NOAA History Banner
gold bar divider
home - takes you to index page
about the site
noaa - takes you to the noaa home page
search this site
white divider

arrow Stories and Tales of the Coast & Geodetic Survey
arrow Ocean Tales

submarine scarp off cape menodino, california

Harold W. Murray, Associate Cartographic Engineer
U.S. Coast and Geodetic Survey

The significance of Cape Mendocino and the nearby area to navigators using these waters is a classic example of the use of submarine and shore configuration in position determination. The Pacific Coast Pilot describes the cape as a “mountainous headland” and “famous landmark of the old Spanish navigators”. The cape is a turning point for nearly all vessels bound north or south and, in view of the dangers in this vicinity, must be approached with great caution in thick weather because the currents and irregular bottom tend to make the ordinary methods of navigation uncertain. It is also in a region of great climatic change, and the meteorological conditions northward of the cape are quite different from those to the southward. Fog, for example, is more prevalent to the southward, whereas rainfall is heavier and the northwesterly winds of summer are more violent to the northward. The combined in fluence of these factors on navigation is emphasized by the number of wrecks which occurred during the earlier years of the present century when the then existing hydrographic surveys were so woefully inadequate.

topographic chart showing cape mendocino submarine canyons
The Mendocino Escarpment. The initial indication of the great seafloor
fracture zones. Surveyed by GUIDE between 1935-1938

A vessel running the courses commonly used in this locality would approach Cape Mendocino from the north on a course of 176° and would expect to pass about six miles off the cape to clear safely the known dangers. When the cape is abeam and bearing about East, the course would be changed to 160, a 16 change to the East. That is, to exaggerate, one might say that a vessel has to run out and around the cape, keeping six miles off. It is evident that if one did not run far enough or, in other words, changed course too soon, the vessel would run onto the shoals or onto the cape itself. The currents in this locality are strong and variable, some wind effects are little known, and fog is more prevalent than in most places on the coasts of the United States, there usually being approximately 1,300 hours of fog per year. A vessel passing the cape may well have been running for many hours in a fog without sight of any fixed object and be entirely dependent on dead reckoning and soundings for an approximate position.

It was during a period of foggy weather in the year 1916 that the master of the Steamer BEAR, “when his reckoning put him about 15 miles northward of Cape Mendocino, began to take soundings to locate his position” and feel his way to the lightship anchored off the cape. The soundings immediately indicated that the vessel was proceeding over deep water of 100 fathoms or more. When the depths began to shoal from 80 fathoms to 34 fathoms and subsequent soundings showed substantially deeper water, it appeared from the published chart then in use that the vessel had safely passed the cape and the course was changed as usual. About an hour later the vessel stranded two miles north of the cape wit ha loss of six lives. The contributing factor in the disaster was that the misleading soundings had been obtained in a reported but unsurveyed and, consequently, inadequately charted submarine depression (Eel Canyon) several miles northward of Cape Mendocino. ( See Coast and Geodetic Survey Special Publication No. 48 (1918), “The Neglected Waters of the Pacific Coast”, for details of this and other incidents.)

Between the years 1899 and 1917, 15 wrecks of strandings occurred in this area and an additional 50 occurred at other points along the California coast. Accidents off the coats of Oregon and Washington in this period totaled 26 and 15 respectively. In each case the lack of hydrographic surveys, insufficient knowledge of currents, and inadequate charts were the contributing not necessarily the sole factors involved.

Hydrography previously executed by the Coast and Geodetic Survey in the vicinity of Cape Mendocino, aside from earlier reconnaissance surveys, consisted principally of 1:!0,000 and 1:20,000 scale surveys accomplished between the years 1872 and 1886. These surveys, generally speaking, extended from six to eight miles offshore and included the heads of all submarine depressions terminating within the above mentioned limits. All soundings were vertical casts and probably did not exceed 20,000 in number.

It was not until the years 1919 to 1921 that the Coast and Geodetic Survey Ships WENONAH and LYDONIA, R.R Lukens and E.H. Pagenhart commanding, surveyed the offshore area of Cape Mendocino and for a distance of at least 67 statute miles. The deeper offshore soundings, totaling about 6,500, were obtained by the laborious and time-consuming vertical cast method, horizontally controlled in part by dead reckoning and by three-point fixes on shore objects. The intensity of hydrography was naturally limited by the methods used, by the length of time available, and by the funds allotted to the vessels, but was accepted as adequate for the needs of navigation and there heretofore reported but mysterious Eel Canyon was now definitely surveyed and firmly secured within the confines of a geographic projection.

In 1935 the Coast and Geodetic Survey’s plan of making modern and more intensely developed surveys, a project begun in 1932 at the southern limit of the State of California, had progressed northward to the vicinity of Cape Mendocino. The submarine topography revealed in the waters contiguous to the cape is shown by the submarine contours in the accompanying illustration.

The hydrography represented in the illustration was obtained at selected intervals during the period from 1935 to 1938 by the Coast and Geodetic Survey Ship GUIDE, F.H. Hardy, O.W. Swainson, and E.W. Eickelberg commanding. These surveys consisted of three series. One series of nine 1:10,000 and one 1:20,000 scale surveys embracing the area between the shore line and the 20-fathom curve consisted of 48,000 soundings. The second series was composed of one 1:20,000 and three 1:40,000 scale surveys extending from the 20-fathom curve to distances of three to seven statute miles offshore and consisting of about 25,00 soundings. The last series consisted of one 1:120,000 scale survey with about 7,000 soundings extending from the last mentioned limits to more than 66 miles offshore. In all, a total of more than 80,000 soundings have been taken within the area of the illustration. The soundings obtained are principally echo soundings supplemented by vertical cast and hand lead soundings in the waters adjacent to the shore line. Horizontal control consisted of three-point fix angles on shore objects in the inshore area and radio acoustic ranging in the offshore area.

The contour interval shown on the illustration is 100 fathoms (600 feet) for depths of 100 to 1,800 fathoms. In depths less than 100 fathoms the 10-, 20-, 30-, 40-, 50-, and 75-fathom contours are shown. The contour interval used in the 1:120,00 scale survey previously mentioned was 25 fathoms, or four times greater. This insured a more accurate contour delineation in areas where echo soundings were practically continuous on rather widely-spaced sounding lines.

The diversity of submarine topography expressed in the illustration is self-evident. Heading the list is the long submarine scarp one-half to one mile in height extending more than 66 statute miles from shore. The western extremity of this feature has not as yet been ascertained. Portion of the face of the scarp plunge downward to the north with a steepness of from 24 to about 100 percent. The downward slope of the top of this scarp, measured from the closed 200-fathom contour to the closed 800-fathom contour, is 1.7 per cent. However, from the western extremity of the 800-fathom contour to the western limit of the 1,000-fathom contour (outside the limit o the illustration), the rate of descent has increased to 3.9 per cent. The ocean bottom to the north and northeast of the scarp is quite flat and about one and three-fourths miles below the surface of the ocean, whereas the bottom to the south of the scarp slopes gently southwestward at a rate of 3.8 to 7.6 per cent accompanied by a depth change of from one-fifth to two miles. This scarp with the three types of contrasting topography constitutes a submarine feature too unusual to possess a known rival on the entire West Coast.

The 100-fathom contour closely approximates the limit of the continental shelf which is broader on the north than on the south. The continental slope beginning at the 100-fathom contour slopes away in the broader areas at a rate of about 4.3 per cent just above Spanish Canyon and is as great as 11 to 19 per cent on either side of Bear Seavalley where it is considerably shorter in length. Here again contrasting topography is presented in that the rate of slope on the north is about three to four times greater than that to the southward.

Submarine canyons are well entrenched on either side of the scarp and protrude several miles into the continental shelf. Mattole and Delgada canyons to the southward of the cape are remarkable in that they extend so close to shore. Eel Canyon on the north has a broad head about five miles wide with five pronounced tributaries. It traverses a distance of 32 statue miles between the 30- and 1,400-fathom contours. The bottom gradient between the 75- and 900- fathom contours, a distance of about 25 miles, is from 5.0 to 3.0 per cent. At the 900-fathom contour the bottom slopes steeply to a depth of 1,300 fathoms with gradients as great as 30.3 per cent after which it lessens to about 2.4 per cent. The submarine knoll existing near the mouth of the canyon, around which the stream channel has had to travel 11 miles, is a phenomenon in deflection of submarine canyon courses. This knoll will evoke an interesting discussion as to whether it is younger or older than the canyon, or contemporary with a portion of the canyon’s history. The fact that the mouth of the canyon including that portion of the canyon just eastward of the knoll approaches a straight line would imply, for example, that the canyon was fault-controlled and that the knoll was a subsequent intrusion occurring at some time after the canyon was well established.

Bear Seavalley is about 18 miles long between the 75- and 1,400-fathom contours. Its gradient is about 14 per cent down to the 500-fathom contour, thence 30 per cent to a depth of 1,000 fathoms after which it levels out from 11 to as low as 2.9 per cent. The name of this feature was supplied by the writer, all other names shown on the illustration being in use on the later editions of the Coast and Geodetic Survey charts of this area. The term “seavalley”, however, is a recent decision of the United States Board on Geographical Names and is applied to submarine depressions that are of valley form but unaccompanied by steep adjacent parallel walls such as are found in canyons.

Although outside the scope of this article, it is nevertheless of practical interest to note that another recent decision is “seamount”. This new term is being applied more frequently off the West Coast and is used to denote a submarine elevation of mountain form. As a specific example, the first feature to receive this designation was a submarine mountain discovered by the Coast and Geodetic Survey Ship GUIDE in 1933 about 75 miles west of Point Piedras Blancas, California. This feature rises from a depth of 1,900 fathoms to 729 fathoms and has a net elevation above the ocean floor of 1,171 fathoms or 7,026 feet. It was named “Davidson Seamount” in honor of George Davidson (1825–1911) of the U.S. Coast and Geodetic Survey.

Mendocino and Mattole canyons join at a depth of around 900 fathoms. Their lengths inshoreward from this point are 14 miles to the 40-fathom contour and 18 miles to the 10-fathom contour respectively. Two alternate outlets into the broad region of the 1,400-fathom contour are possible: one where Bear Seavalley enters, and the other about nine miles farther westward. The total lengths, in the case of the longer Mattole Canyon, to the two outlets are about 39 and 48 miles. Mendocino Canyon is more direct and has a gradient of about 6.4 per cent from a depth of 200 to 1, 100 fathoms after which it levels out to about 2.2 per cent. The gradient along the major portion of Mattole Canyon is about 5.4 per cent or slightly less. Portions of the side walls of these two canyons are similar and yet contrasting. Near the apex of the 400-fathom contour, Mendocino Canyon has a steep wall slope of about 49 per cent on the north side, whereas Mattole Canyon has its steeper side slope of about 52 per cent of the south side where the face of the scarp serves as a side wall. In the same vicinities the opposing walls of each canyon are also similar in that they have lesser slopes of 21 and 16 per cent respectively.

Spanish and Delgada canyons are only partially shown on the illustration. Spanish Canyon is fairly straight and has a gradient of about 7.2 per cent from the 30- to the 300-fathom contour after which it lessens to about 2.4 per cent. Delgada’s gradient is about 15 per cent from the 10- to the 100-fathom contour, thence 11 per cent to the 200-fathom contour after which it changes to about 2.7 per cent.

The contouring of several features represented in the illustration has revealed the desirability of additional development for further geological and seismological researches. Such additional development will necessarily be more comprehensive than that needed for purposes of navigation and will be accomplished by the Coast and Geodetic Survey Ship GUIDE, E.W. Eickelberg commanding.

The new development will consist briefly of:

1. A zigzag echo sounding line extending along the fault scarp to ascertain its present unknown western extremity.

2. A system of short, closely-spaced, vertical cast sounding lines crossing the submarine scarp approximately normal to the depth curves between the junction of Mattole and Mendocino canyons and the western limit of the illustration. These lines will be spaced about five nautical miles apart. The sounding interval will be as close as 100 meters on the steepest portion of the scarp. Bottom specimens will also be obtained at each vertical cast sounding. This development will permit a more exact determination of the face of the scarp and will, of course, be intensified if any unusual features, such as cliffs or terraces, are discovered.

3. Spanish Canyon will be further developed by a system of lines running approximately at an angle of 45 with the depth curves. Other areas to the northeastward, including the longest tributary of Eel Canyon will also receive further development.

While the careful development of these features and the intimate knowledge obtained from them are of considerabl scientific value, probably the greatest present value is to the navigator. Previously in thsi article a short description was given of the stranding of the Steamer BEAR near Cape Mendocino. If the mater of that vessel had been provided with a modern chart of this locality, he would not have made the assumption that the cape had been safely passed and the disaster probably would have been averted.

At that period, of course, soundings were obtained by vertical casts or pressure tubes. Nowadays, sounding is much simplified by the use of an echo sounding device with which more and more commercial vessels are being equipped. With such a device and a modern chart, a ship rounding Cape Mendocino in fog can determine its position almost as accurately as though the weather were clear, simply by adjusting the profile of this unique bottom as given by a continuous line of soundings, to the chart.

It is pleasant to think that Nature, while making Cape Mendocino one of the most dangerous points on our coasts, with its reefs, fogs, and generally unfavorable weather, has so patterned the submarine topography that the very peculiarities of its features increase the safety of navigation in this locality.

Following are some remarks by Captain N.H. Heck concerning the seismological aspects of the submarine scarp.


N.H. Heck, Hydrographic and Geodetic Engineer
Chief, Division of Terrestrial Magnetism and Seismology
U.S. Coast and Geodetic Survey

Recent confirmation of early surveys which indicated a vertical scarp running west from Cape Mendocino in about latitude 40 18' raises the interesting question as to whether or not this is an extension of the San Andreas Fault. Since this means a sharp change in direction of the fault as it leaves the coast, it is important to introduce the seismic evidence. It is to be understood that the San Andreas Fault is a major earth feature extending from the vicinity of Cape Mendocino through California into Mexico and possibly even farther. The only similar formation on land, comparable in extent, is the great fault which begins in northern Syria, extends through Palestine and across the Red Sea into Africa. It is possible, however, that the great submarine troughs are similar phenomena.

It is necessary first to appraise the accuracy of the locations of the known earthquakes of the region. Several elements are essential. The instrument should be suited for recording the earthquakes that occur, there should be a proper distribution of stations, the time scale on the records should be open so that the time of arrival of an earthquake phase is accurately known, and the velocities of the earth waves of the region should be known. If we were discussing the problem in the vicinity of San Francisco or southern California, these conditions would be met, but they are far from being met in the Cape Mendocino region.

During recent years there have been seismographs at Ferndale and Ukiah, and it is probable that the earthquake epicenters are reasonably well recorded off Cape Mendocino, but these are not enough stations for best results; and, since we are dealing with a geological phenomena, the history should be as long as possible. Prior to the installation of the two stations mentioned, only approximations of the epicenter could be made. A number of vessels reported feeling shocks in the vicinity, but the difficulty is that no matter what the relation of the vessel to the epicenter may be, if it feels the shock, it appears to come directly from beneath the vessel. It is probable that the recent determinations of epicenters are 10 or 15 miles in error and the earlier ones in still greater error, so that the direct association of earthquakes with the well-defined scarp cannot be made. However, the definition of the scarp is very important in connection with future more precise determinations. While the evidence is incomplete there is reason to think that there is not the same degree of earthquake activity to the northward of Cape Mendocino, either on land or sea, as there is just south of that vicinity.

Professor Perry Byerly of the University of California, in a paper presented at the Richmond meeting in 1938 of the American Association for the Advancement of Science, made the interesting suggestion that the fault continues only until it meets the region where there is a transfer from continental to oceanic crustal conditions (i.e., from 40 km. thickness to practically zero.) This is a very interesting speculation, and it is in part supported by the westerly trend of the scarp.

In any case this is an excellent example of the need for accurate depiction of submarine features in connection with geological and seismological problems.


Publication of the National Oceanic & Atmospheric Administration (NOAA), NOAA Central Library.

Last Updated: June 8, 2006 9:24 AM

Privacy Policy | Disclaimer