values in the whole Polar Basin, for which reason the tides orig–

inating in the Arctic Sea will be small and almost negligible. The observed

semidiurnal tides, therefore, must be due to waves propagated from adjacent

waters. Since Bering Strait is too narrow to be of importance, the tides

from the North Pacific will not be perceptible except in the vicinity of Bering

Strait and the semidiurnal tides in the Polar Sea must be closely related to

the tides of the North Atlantic.

        The North Atlantic and the Polar Sea are in communication through Baffin

Bay, through the wide opening between Greenland and Spitsbergen, and through

the Barents Sea. The tidal wave which propagates through Baffin Bay will

possibly control the tides in the Canadian Arctic Archipelago, but is not likely

to be of much importance in the Polar Sea itself. The Barents Sea is shallow

and almost cut off from the Polar Basin by the shallow banks which extend from

Spitsbergen to Franz Josef Land and on to Novaya Zemlya. Furthermore the de–

flecting force of the earth’s rotation tends to concentrate the tidal wave

along the northern coast of Russia. The wave that passes between Novaya Zemlya

and Franz Josef Land will, therefore, be unimportant, as is seen by comparing

the tides at Cape Flora and Teplitz Bay. The tidal hours are 5.97 h. and 2.06 h.,

respectively. The tidal hour at Cape Flora indicates that the tide at this

place is connected with the branch of the tidal wave which passes across the

Bar e nts Sea. This is seen if we compare the tidal hour at Cape Flora with the

tidal hours at places in Novaya Zemlya, where at Byellushya Bay we find 5.74 h.

and at Krestovaya Bay 5.53 h. The tidal wave at Teplitz Bay, on the other hand,

003      |      Vol_VII-0615                                                                                                                  
EA-Oc. Fjeldatad: Arctic Tides

is evidently connected with the tide on the north coast of Spitsbergen.

        The deep and comparatively wide opening between Greenland and Spits–

bergen is the chief communication between the Atlantic and the Polar Sea.

It is, therefore, certain that the semidiurnal tide which is observed in

the Polar Sea is almost entirely derived from the tidal wave which is

propagated through this opening.

        The propagation of the semidiurnal tidal wave can be seen by a com–

parison of the tidal hours at different places. The tidal hour means

Greenwich time for high water of the component in question. It is found

by dividing the phase angle X by 30 and adding the West Longitude of the

place expressed in time. In Table 1 we have collected the tidal hours of

the components M ₂ and S ₂ together with the phase difference S ° ₂ - M ° ₂ . The

table is divided into four different sections. In the first part of the

table we follow the increasing values of the tidal hours from the northern

coast of Spitsbergen to Maud Harbour, Cape Chelyuskin. In the second part

are represented the tidal data belonging to the Kara Sea branch of the tidal

wave. The tidal hours seem to indicate that the tidal wave here comes from

the north. The third part of the table shows the propagation of the Norden–

skiöld (Laptev) Sea branch of the tide to the Lena Delta.

004      |      Vol_VII-0616                                                                                                                  
EA-Oc. Fjeldstad: Arctic Tides

Table 1

Station	M 2	M ° 2	S 2	S ° 2	THM 2	THS 2	S ° 2 -M ° 2
Port Virgo	41.4	38°	14.3	70°	0.55 h.	1.6 h.	32°
Mosel Bay	35.0	87	13.1	121	1.82	3.0	34
Treurenberg Bay	28.0	99	10.7	150	2.16	4.9	51
Sveanor	25.1	92	11.1	138	1.85	3.4	46
Brandy Bay	28.0	83	10.0	135	1.47	3.2	52
Teplitz Bay	15.5	178	6.4	229	2.06	3.8	51
Maud Harbor	12.6	324	5.0	3	3.76	4.1	39
Zarya Harbor	17.7	26	8.2	99	6.53	9.0	63
Port Dickson					8.55
Nerpalakh Lagoon	7.3	25	4.6	91	3.90	5.9	66
Sagastyr					9.30
Bykhov	7.1	316	4.2	14	1.90	3.9	58
Tiksi Bay	11.8	15	8.9	65	3.90	5.6	50
Dashka Island	3.0	55	2.1	102	5.30	6.8	47
Bennett Island	35.0	137			6.61
Flaxman Island	6.7	354	2.7	37	9.53	11.6	43
Point Barrow	5.2	336	2.1	16	9.65	11.0	40
Aion Island	1.7	347	0.9	35	0.33	2.0	48
Four Pillar Island	1.0	60	0.5	110	3.17	4.8	50
Pitlekai	2.5	184	1.0	240	5.71	7.6	56
Cape Serdtse Kamen	3.9	221	1.5	307	7.81	9.7	77

005      |      Vol_VII-0617                                                                                                                  
EA-Oc. Fjeldstad: Arctic Tides

        In the last part of the table we follow the increasing values of the tidal

hour from Bennett Island to the northern coast of Alaska and across the

Siberian shelf to Aion Island and Four Pillar Island. We see that the wave

is propagated very rapidly across the deep Polar Basin from Spitsbergen to

Cape Chelyuskin so that the time interval between high water at Port Virgo

and Maud Harbor is 3.2 hours. The distance between the two places is 1,965 km.,

giving an average velocity of propagation of 165 meters per sec. If we

compare this with the simple formula c = √(gh) we find that the velocity cor–

responds to a mean depth of 2,770 m. In reality the depths must exceed 3,000 m.

Such depths were also found by Nansen.

        The difference between the tidal [ ?] hours of Port Virgo and Point Barrow

is 9 hours. This is more than we should expect, for the distance between Port

Virgo and Point Barrow is about 3,200 km., and the time of propagation would

be only 6 hours if the wave were propagated across an uninterrupted deep polar

basin with a depth of about 3,000 m. This point was one of Rollin A. Harris’

strongest arguments for the assumption of a large tract of land in the Polar Sea.

        The time difference between Port Virgo and Aion Island is 11.8 hours. At

Four Pillar Island high water takes place 2.8 hours later than at Aion Island.

These tidal hours are best explained if we assume that the tidal wave is

propagated across the shelf in a southwesterly direction.

        On the other side of the polar basin along the Canadian Arctic Archipelago

few observations of the tide have been made.

        From the northern coast of Greenland and Grant Land we have some observa–

tions made by Peary’s expedition. The tidal hours vary, with apparent ir–

regularity, probably indicating the influence of the tidal wave from Baffin Bay

006      |      Vol_VII-0618                                                                                                                  
EA-Oc. Fjeldstad: Arctic Tides

through Kane Basin and Robeson Channel. The two stations lying farthest away

from the entrance to Robeson Channel are Cape Morris Jesup, on the northern

coast of Greenland, with the tidal hour 0.68 h., and Cape Aldrich, on Grant

Land, with the tidal hour 0.31 h.

        The Canadian Arctic Expedition under the leadership of V. Stefansson, 1913-1918,

observed the tide at several points along the coast of the continent eastward

frost Alaska to Amundsen Gulf, and also on the open shores of the Polar Sea on

Banks Island and in the vicinity of Ellef Ringnes Island (3). The results are:

Collinson Point: spring range, 0.60 ft.; neap range, 0.20 ft.; establishment

of port, 0 h.18 m.. Demarcation Point: spring range, 0.72 ft.; establishment,

12 h.14 m.. The tidal hours are 9.95 h. and 9.23 h., respectively. The ob–

ervations at Cape Kellett gave as results: spring range, 0.43 ft.; neap range,

0.40 ft.. The establishment could be determined only approximately and the

result was 9 h.50 m.. If this be correct, the tidal hour is 5.63 h.

        The propagation of the tide on that part of the Siberian shelf to the

north of the New Siberian Islands can be followed by the results of the hourly

soundings. In Table 2 we have arranged the results according to increasing

values of east longitude.

Table 2

Lat.			M 2	M ° 2	TH	Remarks
I	76°39	139°40	50 cm	31°	3.7h	July 23-25, 1924
II	76 16	142 15	51	58	4.4	July 12-16, 1942
III	75 46	143 25	(55)	-	(5.0)	Fade e vski Isl. Aug. 13, 1924
IV	76 30	143 58	49.7	89	5.4	July 2-3, 1924
V	76 18	144 08	72	95	5.6	July 6-8, 1924
VI	76 41	149 05	35	137	6.61	Bennett Isl., 1881
VII	74 40	166 29	11	159	6.2	May 1-2, 1923
VIII	74 26	167 45	13	153	5.9	April 25-26, 1923

007      |      Vol_VII-0619                                                                                                                  
EA-Oc. Fjeldstad: Arctic Tides

        In this ta lb bl e are also included the results from Bennett Island and the result

of an observation from Fadeevski Island, which serve as a check on the results

of the hourly soundings. We remark that the results from station I, II, and IV

agree completely in regard to the amplitudes, and that the result from Fadeevski

Island also confirms these results, which give larger tidal ranges than those

found elsewhere in the Polar Sea. Station V gives a still greater amplitude,

but this may possibly be erroneous, and to some degree be a result of the

ship’s motion above a sloping bottom.

        The table indicates that the tidal hours change rapidly in this region.

This fact is easily explained when we take into account that the depth on

this part of the shelf is only some 20 meters.

        The direction of propagation in this region is about S. 80° E., as indi–

cated by the results of current measurements.

        In the region to the east of Bennett Island the direction of propagation

is more southerly, and across the shelf the direction must be toward the south–

west.

        In the East Siberian Sea the direction of propagation must also be toward

the southwest. The tidal wave is here stron g ly affected by frictional resistance

and the amplitude in reduced from 12 cm. at the edge of the shelf to about 2 cm.

at Aion and only 1 cm. at Four Pillar Island. The time necessary to cover the

distance across the shelf to Aion Island is 6 hours. The distance is about

530 km. and so the velocity of propagation will be about 24 m. per sec..

       

Diurnal tides

        In Table 3 we have compiled all the available harmonic constants relating

to the components K ₁ and O ₁ in the arctic regions. In the first column of the

008      |      Vol_VII-0620                                                                                                                  
EA-Oc. Fjeldstad: Arctic Tides

table is given the western longitude of the station expressed in time. The

first part of the table gives the harmonic constants for places along the

eastern side from the northern part of Norway, Svalbard, and Siberia to

Alaska. The other part of the table gives the harmonic constants for some

places along the coast of Greenland and Grant land.

Table 3

Station	Long.	K 1	K ° 1	0 1	0 ° 1	THK 1	THO 1
Bodø	23.04	10.3	208°	3.9	32°	12.91 h.	1.17 h.
Kabelvåg	23.03	10.5	212	3.9	54	13.16	2.63
Bear Island	22.72	5.4	230	4.4	66	14.05	3.12
Port Virgo	23.29	2.7	225	1.2	12	13.62	0.08
Mosel Bay	22.93	7.0	245	2.7	72	15.26	3.73
Treurenberg Bay	22.88	7.3	270	2.1	70	16.88	3.43
Sveanor	22.78	8.7	271	1.2	78	16.85	3.98
Teplitz Bay	20.14	3.0	26	1.2	49	21.85	23.42
Zarya Harbor	17.66	3.6	3	2.4	8	17.86	18.19
Maud Harbor	16.97	4.8	24	2.0	4	18.16	17.25
Nerpalakh Lagoon	14.86	0.6	293	0.3	350	10.39	14.19
Four Pillar Isl.	13.17	0.6	12	0.4	76	13.97	18.24
Aion Island	12.81	0.4	355	0.4	12	12.48	13.61
Pitlekai	11 . 57	1.5	53	1.2	63	15.10	15.79
Cape Serdtse Kamen	11.44	1.5	77	1.3	64	16.57	15.70
Point Barrow	10.44	1.5	347	1.5	20	9.55	11.76
Flaxman Island	9.77	2.4	7	2.7	45	10.25	12.72

009      |      Vol_VII-0621                                                                                                                  
EA-Oc. Fjeldstad: Arctic Tides

Station	Long.	K 1	K° 1	0 1	0° 1	THK 1	THO 1
Jan Mayen	0.56	3.4	97	6.1	49	7.00	3.85
Denmark Isl.	1.75	8.7	65	9.3	35	6.08	4.08
Little Finsch Isl.	1.41	11.8	75	7.6	25	6.41	3.08
Cape M. Jesup	2.24	7.6	340	5.7	312	0.93	23.03
Cape Bryant	3.70	9.8	286	4.3	262	22.77	21.15
C. Sheridan	4.08	4.9	298	2.7	278	23.95	22.63
Polaris Bay	4.12	12.2	246	4.6	209	20.51	18.04
Fort Conger	4.32	8.5	222	5.8	199	19.12	17.59
Cape Aldrich	4.67	5.2	311	3.4	281	1.36	23.36

        If we fix our attention on the component 0 ₁ we see that the phase is

everywhere less than 90° for the stations along the coast from Alaska to

Svalbard. The tidal hour increases from about 12 hours at Point Barrow to

18 hours at Maud Harbor, and 23.5 hours at Teplitz Bay. [Pitlekai and Cape

Serdts Kamen are situated near Bering Strait, and the tide here is possibly

influenced by the tidal wave entering the Polar Basin from the North Pacific.

These places are, therefore, not taken into account in the following discussion.]

        The difference of approximately 12 hours between the tidal hours of Point

Barrow and Svalbard indicates that the wave is mainly generated in the Polar

Basin itself and that it is a somewhat retarded equilibrium tide. In a circum–

polar basin like the Arctic Sea the ideal tidal hours will differ little from

the western longitude of the place expressed in time. From the table we see

010      |      Vol_VII-0622                                                                                                                  
EA-Oc. Fjeldstad: Arctic Tides

that the tidal hours and the longitude run nearly parallel for 0 ₁ .

        The great value of the phase angle 76° which we find at Four Pillar

Island is explained by the location of the island. The sea here is very shal–

low and, owing to frictional resistance to the currents, the wave will probably

b e of a progressive character across the shelf. The difference between the

tidal hours at Aion Island and Four Pillar Island is too great, but we have

to bear in mind that the tide here has a very swell amplitude so that the

results are probably affected by relatively large erro r s.

        The tidal hour for the stations on the northern coast of Greenland and

Grant land show the influence of the tide wave from Baffin Bay, which enters

the Polar Basin through Kane Basin and Robeson Channel.

        For the component K ₁ the tidal hour and the longitude correspond fairly

well along the coast from Point Harrow to Teplitz Bay, but from there to Port

Virgo we again find decreasing values of the tidal hour. Thus we find for

Treurenberg Bay 16.88, Mosel Bay 15.26 and Port Virgo 13.62.

        These tidal hours are closely connected with the tidal hours in the Nor–

wegian Sea. The diurnal wave K ₁ enters the Norwegian Sea through the passage

between Scotland and Iceland, and another branch enters through Denmark Strait

between Iceland and Greenland. The observations are too few to trace the

propagation of these two waves in detail, but the principal progress of the

waves can be seen. The tidal hour for K ₁ at Thorshavn, Faroe Islands, is

9.5 h., at Bergen we find 10.98 h., at Bodø 12.90 h., and Port Virgo 13.62

hours, These figures indicate that the wave is mainly of a progressive char–

acter. The other branch of the wave may be traced at Denmark Island, Little

Finsch Island, and Jan Mayen, where the tidal hours are 6.08, 6.41, and 7.00,

respectively.

011      |      Vol_VII-0623                                                                                                                  
EA-Oc. Fjeldstad: Arctic Tides

        These two waves must Interfere in the strait between Svalbard and Green–

land, and so the tidal hours at Port Virgo, and possibly also at Cape Morris

Jesup, are influenced by this. The tidal hour at Cape Morris Jesup may also

be influenced by the wave which enters through Robeson Channel.

        Tidal Currents.

        The shelf region north of Siberia presents unique conditions for the

investigation of tidal currents. In the winter time the rigid ice cover is

only slightly affected by the tidal motion, and there are no disturbances by

surface waves. The Maud was in possession of a number of Ekman current meters,

but these instruments were not well suited for use in the Arctic during the

very low winter air temperatures. Dr. Sverdrup, with the assistance of the

aviator, Odd Dahl, succeeded in constructing an electric registering current

meter, which could be left in the sea for weeks at a time, and which regis–

teredd the velocities and the direction of the currents on deck. This current

meter was put in operation in the spring of 1923, and during spring and summer

they obtained several long series of current measurements from a fixed depth.

        The hydrographic conditions on this part of the shelf are characterized

by nearly homogenous water from the surface down to a depth of same 30 meters.

Then comes a transition layer with rapidly increasing densities, and a bottom

layer with more homogeneous water. The frictional resistance caused by the

ice cover reduces the velocities in the upper layer, so that the greatest

current velocities ordinarily are found in the transition layer. Near the

bottom the velocities are again reduced by frictional influence.

        The current meter was operated in depths where from experience the greatest

velocities were to be expected.

012      |      Vol_VII-0624                                                                                                                  
EA-Oc. Fjeldstad: Arctic Tides

        In Table 4 we give the results of some of the measurements during the

spring and summer 1923.

Table 4

Series	Lat. N.	Long. E.	V 1	V 2	T	X	TH
1	74°12′	168°50′	16.4	−14.8	260°	−75°	9.4
2	74 20	168 33	14.6	−13.0	243	−83	8.9
3	74 24	168 12	9.9	−8.0	241	−66	8.8
4	74 40	166 24	18.6	−14.3	236	−94	8.8
5	75 24	166 00	8.7	−8.3	243	−78	9.0

        In this table, V ₁ is the maximum velocity and V ₂ the minimum velocity of

the component M ₂ . T is the phase angle and X the direction of the maximum

current. (−75° means S. 75° W.)

        The results are naturally affected by relatively large errors with respect

to the angles T and X since the current ellipses are nearly circular.

        At spring tide the velocities reach values of some 30 cm./sec.

        In the summer of 1924 the Maud vas drifting with the ice along that part

of the shelf which lies to the north of the Hew Siberian Islands. Here the

tidal currents are considerably stronger and are able to break up the ice into

smaller ice floes which to some degree are carried along with the tidal cur–

rents. Measurements of the ice drift indicated the existence of tidal currents

which sometimes amount to more than 40 cm./sec.

        The current measurements which were made during the summer of 1924 were

combined with measurements of the ice drift, and thus it was possible to find

the distribution of the velocities with depth.

013      |      Vol_VII-0625                                                                                                                  
EA-Qc. Fjeldstad: Arctic Tides

        We give here the results of three such series of current measurements

(Table 5). On June 30, 1924, at longitude 143°53′ E., latitude 76°34′ N.,

the depth of the sea was 36 meters and the Ekman current meter was working

at depths of 15, 23, and 31 meters. The measurements were continued during

one tidal period. It is seen that the velocities show a decrease toward the

bottom, indicating the influence of friction.

Table 5

Depth	V 1	V 2	T	X	TH
0	17.2	−8.8	53°	−6°	4.2
15	14.4	−9.0	42	−23	3.8
23	14.5	−7.2	71	31	4.8
31	10.1	−0.2	72	44	4.8

        Observations made on July 23, 1924, at longitude 139°40′ E., latitude

76°39′ W., provide one of the few instances when the surface drift of the ice

and the surface elevation have been determined simultaneously. Thus it is

possible to obtain a comparison between the observed amplitude and the ampli–

tude which can be computed theoretically from the observed velocity.

Table 6

Depth	V 1	V 2
0	26.5	−15.9	−5°	−29°
12	24.1	−14.2	−7	−27
20	14.3	−8.1	−12	−22

014      |      Vol_VII-0626                                                                                                                  
EA-Oc. Fjeldstad: Arctic Tides

        The computed velocities agree fairly well with the observed values.

The computed angles can not be directly compared with the observed values,

since the angle in the theoretical solution is reckoned from the time of

high water, and the orientation of the current ellipse is reckoned from the

direction of propagation of the tidal wave.

        The influence of the stratification of the sea on the tidal currents

is demonstrated by a series of current measurements in the region to the

east of Bennett Island in May 1923. The current meter was shifted at sev–

eral depths and the currents thus determined. It was found that the

velocities were strong at a depth of 40 meters, but decreased rapidly toward

the surface and toward the bottom.

        Sverdrup explained this distribution of velocity by assuming that the

eddy viscosity was great in the upper layer, where the water was homogeneous,

and also in the bottom layer, but that the viscosity nearly vanished in the

transition layer, where the stability is great. But the stratification is

also favorable for the formation of internal waves, which might to some degree

be responsible for the velocity distribution. The conditions for studying

the dynamics of tidal waves and currents are very favorable in the shelf

region of the Arctic, and one might wish that new arctic expeditions would

take up such investigations.

       

Jonas Ekman Fjeldstad