Covering approximately three quarters of the Earth’s surface, water presents a formidable obstacle when searching for a shipwreck. The sheer size of the two major oceans, the Atlantic and the Pacific, cover an astounding 105 million square miles. The average depth of these two water bodies is just shy of 13,000 feet. However, technology is constantly making the world smaller and the oceans more accessible.

Remote sensing is broadly defined as the science of acquiring information without direct physical contact. Without direct contact, some substitute method must be utilized for gathering and transferring this information. Being that the oceans are a particularly hostile environment, with frigid temperatures and potentially bone-crushing ambient pressures, remote sensing applications are particularly attractive to researchers. Whether utilizing sonar systems or satellites, remote-sensing devices enable one to safely obtain vast amounts of oceanographic data without getting one’s feet wet.

In particular, the diving community is frequently attracted to remote sensing applications when searching for shipwrecks. In the past, blind luck often played a significant role in determining success when looking for shipwrecks. However, when hunting for shipwrecks now, it is possible to greatly increase the odds in your favor. Individuals can greatly extend the range and scale of their search with the use of a remote sensing device such as a magnetometer or side-scan sonar, covering exponentially more area than with a standard echo sounder (i.e., fish finder). Further, divers can gain important information on an intended dive site, such as wreck orientation and identification of potential hazards, even before suiting up. The following is a quick primer on remote sensing devices and applications germane to the shipwreck diving community.

WHERE THE HELL ARE WE?

Perhaps the most critical component of remote sensing is positioning. Without accurate positioning information, searching for anything would be frustrating and inefficient. In the past, position was determined either by dead reckoning or celestial navigation. Dead reckoning was the process of plotting course and distance from a known position. Obviously, there are numerous sources of error when using this system. Wind and currents can subtly alter one’s course, and these sources of error are compounded the farther away you sail from your point of departure. Celestial navigation is based on the tenet that an unknown position can be determined using the known position of the stars. By employing a sextant and the fixed position of stars as landmarks on the open ocean, mariners could obtain the necessary information to determine their bearing. Unfortunately, while it was possible to establish one’s latitude fairly well using these methods, longitude remained a dilemma until the mid-eighteenth century. Since local time progresses one hour for every 15 degrees to the East or West, to determine longitude one had to know not only the time at the current position, but also the time at another known fixed position. Typical pendulum clocks could not be used aboard ship due to the motion of the vessel while at sea, as well as changes in temperature and humidity. John Harrison solved the longitude problem by developing an ingenious spring-driven clock that accurately kept Greenwich Mean Time. Thankfully, there are now more accurate ways of determining one’s position on a featureless ocean.

There are basically two current methods for determining position at sea: long-range navigation (LORAN) or the use of global positioning system (GPS). LORAN is a federally run system that utilizes land-based radio navigation transmitters to provide accurate position and timing information. Analogous to triangulation, a survey vessel can determine its position by receiving and comparing the time differences of the low frequency signals (10-300kHz) emitted from a chain of these fixed-position transmitters. LORAN transmissions and receivers typically provide a repeatable accuracy of 60 to 300 feet. Repeatable accuracy is the range to which one can return to a saved position. Some of the drawbacks to LORAN are signal dispersion and degradation over long distances. Accuracy in LORAN is achieved by the use of multiple secondary transmitters in a chain that allow a vessel to triangulate its position. When far out at sea, the intersections of these signals develop a very wide spread and hinder the ability to determine an accurate position.

In contrast to LORAN, GPS utilizes a network of satellites to determine a precise position anywhere on the globe. By obtaining the signal of any three satellites, which travel in a very precise orbit, it is possible to triangulate your position anywhere on Earth. However, since satellite triangulation is based on geometry, error can still occur that will diminish accuracy. If the satellites are not spread out widely enough, it is possible to incur some error. Obviously, accuracy will be improved by relying on as many satellites as possible to triangulate your position.

The GPS satellites broadcast two signals, a coarse acquisition (CA) signal and a precise positioning system (PPS) signal. With just the CA signal, and in the absence of selective availability, an intentional randomness that can affect repeatability, GPS has an accuracy of approximately 66 feet. The PPS signal has an accuracy of less than three feet, but it is available only to military and other licensed users. Differential GPS (DGPS) is another means to improve accuracy of the standard CA signal. While accurate, GPS satellites still produce some error. As with land-based signals, geometry can be critical with GPS signals as well. In order to reduce GPS signal error, users can receive a second corrected, or differential, position signal such as the satellite-based wide-area augmentation system, which will enable accuracy to within three feet. Differential GPS (DGPS) generally employs a network of reference stations to re-broadcast the GPS signal, which allows for accurate and repeatable positioning information.

ACOUSTIC SYSTEMS

Now that we know where we are, it is necessary to find a target that is lying somewhere under potentially thousands of feet/meters of water. One of the most basic ways to locate a wreck is through the use of a fathometer (e.g., depth finder, echo sounder, etc.), a basic form of sonar. Sonar (sound navigation and ranging) utilizes transmitted and reflected sound waves to measure distance underwater. Just as the name implies, an echo sounder utilizes acoustic energy to measure bathymetry, or water depth. By measuring the time it takes for a pulse of energy with a known velocity in water to reach the bottom and be reflected back, it is possible to determine the distance to the seabed. The emission of a cone-shaped beam from an acoustic transducer allows the production of a picture of the seafloor along a constant track line. As one travels over a shipwreck, a vertical “spike” will be produced due to the structural relief of the wreck.

However, while affordable, the use of an echo sounder is not very efficient for finding shipwrecks in the absence of relatively accurate positioning information. Since the energy emitted from an echo sounder’s transducer is emitted vertically, one would basically have to run directly over a wreck in order to locate it with this device alone. In the absence of coordinates that put you “in the ballpark,” one would need to depend on a more efficient alternative than an echo sounder.

Thankfully, sonar has been adapted to cover much larger portions of the seabed. Side scan sonar utilizes similar acoustic transducers to echo sounders, however, side scan sonar transducers are oriented 90 degrees from the path of the survey vessel. By emitting a narrow, focused sonar beam to the sides of the track line, the beam propagates out across the seabed. As it travels outward, the seabed and other obstructions eventually reflect the acoustic energy, which allows the signal to return to the sonar system. The travel time of the acoustic pulses are recorded together with the amplitude of the returned signal as a time series. The recorded amplitudes are a measure of the acoustic backscatter and the reflection from the sea floor, which is dependant on the seafloor type or roughness and the nature of objects on the seafloor. A digital signal processor displays this sonar data, most commonly in a digital format on a personal computer. When a large object, such as a shipwreck, reflects the side scan sonar’s pulse, it produces a “shadow” where the remainder of the acoustic energy is blocked. By employing some relatively simple geometric formulas, it is possible to estimate the size of a scanned object. However, locating and imaging a wreck with a side scan towfish is not as simple as just throwing the instrument over the side and blissfully driving around.

To utilize side scan sonar efficiently, the user must first be aware of the instrument’s limitations. Side scan sonar, while excellent at covering large swaths of the seabed, is not particularly fast. In order to allow adequate signal coverage, the speed of the survey vessel must not exceed the transducer’s rate of emission. As a general rule of thumb, three knots is an ideal survey speed for most side scan units, with four knots a recommended maximum survey speed. While speed has an effect on image quality, it also plays a more direct effect on the side scan towfish. Water drag caused by increased speed will tend to lift the towfish off the bottom, causing a larger blind spot directly under the towfish. If the towfish is lifted high enough, water column disturbance from the survey vessel’s wake will wreak havoc on the towfish. This lifting effect can be controlled a bit with the use of a depression wing. These are attached to the side scan towfish in order to produce negative lift, which will pull the towfish deeper. These devices are particularly useful when there is a limitation on available towfish cable, though it is generally advantageous to have at least a three-to-one ratio of deployable cable when surveying. That is, for every one meter of depth, there should be three meters of cable available. This will allow the towfish to reach a sufficient “altitude” above the seafloor when the survey vessel is in motion. At approximately $15 per meter of towfish cable, this can be a particularly expensive consideration.

Another important consideration when surveying with side scan is the effect that sea surface conditions can have on image quality. A side scan sonar towfish operates best when towed straight and level, which allows proper transducer beam emission. However, when the surface of the sea is rough, as it frequently is, the quality of the sonar image suffers. When the surveying vessel is heaving and rocking due to surface swell, that effect is translated down the cable to the towfish. Heavy sea conditions are not required to negate the effectiveness of a survey. If the seas are running as little as three to five feet when operating off a small vessel, it can cause the towfish to pitch and yaw enough so that the transducers do not project off their intended axis, thus creating gaps in the data. It is possible that this effect can be so pronounced that even large shipwrecks can be missed along the track line. One method to suppress this effect is to utilize a form of “shock absorber” along the tow cable, which helps to minimize the translation of surface conditions to the side scan sonar towfish.

A common query pertains to the range of a side scan sonar unit. This is entirely dependent on the transducer frequency of the towfish. The lower the frequency of the transducer, the greater the range. For example, a 50 Hz pulse can be transmitted several thousand kilometers in the ocean, but a pulse of 300 kHz, a readily-utilized side scan sonar frequency, can be transmitted less than 3,280 feet. However, low frequency acoustic energy does not have the capability to produce high-resolution images. Therefore, a common dilemma is what sacrifice to make with a side scan sonar unit: range or resolution? Fortunately, many companies now produce dual-frequency towfish to help remedy this situation.

As may be expected, side scan sonar is highly effective at finding shipwrecks. Furthermore, this remote sensing tool is also fairly economical, with price ranging anywhere from $7,000 for a lower-end unit, to approximately $30,000 for a dual-frequency towfish and laptop computer package.

A further refinement to basic sonar, or even side scan sonar, is what is known as multi-beam sonar. As mentioned above, side scan sonar emits a single acoustic beam at an oblique angle to the track line. Due to the limitation of this single beam, side scan sonar is limited to a speed that is slow enough to allow the signal processor to survey the bottom. However, by employing several sonar beams simultaneously, speed can be greatly increased without loss of resolution or range. Depending on the number of beams utilized, the speed of the survey is not compromised by limitations of the multi-beam equipment. While some manufacturers offer a multi-beam towfish that is operated similar to a side scan sonar towfish, most multi-beam applications are conducted with a transducer mounted directly to the vessel.

Another advantage of multi-beam sonar is the capability of the data processors to produce a three-dimensional effect in the data presentation. Color is generally added to depict bathymetric contours along the survey, which enhances and facilitates data interpretation. While the texture of the seafloor can be readily depicted in a side scan sonar image, side scan sonar does not directly add the depth and scale reference that multi-beam sonar does. Since the multi-beam transducer array is generally mounted to the survey vessel, accurate positioning information (within 20 inches) can be acquired. Furthermore, survey swaths can be overlapped to produce an overall survey mosaic of total seafloor coverage. Unfortunately, multi-beam sonar systems are dramatically more expensive than side scan sonar systems.

MAGNETIC ANOMOLIES

Another popular remote sensing tool that is utilized in finding shipwrecks is a magnetometer. Anyone that has dived on a shipwreck is probably aware of the effect a wreck may have on one’s compass. The large amount of ferrous metal can throw off the magnet in the compass, thus producing an erroneous reading. Believe it or not, but a diver’s compass is a primitive form of a magnetometer. A magnetometer is a device that measures fluctuations in the Earth’s magnetic field, or, rather, it measures the intensity of the local magnetic field. Ferrous material, such as iron and steel, cause localized magnetic field fluctuations, or anomalies, that may indicate the presence of a shipwreck. The amount of the fluctuation is a direct function on the mass of the object. While magnetometers can easily identify iron- or steel-hulled wrecks, they can also be useful in locating wooden wrecks due to the presence of iron fittings and other ferrous accoutrements associated with the vessel.

There are three main types of magnetometers in use today, the proton precession magnetometer, the Overhauser (OVH) magnetometer, and the alkali vapor magnetometer. The proton precession magnetometer is the simplest and least expensive of the three units, but it also rapidly becoming obsolete. The proton precession magnetometer operates on the principal that protons in atoms spin on an axis aligned with the Earth’s magnetic field. However, when these protons are subjected to an artificially induced magnetic field, the protons will align themselves with the new field. When this new field is interrupted, the protons return to their original alignment with the Earth's magnetic field. As the protons revert their alignment, the spinning protons precess, or wobble, similar to a spinning top as it slows down. This precession frequency, measured in Gammas, is equal to the rate of the ambient magnetic field, and any anomalies caused by shipwrecks or other ferrous material will deviate from that rate.

The sensor component of the proton precession magnetometer is typically a cylindrical container filled with a liquid rich in hydrogen atoms (e.g., distilled water, kerosene, alcohol) surrounded by a copper coil. This coil is energized by a direct current to produce a strong magnetic field. When the current is shut off, the precessing protons induce a very weak signal into the coil, which is connected to a suitable output device. The output can easily be incorporated into a simple computer program run on a laptop, which can also overlay GPS data to record the position of any magnetic anomaly.

The OVH magnetometer is very similar to a basic proton precession magnetometer, but adds a simple technological improvement. While the magnetometers work on the same principle, the difference is found in the sensor design. Instead of simply utilizing a hydrogen-rich fluid, an OVH sensor adds a free-radical to the solution, which produces a two-spin system that enhances the reactiveness of the protons in the fluid to an electrical field. This results in a faster data cycle rate. Additionally, instead of applying a direct electrical current to the sensor, the OVH sensor only requires a low power radio frequency to induce a magnetic field. Therefore, the OVH magnetometer requires less power, is more sensitive, allows for faster sampling rates, and requires a smaller sensor than a standard proton precession magnetometer. Due to these advantages, the OVH magnetometers are largely becoming the standard choice for magnetic anomaly detection, and typically cost less than $20,000.

In contrast to the proton precession and the OVH magnetometers, alkali vapor magnetometers utilize an optically pumped magnetic sensor. Optically pumped magnetometers use alkali metals from the first column of the periodic table (e.g., cesium, potassium) in a gaseous state, but they operate on virtually the same principle as a proton precession magnetometer. Since the alkali metal needs to be gaseous to operate the magnetometer, the cell that contains the metal must be continuously heated to approximately 113° Fahrenheit. These sensors are much more sensitive and have a faster response time than a proton precession magnetometer sensor. However, the cesium magnetometers are extremely sensitive to changes in direction, and, as a result, are largely used on relatively stable platforms such as aircraft. Furthermore, as may be expected, anything that is faster and more accurate is likely to be more expensive, and an alkali vapor magnetometer is approximately twice as expensive as an OVH magnetometer.

Successful operation of a magnetometer relies on numerous considerations. Since the magnetometer is sensitive to magnetic fields, it is prudent to isolate the sensor as far away from the vessel as possible, in order to avoid any disturbance from the vessel’s engines and metal hull, if applicable. This is typically accomplished by utilizing a towfish, similar to a side scan sonar unit. Another advantage to using a towfish is that places the sensor in closer proximity to potential minor anomalies, such as wooden shipwrecks; distance is a critical factor in detecting magnetic field anomalies. However, a towfish requires sufficient cable to deploy the sensor at least four boat lengths behind the vessel or at a sufficient depth, and it also negatively affects the speed of the survey. Another alternative to a towed sensor is to mount the sensor directly to the boat, preferably as far away from the engine as possible. Obviously, this option is only viable when using a conventional fiberglass-hulled boat. This arrangement allows the vessel to survey at a high rate of speed, and also provides accurate geographic positioning. However, the drawback to this format is the greater distance from the anomaly in comparison to a towed sensor, which becomes more important when one is looking for shipwrecks that may produce a small anomaly or are located in deeper water. Regardless, if one is looking for large twentieth century shipwrecks, a magnetometer is an affordable and relatively efficient survey method.

LASERS

In contrast to the particularly low-tech and inexpensive magnetometer, expensive laser technology is perhaps isolated to academic and governmental research agencies. Regardless, lasers are readily being used, and with remarkable results. Two of the more interesting applications are light detection and ranging (LIDAR) and laser-line scan (LLS).

LIDAR is similar to RADAR, in that is transmits and receives electromagnetic radiation. However, LIDAR does so at higher frequencies than RADAR, typically at the ultraviolet end of the spectrum. A laser pulse is transmitted from an airborne platform to the surface below. A large portion of that pulse is reflected by the water’s surface, while the remainder of the energy penetrates the water column, eventually to be reflected by the seabed. Absorption, refraction, and scattering all combine to affect the laser’s penetration depth. Generally, LIDAR works best in clear water; the maximum penetration depth of most LIDAR systems is approximately 160 feet, or two to three times the secchi depth (a measurement of turbidity). Since the speed of light is a known constant, the depth of the water can be calculated by measuring the time interval between the laser’s reflection off the water’s surface and the reflection from the seabed. This information, along with DGPS data from the aircraft, can be utilized to generate very precise bathymetric mapping. Since the resolution of LIDAR is on the order of 10 - 20 inches, any dramatic spike in depth that is produced from a shipwreck would be very pronounced.

Due to the high speed, high resolution, and ability to cover large areas of real estate in a relatively short period of time, this system is highly effective in mapping benthic habitat. However, the depth of LIDAR is generally restricted to shallower water. While it is possible to increase the laser strength of the LIDAR to increase penetration, it is not practical, especially considering the deleterious effects that a powerful laser may have on a recreational vessel and its passengers. As may be expected, this remote sensing application is fairly expensive. Although most wreck hunters may not have access to a LIDAR system, it is possible for an industrious researcher to review LIDAR data from various surveys conducted by government agencies and academic institutions. Most of these surveys are oriented toward coastal processes (e.g., beach erosion) or benthic habitat mapping, however, there is a potential to find information on shipwrecks that were inadvertently imaged during the LIDAR survey. These avenues should be explored, especially considering that LIDAR surveys are more frequently employed in an ever-expanding range of disciplines.

While LIDAR is particularly well suited at imaging large areas to accurately document bathymetry, LLS allows researchers to map discrete areas, such as a shipwreck, in amazing detail. An electro-optic imaging technique, LLS has the ability to produce high contrast underwater images at millimeter scale resolution, and it has as much as five times the range as conventional video and photographic systems; the performance of LLS far exceeds acoustic survey techniques such as side scan and multi-beam sonar. Unfortunately, in order to obtain these high-resolution images, LLS needs to survey at a relatively low velocity, with typical tow speeds of two to three knots. Due to the cutting-edge technology of this remote sensing application, LLS is extremely expensive at approximately $700,000, and is therefore confined almost entirely to governmental agency and academic institution projects.

REMOTELY OPERATED AND AUTONOMOUS UNDERWATER VEHICLES

A remotely operated vehicle (ROV) is an underwater robot that allows an operator to remain in a safe environment while the ROV works in the hazardous ocean environment. An ROV basically consists of the vehicle and an umbilical that connects the vehicle to the surface and the operator. The ROV can greatly vary in size and cost, depending on the tasks it is designed for and the equipment it carries. A new generation of low-cost ROVs may be particularly appealing to those searching for shipwrecks. Some of these systems consist of nothing more than thrusters attached to a conventional video camera housing, while others are a bit more refined. These low-cost units range anywhere from $10,000 for a primitive system, to over $40,000 for platforms with moveable cameras and attached lighting.

More complex and expensive ROVs have the capability to actually conduct work at depth. Manipulator arms and other specialized extensions can be added to the vehicle chassis that enable the ROV to perform tasks other than just relaying images to the surface. Other remote sensing applications, such as multi-beam sonar or LLS, have been incorporated into some ROVs, allowing for a more dynamic and functional ROV. The limitation to many ROVs is the need for a connection to the surface via an umbilical. However, technology has advanced to a point where that umbilical can be severed, freeing the vehicle to conduct unfettered research. These vehicles, close relatives to ROVs, are known as autonomous underwater vehicles (AUVs).

While development of AUVs began in the early 1960s, they have only recently become practical. Tasks are completed by an AUV through internal computer commands and extensive software controls, or by real-time communication from the support vessel via an acoustic link. An obvious advantage of an AUV over an ROV is the lack of the umbilical. This allows the AUV to work around the clock, even in rough weather. Since an ROV requires an umbilical to receive commands, heavy surge and deteriorating weather would require the extraction of the vehicle. However, an AUV can operate regardless of surface conditions. When combined with other applications, such as side scan or multi-beam sonar, an AUV can be directed to survey vast areas of the seabed, concluding only when power supplies require recharging. When pre-programmed, an AUV can conduct its assigned tasks while researchers can focus on other work. This flexibility is especially attractive when at-sea time is limited. As with any other beneficial tool, there are always drawbacks. One of the most glaring to the amateur wreck hunter is price: AUVs are extremely expensive. Another drawback is the relative infancy of this technology. Additionally, should an AUV have a catastrophic failure, or even a minor failure, it is possible the vehicle will be lost. Regardless, AUVs are definitely on the cutting edge of remote sensing technology, and will undoubtedly be refined and advanced.

THE FUTURE IS NOW

So where will remote sensing applications take us in the future? Bold predictions by at least one well-known oceanographer has forecast that it will soon be unnecessary for man to enter the ocean at all. While this may be particularly troublesome for avid divers, perhaps there is a bit of credibility to this statement. Already it is possible to find and image a shipwreck without getting wet. Furthermore, it is quite possible that the removal of the “human factor” may be highly likely for some disciplines, such as underwater archaeology. High-resolution side scan sonar and LLS has the ability to produce an extremely detailed site plan of a wreck in a relatively short amount of time. Should tasks need to be accomplished or samples taken, ROVs, with a virtually unlimited bottom time, can quickly be deployed. Even the use of space-borne platforms are not too far-fetched. This is illustrated by an incident that recently occurred at Marine Corps Base Quantico, after a reconnaissance satellite detected several submerged anomalies in Quantico Creek. After ground-truthing the sites, it was determined that two of the anomalies the satellite detected were Civil War-era shipwrecks. While it may currently not be feasible for the general public to utilize satellites for identifying submerged cultural resources, it is not outside the realm of possibility in the near future.

It is obvious that no one remote sensing tool is a panacea; every system has advantages, as well as drawbacks. However, by incorporating some of these tools along with basic archival research, it is possible to improve your success rate at locating and documenting shipwrecks. If nothing else, remote sensing applications prove that it is a small world after all.