50 Years of Multibeam Sonar


First conceptual diagram of slant range and sonar observed area in BOMAS (Bottom Mapping Sonar) proposal to the US Navy from General Instruments, the system developer.

An article in Hydro International magazine tells the story of how multibeam sonar was developed in the early 1960s. Originally conceived as an airborne radar mapping system, the concept of beam formed imaging was proposed to the US Navy, which immediately understood its application for use in hydrographic surveys.

Diagram from General Instruments proposal to the US Navy.
BOMAS (Bottom Mapping Sonar), later called SASS (Sonar Array Sounding System), was first installed in 1963 aboard USS Compass Island, a Navy electronic system test vessel. SASS, developed by the contractor General Instruments, was intended to produce contour maps of the ocean bottom. Sixty-one beams, each one degree wide, fanned out beneath the ship. Gyro information was used to stabilize the return echoes for roll and pitch.

Contrary to reports elsewhere on the web, the first SASS survey ship deployment was on  USNS Michelson at the Brooklyn Navy Yard (NY Naval Shipyard) in early 1964. Michelson was to proceed directly to its new operating area in the Pacific, but as the SASS (and associated systems) required further testing. Operating out of ports in Florida, sea trials were conducted in the waters around the Bahamas during spring and summer of 1964. Michelson finally deployed to the Pacific in the autumn, via Panama. SASS was  subsequently installed aboard sister survey ships USNS Dutton and USNS Bowditch.

Usefulness of SASS was initially limited by the lack of  large scale digital data storage at that time. Increased computer power, digital signal processing and the use of GPS for control have improved multibeam sonar. Michelson's SASS was the grandfather of all survey systems, side scans and fish finders in use today.

Read the story in Hydro International here.



Mystery Seamount

From All Hands magazine, November 1966.





































Before Sonar

So, how did they conduct hydrographic surveys before sonar was invented?

Until sonar echo sounders came along in the mid 1930s mariners known as "leadsmen" used a lead line to measure water depth. That's lead, the metal, used as a sinker. Made of perhaps 25 to 30 fathoms of line, known to landlubbers as rope, a lead line was tagged at intervals of every two or three fathoms up to 20 fathoms of depth.

 


The tags of leather and differing types of cloth were such that they could be "read" by feel at night when not visible. Below is a chart of the traditional lead line markers. Sometimes the 13 fathom tag was made of blue cloth rather than leather.
 
 

After casting the line, the leadsman would call out the indicated depth. The great American novelist Samuel Clemens, author of the Huckleberry Finn and Tom Sawyer stories, took his pen name from the callout for two fathoms of depth: "by the mark, twain".

Here are some examples of the leadsman's callouts:


Here is a description of how the leadsman cast his line and probed the briny deep. This excellent article was liberated from the NOAA website noaa.gov/transformations/hydrography/.

Determining depth by lead line is clearly limited to shallow areas, 20 fathoms (120 feet) or less. This suggests that this method was used more to avoid running aground than conducting surveys. Other than tying two leadlines together end-to-end, other methods were needed.

Chain soundings were one variation, with tags attached to the chain at certain intervals. The first illustration shows markers at 100 centimeter spacing. Alternately, chains could be tagged in feet, fathoms or even surveyor's chains, a unit of length.

It became clear to installers of the first undersea telegraph cables that ocean surveys would be required prior to laying cable. Some interesting machines were developed in the US and UK to probe deeper into the oceans. Common to them was the idea of measuring the amount of wire or line paid out instead of marking it with cloth, leather strips or tags. Wheels or pulleys were machined such that their circumference equaled one unit of length, whatever that unit might have been. The wire ran once around the calibrated wheel. Gears were attached to pointers that indicated the number of wheel revolutions, metering the length paid out into the sea.

Naval officer Charles Sigsbee developed his sounding machine in the late 1870s (second picture). For some reason the globular lead weight was supposed to be left behind on the ocean bottom after each cast, drawing the cylindrical part back up to the surface. Sigsbee later was commander of the USS Maine when in 1898 it exploded in Havana harbor during the Spanish American war.

The Passerelle sounding machine is shown in the third picture. The metering apparatus is marked "A" in the drawing.

 During the Napoleonic wars every ship in the British navy had one of  Edward Massey's sounding machines (fourth picture). While intended more for shallow waters, he sold 1,750 of them to the British government after it was recommended by the Board of Longitude, which was supposed to be inspecting nautical clocks.

Aboard Michelson a device much like the Massey machine was used to measure wire line paid out in Nansen bottle casts.

A bit of research reveals that all sorts of "sounding machines" were invented by just about everyone. Even William Thomson, better known as Lord Kelvin, got into the act with his 1907 Motor Sounding Machine, later installed on RMS Titanic and her sister vessels. Kelvin was renown for his work in thermodynamics, magnetism and electricity. However, he was a failure at aeronautics, declaring in 1895 that "heavier-than-air flying machines are impossible".






 


Brooklyn Navy Yard (3)

During late winter or early spring of 1964, while in the Brooklyn Navy Yard, Michelson was drydocked for the periodic hull cleaning and repainting that all ships require. Also, two sets of transducers for the new multibeam sonar were installed while in drydock, the transmitting set being longitudinal along the keel and another set across (athwartship) for the receiver. All were placed forward of the superstructure, beneath the former number three hold.

Photo of a Victory Ship model
A ship in dry dock is an awesome sight. While not really large, Michelson looked like a nautical giant when viewed from the dry dock floor. There is a lot of ship below the water line that one rarely gets to see. Walking down the steps into the dock the rudder and propeller looked bigger and bigger, but ineffectual, out of their element. From the keel, resting on wooden blocks, it was 40 feet of steel up to the main deck. I walked around the ship and, crouching down, under it! This was very impressive. Work had finished on the hull and within a day or two the dock was flooded and Michelson was eased on out.

Within Michelson's navy detachment men had transferred out to other ships and shore stations while new ones were reporting aboard. Since the usual tour of duty was just one year aboard, few of those from early 1963 remained. A few of us volunteered for another tour and were retained. New officers arrived. We continued to reside in less than palatial accomodations at the navy receiving station (RECSTA) across the street from the yard. 

This was a good time to select staterooms and bunks as we were soon to move back on board. I claimed a lower bunk having a secret booze compartment concealed below a custom made book rack. Some previous sailor had done a fine job on this one. Perhaps he was a bibliophile bookworm as well as a serious consumer of spirits.

Yardbirds were finishing their work and others cleaning up their mess. New work spaces for electronic stuff had been created on the two decks below our living quarters and more staterooms added in in number two hold. More tech reps, a/k/a field engineers, were expected to occupy the new living spaces.

Some old and new electronics were installed and ready for test:

  • In the SCC (Survey Control Center), the two Loran C receivers were reinstalled. 
  • NIC (Navigation Information Center) moved to the former Hydroplot location on the third deck. The NAVDAC computer was reinstalled, along with the new SINS Mark 3 Mod 3 inertial navigator, Transit satellite navigator, Sperry mark 19 gyrocompass and a Bunker Ramo CP677 computer.
  • Hydroplot and its Bendix G-15D computer moved down to the fourth deck, formerly home of the disused doppler sonar.   
  • A new sonar room on the fourth deck contained the SASS (Sonar Array Sounding System) multibeam sonar equipment.
  • The former NIC on the 04 level became semi abandoned, although some electronics remained. Henceforth it was called "old NIC".

Being the overall project manager for navigation systems, Sperry Gyroscope ran a school on Long Island to familiarize technical detachment members on how all this stuff was going to work. Every morning for a couple of weeks we hopped on a bus from RECSTA to Sperry's classroom in the basement of a strip mall in New Hyde Park.

This was an busy time in New York while the 1964-65 world's fair was preparing to open at Flushing Meadows, also site its 1939 predicessor. Our Sperry bus ride went by there each day.

Michelson was scheduled to move to the Western Pacific to begin survey operations, working out of the navy base at Yokosuka, Japan. This was exciting! 

The old ship began to take on a new crew of MSTS Pacific merchant mariners based in the San Francisco area, rather than the New York sailors who operated the ship while in the Norwegian Sea and Mediterranean. All, both licensed and unlicensed, were new. Military Sea Transportation Service (MSTS), now called MSC (Military Sealift Command) was the navy's own proprietary steamship company, operator of troopships, auxiliary vessels and odd ones such as Michelson.

Once we left the Brooklyn Navy Yard we would move to the naval supply center in Bayonne, New Jersey to take on fuel and provisions. Then off to California and the Pacific by way of the Panama Canal. 

That was the plan, but it didn't happen quite that way.


Survey Pattern

What is an ocean survey all about? Did your ship steam up and down, sideways or around in circles?

Michelson surveyed areas of interest by steaming in a north/south and east/west grid pattern pinging along with its single beam sonar and guided by the Loran C radio navaid. This is called a trackline survey.

The NAVDAC computer looked at Loran once a minute and after 35 seconds of deep thought began to print out the computed latitude, longitude and UTMs (universal transverse mercator) map coordinates as well some other data. This took another 25 seconds, then it was time for another look at Loran. Meanwhile the ocean below was probed by the sonar. Depth plus date and time were recorded on a precision depth recorder (PDR).

In the survey control center the duty oceanographer plotted position information on a chart and issued course corrections to the mate in the pilot house. Due to winds, currents and who knows what else a 000º (north) heading sometimes required course changes of a few degrees east (001º, 002º, 003º) or west (359º, 358º, 357º) to stay on track. There was kind of an art to all this. Too much or too little correction could mean having to run the line again. The delay built into the system by the painfully slow computer didn't help. 

 Generally, tracks 1000 yards apart were run inside of ten nautical mile squares, with additional tracks between those already surveyed as required to increase granularity. It took the better part of two days, sometimes longer, to complete one survey area, then we steamed off to the next selected site, sonar still pinging away.

This trackline grid survey was the norm while I was aboard Michelson. Multibeam sonar came into use in late 1964 which likely changed the survey methodology.
 
Typical Survey Pattern



Sonar Bathymetry

Mapping the ocean bottom requires continuous measurement of depth using sonic means and accurate recording of the data, referenced to time and geographic position. From when Michelson was put into service through c. 1964 a single beam sonar was used to acquire depth information. This "trackline" method involves steaming along preset North/South and East/West lines in a checkerboard pattern to cover the area to be surveyed. Accumulated data is later plotted to create maps (charts) of the underwater topography.


The sonar transmitter/receiver, was located in the survey control center. Michelson was equipped with the SQN-6 (XN-1) sonar, essentially a modified UQN-1B depth finder set. Nearly every naval vessel of that time had a UQN-1B on the bridge. This was the successor to the "lead line" used to measure depth in the days of sailing ships.

UQN-1 Sonar Depth Finder with cover off (left) and closed.
According to the UQN-1B tech manual the sonar transmitter delivered 800 watt pulses at an audio frequency of 12 khz. Somehow I remember that our sonar, for whatever reason, pulsed ("pinged') a bit higher at approximately 14 khz, controlled by a pair of quartz crystals. The indicator scope and recording mechanism in the SQN-6 cabinet was not used. The unit's transmitter ping was keyed by a separate precision depth recorder (PDR) which also took the SQN's receiver output to be recorded. Two large glass tetrode tubes furnished the powerful audio frequency output. These 4-65A tubes were commonly used in radio transmitters of that time. 

Few people could hear the 14 khz sonar pings, at about the human ear's high frequency hearing limit. Most of the time I could hear it below decks in the living areas, barely audible if listening carefully, a ping every 1 1/2 seconds. American TV sets had horizontal sweep oscillators running at 15.750 khz. I could hear that sound from TVs when I was younger!

4-65A Tetrode
The second sonar element was the transducer. This was the thing that transmitted the pings down into the water and received the resultant echoes. An off-the-shelf UQN-1B came with a transducer made for simple depth finding use. Ours was stabilized such that the transducer always pointed straight down, in line with the local vertical. This electro mechanical apparatus was located down a vertical trunk below the fourth deck, two levels beneath our living area. It probably had some sort of gyro reference to help keep it stable.

Electrically, the transducer was magnetostrictive. Electrical pulses (pings) applied to it changed its shape, converting the transmitter's pulses to acoustic mechanical energy in a narrow beam.

Looking at the stabilized transducer could make you seasick. It stood straight up while the ship rolled and pitched around it. The whole affair worked quite well, keeping the sonar beam more or less straight down unless we experienced really heavy seas. 

Our two precision depth recorders (PDR) were in survey control. Mechanical monsters, each about the size of a washing machine, initiated the transmitter's pings and recorded the results. These worked in a similar manner to plotters but with an electrified stylus instead of ink and printhead. The sonar guy spent a lot of time fiddling with these intensive care machines. A company called Timesfax made them, which identifies the origin of the technology. Fax (facsimile) was invented to transmit low resolution newspaper quality photos over phone and radio circuits.

A wide (26-28") roll of electrosensitive paper was loaded into the PDR's left side. The paper ran horizontally across the top to the take up roller on the right side. A track for the styluses ran from front to back on the machine's left. When turned on, a stylus would start its trip across the width of the paper.

The first point to be marked was the start line corresponding to sea level. After moving a very short distance the stylus encountered the keying block. Here contacts closed very briefly, initiating the transmitter's output (ping). The stylus would then record a short black mark where the sonar heard its own outbound ping. The stylus continued to travel across the paper. If the sonar receiver heard a return echo it would cause the stylus to mark the paper accordingly. One trip across the paper's 24 inch wide track took 1 1/2 seconds and represented a scale of 0 to 600 fathoms. One fathom equals six feet.

Timesfax Precision Depth Recorder (PDR) Operation. Click for larger image.


As the first stylus reached the 600 fathom mark (exactly 24 inches from the start line) another stylus started its trip across the paper, marking the start, keying the transmitter and recording return echoes. Another stylus appeared every 1 1/2 seconds, the equivalent of 600 fathoms of depth. The PDR also recorded some sort of time codes as well as manually inserted event marks. Trace intensity was adjustable.

Another type of PDR, this one from EDO Corporation.
All this actually worked well if everything was adjusted correctly. Before each trip our sonar guy had to check with the first mate or the bosun to get the ship's draft in order to set the keying block to correspond with how deep we were in the water. Precise stuff, indeed.

So what happens if the depth is greater than 600 fathoms? I knew somebody would ask that. Well, if the depth below the keel were say, 1000 fathoms, then the echo return would appear on the next pass across the paper, making that the 600-1200 fathom range. Periodically the oceanographer was required to make a phase check. He could disable the normal pinging, initiate a single ping manually from the next stylus pass, then listen with headphones for the return echo, noting on the PDR paper in which pass (or phase) the return was heard. Simple, huh?

Again, all of this electro mechanical stuff sounds hopelessly complex but it was the best we had and best there was at the time!

An example of a PDR bottom profile trace. Paper moved through this PDR from left to right, giving a continuous profile of the ocean bottom.



Surveying Undersea Highways


This article from the May 1963 edition of All Hands magazine describes how regular navy survey ships USS Maury (AGS-16) and USS Serrano (AGS-24) conducted surveys in the Gulf of Thailand. A field deployable Lorac radio navigator chain supplied control. 

In 1964 USNS Michelson used Lorac for position data while conducting seas trials in the Bahamas during the spring of 1964.
 
Note the UQN sonar used as depth recorder.

















































































Back to the Bahamas

One day during Michelson's seemingly unending 1964 Bahamas sea trials I was told that we were finished using the Lorac equipment. After a port call we would be heading for the Panama Canal, the Pacific, California and Japan, as per our delayed schedule.

I was to shut Lorac down and put it all back in the transit cases. Asking what to do with the custom made rack that came with it (a Lorac rack?) I was told to dispose of it. Dutifully, I packed up the equipment and threw the rack over the side.

Michelson returned to port in Florida. The Lorac tech rep left for Oklahoma. After a couple of days we were told that contrary to what we had been told, sea trials would resume and we needed the Lorac for navigational information! A different Lorac rep arrived. With its mounting rack deep sixed, I got some line and lashed the Lorac stuff down to the big chart table in the Survey Control Center. Once again we were off to the Bahamas.

The newly installed electronics still was not working as intended and not necessarily working together. While the new inertial system (SINS) had already been deployed elsewhere, the SASS (array sonar) had never before been installed on a survey ship. This was supposed to furnish a plot of the ocean bottom contours across a wide swath rather than the bottom profile we got from the narrow beam stabilized sonar. SASS continued to need work, given the limits of computer power and data storage (magnetic tape) at that time.

There is much to read on the internet about the history of array sonar. Michelson was the test bed for this new technology, grandfather to all side scan sonars and fish finders since then. In the summer of '64 it was unproven.

We also had an early version of the navy's Transit satellite navigator. At that time there were few satellites available, so position fixes could be obtained infrequently, but often enough to reset SINS, then the world's most expensive dead reckoning navigator. SINS, of course, was another technical grand daddy, to all naval and aeronautical inertial systems.

One of the new things that had been installed in Brooklyn was a large meter by the helm in the pilot house. SINS was supposed to feed it steering direction, left (port), center (on course) and right (starboard). The object was to keep the meter centered. All of this was supposed to help the duty oceanographer and the AB at the helm maintain a steady course over a desired survey track, but I never saw it work. We had a lot of stuff that didn't work. That was the reason for the four months of sea trials.

During the whole time, May through August, Michelson never called at any Bahamas port, instead operating out of Port Canaveral, the Port of Palm Beach and Port Everglades in Florida. Meanwhile, nearly every radio aboard ship was tuned to the Nassau radio station ZNS, pronounced "Zed-N-S", playing the the pop music of summer 1964.

Pacific Sea Floor Mapping

A paper published in the May 1969 proceedings of the Sixth US Navy Symposium of Military Oceanography describes an ocean survey in the Pacific. The Scripps Institute of Oceanography was contracted by the US Navy Oceanographic Office (NAVOCEANO) to prepare an atlas of charts based on new surveys as well as data from other existing sources.

The results were published in 1969 as H.O. Pub. 1301, Bathymetric Atlas of the North Pacific Ocean.
 

MAPPING THE NORTH PACIFIC OCEAN SEA FLOOR

T. E. Chase, Associate Specialist in Marine Geology 

S. M. Smith, Associate Specialist in Submarine Geology 

Scripps Institution of Oceanography 

Abstract: A series of sea floor topographic and physiographic charts of the North Pacific Ocean are being prepared for the Under-sea Surveillance Oceanographic Center (NAVOCEANO). In addition, reliable bathymetric and magnetic data throughout the Pacific are being put into digital storage. 

Divided into three areas, the coverage consists of 160 topographic charts originally prepared at approximately 1:1,000,000 scale, then reduced to about 1:2,500,000 for printing into three atlases; 10 composite contour charts, and 10 physiographic diagrams will be prepared for distribution throughout the scientific and naval communities. 49 charts between latitudes 0° to 60°N and longitudes 100°E to 160°E are completed and published in an atlas form. Three composite charts covering the area are also completed and undergoing color separation and printing. 

Computer programs have been developed to aid in processing data, and to date 392,000 miles of bathymetry and 206,000 miles of magnetics plus navigational tracks have been digitized. 

On July 1, 1967, the Scripps Institution of Oceanography began a contract with the Undersea Surveillance Oceanographic Center of NAVOCEANO to prepare a series of charts and a digital tape of bathymetric soundings of the North Pacific Ocean. 

The purposes of these efforts are many, and two are of immediate importance. 

Increased studies of the origin and history of the Pacific basin are important to the fields of Marine Geology and Geophysics. Detailed bathymetric sea floor charts are critically needed in such studies. 

Within many research organizations, studies of sound trans-mission in water along selected paths are often interrupted by topographic features. Here, reliable charts are also important. The majority of existing sea-floor contour charts, however, have been found to be of a smaller scale, contour interval, or too old to aid in these studies.

Four series of charts are being produced: physiographic charts (10 sheets); bathymetric contour charts (10 sheets); bathymetric contour charts at approximately 1:1,000,000 scale (160 sheets); and 3 atlases of bathymetric contour charts. 

The charts are a compilation and interpretation of the best bathymetric contours and sounding data available, and they present a complete series of topographic and physiographic charts for ease in referencing the floor of the North Pacific Ocean. They are designed for use by interested persons or groups working in the marine sciences as a quick reference to bathymetric contours and physiographic provinces. 

Figure 1 shows the distribution of and numbering system for location of the charts. The heavy borders are small-scale regional contour and physiographic charts. Areas I, II, and III are boundaries of the three atlases. Area I is contoured, and 49 charts are published in the Bathymetric Atlas of the Northwestern Pacific Ocean. Figure 2 shows a page from that atlas. (H. O. Pub. 1301.) 

The atlas of area I contains 49 bathymetric charts covering nearly 7-1/2 million square miles of the North Pacific Ocean between 4°S and 60°N latitude and between 100°E and 160°E longi-tude, including the South and East China, Yellow, Japan and Okhotsk Seas. 

SOURCES OF DATA 

Figure 3 shows the different sources of data used. The most useful were original echograms or soundings with high navigational quality.

J. C. Sylvester, Bathymetry Division (NAVOCEANO), supplied random soundings for each chart plus microfilm copies of original echograms and ships' navigational tracks.

Marine Geophysical Surveys, NAVOCEANO, supplied original echo-grams of detailed cruises in the North Philippine Sea. The ASW/USW Project, NAVOCEANO, under Mr. W. T. Hammond, supplied soundings of extensive surveys in the Sea of Japan. Published charts from the Maritime Safety Agency, Japan and the USSR gave detail in areas where soundings were sparse. 

Echo sounding equipment used to gather data for the charts included the Edo Corporation Sonar Sounding Set AN/UQN-1B, Westrex Corporation Mark V, X, and XV Precision Depth Recorders (PDR), Thomas Gifft Company Depth Recorder (GDR), Alden Electronic and Impulse Recording Equipment Company, Inc., Precision Graphic Recorder (PGR), Alpine Geophysical Associates, Inc., Precision Echo Sounder Recorder (PESR), and Kelvin-Hughes Echo Sounder. 

EVALUATION OF DATA 

All data was evaluated for its importance in preparing the contour charts and for digitizing into the data bank. Unfortu-nately, many of the random soundings were of value in predicting topographic trends but lacked enough orecise navigational information to warrant inclusion in the digital data collection. 

A large segment of the data was shown on charts and sounding sheets contoured in meters, both corrected and uncorrected for sound velocity. These were converted to "uncorrected" fathoms by use of Matthew's Tables of the Velocity of Sound in Pure Water and Sea Water. 

Initially, efforts were made to use digital (computer) output for the chart preparation. However, the random spacing of pre-cision sounding information with precise navigation was insuffi-cient. Therefore, it was necessary to prepare a data tape separately and to develop the charts by visual interpretation and manual cartographic methods with extensive use of stable base materials. 

DIGITIZING OPERATION 

Data from smoothed navigation plots were digitized onto IBM cards using the Benson-Lehner OSCAR 5-2 XY Reader. After the cards are run through data checking programs and errors removed, they are input to a program which produces a magnetic tape containing time in accumulated minutes and positions at these times. A parallel series of operations are carried out on the echograms and magnetic records, using a CALMA 480 in addition to the B-L OSCAR, the final product being another magnetic tape having time in accumulated minutes and the depth or magnetic field values at these times. Most of the depth data were digitized on the OSCAR at 3 to 5 minute intervals of ship time (1/2 to 1 mile apart), the remainder being done on the CALMA at 1 minute intervals. The interval for magnetic data was 6 minutes. The navigation and depth on magnetic value tapes are used in turn as input to another pro-gram which merges the navigation and depth or navigation and magnetics onto another tape that becomes the primary storage for the expedition. The merged data can then be used for profiles at various scales and vertical exaggerations (Figure 4), sounding plots of single expeditions, sounding compilations of many cruises within a given area (Figure 5), statistical analyses, and trans-mission to other agencies. Additional programs are being developed for producing computer plotted index track charts of digitized data as well as bathymetric and magnetic profiles plotted along tracks on Mercator projection. As of March 1969, 392,000 miles of bathymetry and 206,000 miles of magnetics have been processed. 

CONCLUSIONS 

The efforts expended during the chart preoaration has resulted in contours of many previously uncharted seamounts, trend determinations of other primary structural features like trenches and ridges, plus limits of large physiographic provinces. 

Basic knowledge of structural and tectonic conditions were used throughout the chart preparations. Where sounding data was not sufficient to give detailed portrayal, interpolation was ex-tended to gain the most accurate configuration. 

The charts presented here do not represent the final configuration of the sea floor, for many precise and detailed surveys are needed to give complete coverage. It is felt, however, that the different scales, contour interval, and physiographic interpretation used is sufficient to give as complete a sea-floor portrayal as possible with the present data available and knowledge of geologic features.

(Note: Figure 1 is not shown here but may be seen along with the entire report on the web at: usnavysymposiumo669usna.pdf, pages 345-53.)


One of the illustrations included shows the ocean bottom contours of the Mariana Trench,
part of which is six miles deep. Click for larger image.