An official website of the United States government
Here's how you know
A .mil website belongs to an official U.S. Department of Defense organization in the United States.
A lock (lock ) or https:// means you’ve safely connected to the .mil website. Share sensitive information only on official, secure websites.

April 10, 1963: Search for the USS Thresher

April 10, 2013 | By Frank A. Andrews, Captain, U.S. Navy
This article was published in the May 1964 issue of Proceedings as "Searching for the Thresher" by Frank A. Andrews, Captain, U.S. Navy.
The Thresher search was very much an ad hoc operation. On 10 April 1963, the day of the Thresher's loss, there was no real search organization, no search technique, nor specific operating procedures for locating an object lying on the ocean bottom at 8,400 feet. In the first frantic hours after the Thresher's loss, a full scale search effort consisting of 13 ships was laid on with the aim of scouring the ocean for possible life or floating signs from the Thresher. Within 20 search hours, all hope for survivors had passed, and the entire Thresher project began to change character from that of a standard Navy search and rescue operation to that of an oceanographic expedition. This special expedition soon consisted of three ad hoc elements, which, as later events were to show, combined in a most successful and harmonious manner in support of searching out the Thresher's hull.

Diagram of the search for the lost USS Thresher

The first was the sea-going element. This group, called Task Group 89.7, was ever changing in number and types of ships. At its maximum at-sea size, it consisted of 13 men-of-war (including two submarines) and many search aircraft rushed to the disaster scene on the day of the Thresher's loss. At its minimum, TG 89.7 consisted of one lone oceanographic vessel?the Conrad on one occasion, the Atlantis II on another?left toiling away on station while the task group commander and staff (usually one officer and one chief radioman) were ashore conferring with others in preparation for the commencement of a new phase of the search. In all, 28 naval warships and five oceanographic research, or service, vessels participated in Task Group 89.7 from 10 April 1963 until 6 September 1963, when a substantial portion of the Thresher wreckage was located by the bathyscaph Trieste.

The second of the expedition's three elements was an 11-man shore-based brain trust called the CNO Technical Advisory Group. Its mission was to provide technical guidance to the at-sea search effort. In actual fact, this Advisory Group did much more than propose ideas. Its members also procured ships and hardware, and, in the case of certain individual members, came to sea with the ships to assist in searching. The Chairman of the Advisory Group was Dr. Arthur Maxwell, Senior Oceanographer in the Office of Naval Research. Captain Charles Bishop, U.S. Navy, the senior submarine officer in the Office of the Deputy CNO for Research and Development (OP-07), served as Co-Chairman and CNO liaison officer. The membership of the committee consisted of senior representatives from the Naval Oceanographic Office, the Lamont Geological Observatory, the Bureau of Ships, the Hudson Laboratories, the Naval Research Laboratory, the Oceanographic Department of the University of Rhode Island, the Woods Hole Oceanographic Institution, the Naval Reactors Branch of the AEC, and the Oceanographic Group at the University of Miami.

The third special element was the Thresher Analysis Group which set up operations in the Walsh House at the Woods Hole Oceanographic Institution, Woods Hole, Massachusetts. This Group soon became known as TAG WHOI, pronounced Tag Hooey. Its leader was Mr. Arthur Molloy of the Navy's Oceanographic Office in Suitland, Maryland. TAG WHOI had a varying complement but, over-all, 15 civilians or naval officers spent three or more weeks with this element. These men represented the Submarine Development Group at New London, NAVOCEANO, NEL, NRL and WHOI; they were all obtained from their many parent organizations simply by asking.

The mission of TAG WHOI was to receive and analyze all data obtained at sea and to prepare appropriate search charts showing the location of pertinent clues previously found by the search group. In addition, this group acted as a source of briefing for senior naval officers or chief scientists of various oceanographic search vessels prior to sea trips. They also carried on a major effort in expediting hardware demands from the Task Group at sea. Their location at Woods Hole permitted direct liaison with WHOI scientific personnel as well as quick radio communication via single side band with the Search Task Group 220 miles to the east of Cape Cod.

The initial plan for locating the Thresher was formulated on Friday 13 April, on board the Atlantis II, in an all-day conference which I, as Search Group Commander, held with Mr. Sidney Knott, Senior WHOI scientist of the Atlantis. This conference also included a lengthy radio telephone conversation with Dr. Brackett Hersey, Chief Physical Oceanographer of Woods Hole Oceanographic Institution, who confirmed that the basic plan conceived that day was in essential agreement with most of the thoughts then prevalent in the oceanographic scientific community. In the previous 36 hours, apparently, many minds had been thinking about how exactly to locate the Thresher. That night, in a dis?patch to Vice Admiral E. W. Grenfell, Commander Submarine Force, U. S. Atlantic Fleet, the following over-all plan was suggested:

Phase I. Search. A fine grain, bathymetric survey would be conducted. The term "bathymetric survey" is used by oceanographers to denote a sonic depth finding study of a given area. "Fine grain" meant that a Fathometer of 300 yards sweep width would have to cover the length of the search area in strips 300 yards wide. The type of Fathom?eter to be used had a precision recorder read out and in theory would show a bottomed submarine as an hyperbolic trace perhaps 200 yards long, and standing at mod point 15 to 30 feet in relief above the ocean floor.

Phase II. Initial Classification. All Fathometer echoes classified as "possibles" would be investigated with a deep-towed Geiger counter, or side-looking echo-sounder, or magnetometer.

Phase III. Final Classification. All contacts passing Phase I and Phase II would be photographed by either a deep TV camera, or a still camera. Hopefully, the Thresher's hull or parts thereof, would be in the resulting pictures. Later, a Phase IV was added to the plan. The title of this phase was "Close Examination and Study of the Thresher Hull". When positive photographic evidence was found, the bathyscaph, Trieste, would dive to permit her crew to study the wreckage. The over-all plan was imaginative and certainly appeared to possess a degree of logic, except for the following:
  • No one was quite sure whether or not the Thresher would return an echo from the search Fathometer. In fact, some suspected that the Thresher was buried deeply in the sand, while others suggested that the hull was in many pieces spread over a large area.
  • No one was quite sure that the navigation in the disaster area could be carried out accurately enough to insure 100 per cent Fathometer coverage of the ocean bottom with a minimum of duplication. After all, the sweep width of the Fathometer was about 300 yards, whereas the search area had been defined as an area 10 miles by 10 miles (4,000,?000 square yards) with the center at the Thresher's most probable location. The datum or most probable position had been chosen as the position of the USS Skylark (ASR-20) at 0917 on 10 April when the Thresher was last heard from on underwater telephone
  • No one possessed any real operational experience at towing a magnetometer, Geiger counter, TV camera or a side-looking, echo sounder 15 to 200 feet off the bottom at depths of 8,400 feet. In fact, the design of the various sensors was, in April, merely a topic for discussion, or at most in the preliminary "purchase-of-parts" stage in the various oceanographic laboratories throughout the country. The notable exception to this was the deep camera. There was ample experience in still photography at 8,400 feet and deeper, and all of the oceanographic groups were extremely handy with this technique. Until 10 April, however, most of the bottom photography work had been concerned only with pictures taken wherever the camera happened to chance as it was towed astern on a random wiggling tow cable. The act of selecting an exact geographic position and then placing a camera within 50 yards of this position in a depth of water one and one-half miles deep, with a surface current of one to three knots running, had never before been accomplished successfully.
But all scientists are optimists, and oceanographers are super optimists. In any case, the problems listed above certainly provided no discouragement to the various oceanographic teams who had to solve them. The plan was therefore approved by COMSUBLANT and the CNO Technical Advisory Group. A program to procure ships, men, and equipment was commenced. The actual search conducted over the next three months can be divided into five parts.

Part I, conducted from 13 April until 1 May, was originally named a "fine grain" survey. In actual fact, because of fast and varying random surface currents, and poor navigation, this survey using the precision Fathometer was really a very "coarse" grain survey. In the search area, the Labrador Current, flowing generally southwest, parallels the Gulf Stream, moving generally northeast. The result is a very chaotic surface current situation with flow over 24 hours from any direction around the compass, varying in magnitude between zero and four knots. The major navigation tool which was used in this survey was Loran A which under the best of conditions is probably accurate only to within about 2,000 yards.To overcome these problems, the "cluster" technique familiar to the mine warfare people was tried. All echoes were plotted as reported by the survey ships. A second, third, or even fourth survey ship pass would be ordered over previously reported echo positions, and each time the navigational position of all echoes would be again plotted. If a cluster of echo positions began to develop on the chart within 500 yards of each other, it was assumed that these could have resulted from one and the same object on the bottom.

The cluster was therefore worthy of being called a "possible" Thresher hull posit. In this manner, 12 posits were defined named ALFA, BRAVO, CHARLIE, DELTA and so forth. The major conclusion of Part I of the search, was that some possible "hull posits" had been developed, but the entire survey would have to be redone because of navigational inaccuracy. A "fine grain" survey using Decca and Loran C navigation was soon organized. Part II, the "fine grain" survey was accomplished in 13 days.

For this work, four ships were made available; the USNS Mission Capistrano (AG-162), a former T-2 oil tanker hull now used by ONR in Project ARTEMIS, the USS Prevail (AGS-20), an ex-fleet mine sweeper, the USS Allegheny (ATA-179), a fleet tug hull used by the Hudson Laboratories as an oceanographic vessel, and the USS Rockville (EPCER-851), a PC used by the Naval Research Laboratory as a research vehicle. Each of these ships was assigned one-quarter of the 10-mile by 10-mile search area for survey with its precision Fathometer. A NAVOCEANO team-was assembled on board the Mission Capistrano, and boating arrangements were made to bring the recording traces of the other three ships to the Mission Capistrano each day. On the Mission, in a large below-decks laboratory space, the data was screened and all possible echoes or hull posits were plotted and named for subsequent follow-up by the oceanographic vessels used as classifier ships - the RS Atlantis II, the USNS Conrad (AGOR-3) and the USNS Gilliss (AGOR-4). During this "fine grain" survey, it was clear that the Decca Green Line (from a station in Halifax) was extremely stable and would permit locating a ship within ?100 yards of its actual north-south geographic line of position. There seemed good reason to believe that the four survey ships had indeed placed a Fathometer of 300 yards sweep width over every square yard of the search area.

Unfortunately, the result of the fine grain survey was the definition of 90 "possible" Thresher hull positions instead of the 12 originally defined in Part I of the search.

Apparently, bottom topography was also returning echoes which looked similar to those expected from the Thresher. During Part II, the classifier ships had not been engaged in the Fathometer sweep. Instead, they had been busily engaged in trying to photograph or investigate with a deep magnetometer some of the hull posits defined in Part I. They had no success. Faced with the chaotic surface currents, and with no precise knowledge of the exact location of a sensor on the end of a 9,000-foot wire, the classifier ships were simply unable to pass a camera or magnetometer over or near any of the positions from which echoes had been received.

Naturally, the goal of investigating the 90 "possible" Thresher hull positions using the camera or magnetometer seemed unattainable unless the classifiers could become more proficient. A solution had been proposed which would permit the classifiers to become more proficient, and indeed, would check out the entire concept of the precision Fathometer as a search tool. We would bottom a World War II submarine hulk north of the search area and give the scientific ships some "Type Training," as the man-of-war Navy calls it. This proposal was quickly approved by the CNO Advisory Group and by COMSUBLANT. The ex-USS Toro was nominated, procured, sailed to Boston, and readied for sea. Thus, Part II, the fine grain survey ended with the return to port of all units except the Atlantis II, with plans for some of TG 89.7 to return the following week to place the Toro on the bottom.

Part III should properly be called the finding of clue No.1.

The Atlantis II did not return to port immediately with the other ships, but instead stayed at sea to do more photography work around posit DELTA which the Atlantis II had discovered in the coarse grain survey. DELTA had all the required echo characteristics which the Thresher hull theoretically should have produced; proper dimension and proper energy return relative to the energy returned from the surrounding bottom. Hence, the Atlantis II, with great pride of authorship, was determined to place a camera on DELTA and photograph the Thresher.

Although the Atlantis II never actually photographed DELTA, she did, on 14 May, obtain photographs of very suspicious looking debris about 700 yards north of DELTA. This debris consisted of paper, wire, and bits of twisted metal, and it is properly called the first clue in the bottom search for the Thresher.

DELTA subsequently has turned out to be topography little luck doesn't hurt anyone. In fact, this illustrates the well-known scientific procedure of discovery while carrying out the plan and not as a direct result of the plan. The news of the debris photography caused cancellation of the Toro drop and led directly to the next part of the search.

Part IV should be called "The Classifier Ships Concentrate at DELTA."

A 2-mile by 2-mile area was defined with center at DELTA. The Conrad, the Atlantis II and the Gilliss returned immediately to the scene, prepared to survey with camera and magnetometer. About a week to 10 days was spent in this phase with very little more to show than had originally been photographed by the Atlantis II. The Conrad, however, in a fit of experiment and while her camera rig was under repair, did dredge, using a deep scallop rig. The result after a dozen or so attempts, was a packet of 12 0-rings with certain name plate data written on each of the 0-ring envelopes. It was subsequently determined that these o-rings could have come from the Thresher or certain Navy aircraft.

The Atlantis II, spurred on by the Conrad's success, also rigged a dredge to be attached to her camera. This home-made device, literally made of baling wire and coat hangers, brought up small rocks, sea life, and finally a section of battery plate six inches in length. The battery plate was later identified as being definitely from a Thresher-class submarine. All classifier ships except the Conrad returned to port about 26 May. On 29 May, the Conrad reported the photographing of definite parts of the Thresher hull. The news was released to the press, and subsequently had to be denied. In actual fact, the Conrad had only photographed a part of her own camera. All faces were red. The Conrad returned to the battle, prepared to do or die, and in the next 10 days, kept camera and magnetometer going in and out of the water around the clock.

Part IV, "Classifiers on Station," ended about 15 June with the discovery by the Conrad of clue No.2: an oxygen bottle, a sonar hydrophone, and a piece of 10-foot sheet metal which appears to be sonar baffling used on the exterior of a submarine hull. These were the first large pieces of the Thresher's hull to be located. In addition, the Conrad located a large magnetometer contact on three separate occasions. Part V, the Trieste's operations, followed naturally through the positive encouragement of clue No.2.

The Trieste, with the USS Fort Snelling (LSD-30) and the USS Preserver (ARS-8), had been standing by in Boston since mid-May. An operation order to conduct at sea dives with the Trieste had been written and promulgated. Hence, with evidence on which to base the Trieste dives, TG 89.7 sailed on 21 June to exploit the previous search work. Between 21 June and 8 September, TG 89.7 was at sea for two periods of approximately three weeks each. During this period the Trieste made a total of 10 dives, five on each sea trip. The over-all result of these dives was the final discovery on 6 September of a substantial portion of the Thresher's wreckage, located as far as navigational error is known, within yards of the magnetometer contact reported by the Conrad. This area contained, amongst many other large pieces of sheet metal, the same piece of 10-foot by 10-foot section photographed previously by the Conrad. The area was described by Lieutenant Commander Donald Keach, the Officer in Charge of Trieste, as "a huge auto?mobile junk yard with big pieces of heavy metal all over the place."

TRIESTE surfaced. Blowing water out of entrance changer to permit observers and pilot to climb out

The use of the Trieste in the role of deep search and study vehicle is a story in itself. This remarkable and simply constructed vehicle will certainly revolutionize the exploration of the deep-ocean bottom in the years to come. The ability of men to go to the bottom to see for themselves, or to combine visual observation with photography, or sonar detection or magnetometer search, or radiation measurement is the most flexible and competent of all oceanographic techniques. There were, of course, many minor operational problems associated with towing the Trieste to sea and making 10 dives in 8,400 feet of water. The Trieste cannot be rigged for diving in heavy weather; her surfacing operation is dangerous in fog or after dark, her crew up to this time had never made more than one dive per week.

In general, the Trieste had never really been used in a search role, but rather had merely investigated any particular area in which she happened to land. The Trieste crew, however, had worked together for some months in the San Diego area, and had good answers for most of the problems which arose.

The wisdom of using a large ship, the Fort Snelling, as a seagoing base of operations, and the salvage ship Preserver as an immediate mother and tow ship became increasingly apparent throughout the operation. The whole operation would have been hopeless without the housekeeping, communications, and repair facilities provided by these two ships. The major problems associated with the actual Trieste dives were two in number. The first, navigation, was finally solved in a reasonably acceptable manner, although great improvement is still possible. The second, the slow deterioration of the Trieste's topside wiring and hull parts simply by being at sea for a long period of time, was one which we learned to live with, until finally it appeared that the Trieste had to secure and go home. Hopefully, a redesigned Trieste will make this second situation less of a problem.

To place the Trieste in the proper position for diving. The Preserver with the Trieste in tow, was navigated by radar ranges and bearings on a "taut wire" buoy planted 1,000 yards north of the magnetometer contact previously reported by the Conrad. After diving, the Trieste was maneuvered on dead reckoning for the most part. However, 1,441 colored and numbered markers placed methodically on the bottom in a prearranged grid throughout the area served to tell the Trieste where she had been. Each marker consisted of a window sash weight, as anchor, and a colored plastic sheet attached to the anchor by two to three feet of nylon cord. The markers had been dropped several weeks before by the Allegheny on the basis of Decca navigation and seemed reasonably close (perhaps 300 yards) to their intended ground position. The markers were called "fortune cookies." The Gilliss, also on the scene with the Trieste, had a 3-D tracking system designed by the Applied Physics Laboratory at the University of Washington. This system transmitted a sonar signal to a transponder located on the Trieste, which in turn re-transmitted a signal to all of three hydrophones located on the Gilliss' hull. By measuring the time difference of arrival, the Gilliss could compute a very accurate range and bearing of the Trieste from the Gilliss. Simultaneously, the Gilliss could locate a transponder fixed to the ocean bottom, and thus give an accurate range and bearing of the Trieste from a fixed geographic position.

Unfortunately, this system was just beginning to work when the Trieste's slow hull deterioration took over, and we all had to go home. The Washington APL System appears to offer the greatest promise for future bathyscaph work. In the diagram on page 72 you can see the buoy location, the Atlantis II debris area, the location of the Conrad's debris and magnetometer contact, and the location of the large wreckage area sighted by the Trieste. The brass pipe which is seen in the photograph on page 73 had been marked while the Thresher was under construction with certain drawing and job order numbers, as well as the words "593 boat." The pipe apparently came from a hot water flushing line.

Two other very valuable contributions to the Thresher operation must be mentioned before closing. One was the weekly or semi-weekly conference held at sea between the Task Group Commander and the various chief scientists on scene. Many a boat trip or high-line transfer was made to carry them out. These get-togethers served as very valuable idea sessions and certainly made the job easier for the Task Group Commander. The second contribution was the information obtained from the high-powered receiver and transmitter at the Hudson Laboratories, manned almost continuously night and day by Mr. Gerry Fisher of that laboratory.

This unofficial communication center operating on 3385 KCS SSB could reach anybody in the search operation at any time, thanks to the faithfulness and efficiency of ROBIN as the Hudson Laboratory voice call was known. There were many moments when all official Navy circuits were down, but the scientific circuit, for which 3385KCS was designated, could put the at-sea group in touch with the CNO Tech Advisory Group in Washington, or the Analysis Group at Woods Hole, or even some member of the Advisory Group who was traveling around the country at the time.


It is my understanding that the Trieste II, a new Trieste superstructure with the old Trieste sphere, will go to sea in April of next year to finish the Thresher search operation. Hopefully, the new WHOI research submarine Alvin will also be available. The Thresher Search Operation shows that, with patience, the deep search and study problem can be solved. There are major problems yet to be solved, however, if the Navy is to become proficient at this type of an operation, whether it be search for a bottomed submarine, satellite, missile, or any other object. Most important, we must develop improved search techniques with the capability to locate the search sensors precisely in a geographic frame of reference.