How Sonar Maps the Seafloor

As of the GEBCO Seabed 2030 project's 2025 progress report, 27.3 percent of the world's ocean floor has been mapped to modern high-resolution standards, up from 26.1 percent the year before. Nearly all of that mapping, and nearly everything humans know about the shape of the seafloor at all, was done with sound rather than light or radar, because seawater absorbs light within roughly 200 meters (656 feet) while sound can travel for tens of kilometers. Sonar, an acronym for Sound Navigation and Ranging, turns the travel time of a sound pulse into a depth measurement, and the same basic trick that once located icebergs and submarines now builds three-dimensional maps of trenches, volcanoes, and shipwrecks kilometers below the surface.

Why sound, not light or radar

Radio waves and visible light are both forms of electromagnetic radiation, and seawater is remarkably good at destroying both. Even in the clearest tropical water, sunlight is scattered and absorbed within the top 200 meters, the boundary oceanographers call the euphotic zone, below which photosynthesis becomes impossible and cameras need their own light source to see anything at all. Radar suffers the same fate almost immediately, since conductive, ion-rich seawater absorbs radio-frequency energy within centimeters of the surface, which is precisely why submarines are invisible to airborne radar the moment they submerge.

Sound behaves in an entirely different way underwater because it does not travel as an electromagnetic wave at all. It travels as a mechanical pressure wave, a rhythmic compression and rarefaction passed from one water molecule to the next, the same basic phenomenon that lets sound travel through air, steel, or the ground during an earthquake. Water molecules are packed roughly a thousand times more densely than air molecules, and that density lets a pressure wave propagate through water far more efficiently and far faster than it can through air, with only gradual energy loss over long distances.

That efficiency is the entire reason ocean-depth measurement is a sound problem rather than a light problem. A pulse of sound generated at the surface can reach the bottom of the Mariana Trench, nearly 11 kilometers (6.8 miles) down, bounce, and return with enough energy left to be detected and measured, something no beam of light or radar pulse can come close to doing. It is worth being precise about what "sonar" actually names: it is not one instrument but a family of techniques, all built on the same underlying principle of timing a sound pulse's round trip, ranging from a simple single-beam echo sounder on a fishing boat to a synthetic aperture array towed behind a survey ship.

Langevin, the U-boat war, and the birth of active sonar

The technology traces to a specific and well-documented crisis: German U-boats were sinking Allied shipping faster than it could be replaced during the First World War, and no existing technology could detect a submerged submarine before it fired a torpedo. French physicist Paul Langevin, working with Russian émigré engineer Constantin Chilowsky and the Ecole de Physique et Chimie de la Ville de Paris, spent 1915 to 1918 developing a device that could both transmit and receive pulses of ultrasound underwater using piezoelectric quartz, crystals that generate a tiny electric voltage when mechanically stressed and, in reverse, vibrate mechanically when an electric voltage is applied to them (Physics Today, 2022; ethw.org, "Milestones: Invention of Sonar, 1915-1918").

Piezoelectric transducer — a device that converts electrical energy into a physical vibration, and physical vibration back into an electrical signal, which is what lets a single quartz crystal both emit a sound pulse and later "listen" for its echo. In February 1917, Langevin first used a quartz blade as a receiver; four months later, the same quartz worked as a transmitter once his collaborator Marcel Tournier solved the technical problem of rapidly switching a single crystal between emitting and receiving. By February 1918, Langevin's team was successfully detecting echoes of an ultrasonic pulse bounced off a target, and on May 4, 1918, the system detected the submarine Messidor at a range of 500 meters (1,640 feet) — the first confirmed active-sonar detection of a submerged vessel using piezoelectric ultrasound (cnrs-scrn.org, Northern Mariner, "Paul Langevin and the Discovery of Active Sonar or Asdic").

Just before the November 1918 armistice, the Allies formalized this work by creating the Allied Submarine Detection Investigation Committee, which gave the resulting British technology its enduring name: ASDIC. The war ended before ASDIC saw widespread combat use, but the underlying physics, an electrically driven quartz or later ceramic transducer emitting a directional pulse and timing its echo, became the direct ancestor of both military sonar and civilian depth sounding.

The first confirmed underwater sonar detection of a submarine happened on May 4, 1918, when Paul Langevin's piezoelectric quartz system located the French submarine Messidor at a range of 500 meters (1,640 feet) — seven months before the war that motivated it had ended.

A parallel, independently motivated thread of development ran through civilian shipping rather than naval warfare. The sinking of the RMS Titanic on April 14, 1912, after it struck an iceberg that lookouts spotted too late to avoid, convinced several engineers that ships needed a way to detect large underwater obstacles before visual contact. Within a week of the disaster, English meteorologist Lewis Fry Richardson filed a patent for using echoed sound to determine distance in air, and filed a second patent a month later for doing the same thing underwater. Canadian inventor Reginald Fessenden built a working system in 1914 that used an oscillator to detect a 130-foot (40-meter) iceberg at a range of more than two miles (3.2 kilometers) and also successfully measured seafloor depth at 31 fathoms (186 feet), demonstrating that the same iceberg-detection principle worked equally well for depth sounding (dosits.org, Discovery of Sound in the Sea, "The First Practical Uses of Underwater Acoustics: The Early 1900s"). German physicist Alexander Behm patented his own echo sounder design in 1913. Two unrelated emergencies, one a maritime disaster and one a submarine war, converged on the same physics within a few years of each other, which is why sonar's origin story genuinely has two separate starting points rather than one clean invention narrative.

How echo sounding turns time into depth

Every sonar system, from the simplest fish finder to a research-grade multibeam array, rests on one equation. A transducer emits a short pulse of sound, a hydrophone (an underwater microphone) listens for the echo bouncing off whatever the pulse hit, and the system measures the round-trip travel time. Depth is then calculated as d = (v × t) / 2, where d is depth, v is the speed of sound in water, and t is the total round-trip travel time; the division by two accounts for the fact that the pulse travels the distance twice, once down and once back up.

The speed of sound in seawater, v in that equation, is not a fixed constant, and getting it wrong produces depth errors that compound over an entire survey. Sound travels through seawater at approximately 1,500 meters per second (4,921 feet per second, or roughly 3,375 miles per hour), about 4.3 times faster than the roughly 343 meters per second it travels through air at sea level and room temperature. But that figure shifts meaningfully with three variables: temperature, salinity, and pressure. Warmer water, saltier water, and higher-pressure (deeper) water all transmit sound faster, and across the range of conditions found in the open ocean, sound speed can vary from roughly 1,450 to 1,540 meters per second depending on where and how deep the measurement is taken.

Because that variation is large enough to matter, precision hydrographic surveys do not simply assume 1,500 meters per second and move on. Survey teams lower an instrument called a sound velocity profiler into the water column at the start of a survey, and sometimes repeatedly throughout it, to directly measure how sound speed changes with depth at that specific location on that specific day. That profile then corrects every subsequent depth calculation, layer by layer, for the water's actual acoustic properties rather than a textbook average — the difference between a rough estimate and a survey-grade chart that other ships and cable-layers can trust with their lives and equipment.

Why can't sound speed just be assumed constant?

Assuming a flat 1,500 meters per second across an entire water column can introduce errors of tens of meters in deep water, because sound genuinely bends and speeds up as it crosses layers of different temperature and salinity, an effect called refraction. Real surveys measure the actual sound-speed profile with a dedicated instrument rather than relying on a single average value.

From a single ping to a swath: single-beam versus multibeam

The earliest generation of echo sounders, and the cheap depth finders still standard on recreational and small fishing boats, use single-beam sonar: one transducer sends one pulse straight down and records one depth measurement directly beneath the hull. That is enough to avoid running aground, but it means mapping a stretch of seafloor requires steering the boat back and forth in a dense grid, one narrow line of soundings at a time, with wide gaps of un-surveyed seafloor between each pass.

Modern bathymetric mapping is built almost entirely on multibeam sonar instead. First developed for military and research applications starting in the 1960s and standardized on oceanographic survey vessels through the 1980s and 1990s, a multibeam system sends out a fan-shaped array of dozens to several hundred individual sound beams in a single ping, angled outward from directly beneath the ship to well off to each side. Each of those beams returns its own independent depth measurement, so a single pass produces not one point but an entire cross-sectional swath of the seafloor, typically several times wider than the water is deep. A survey ship in 2,000 meters (6,562 feet) of water might map a swath 6 to 8 kilometers (3.7 to 5 miles) wide with a single line of travel, instead of the single narrow track a single-beam system would leave behind.

Sonar type	Coverage per pass	Typical resolution	Primary real-world use
Single-beam echo sounder	One depth point directly below the vessel	Coarse, single-line	Basic depth soundings, recreational and small fishing boats
Multibeam sonar	Fan-shaped swath, dozens to hundreds of points per ping	Meter-to-sub-meter scale, full seafloor coverage	Detailed bathymetric mapping, nautical charting, trench surveys
Side-scan sonar	Wide swath, angled outward and down	High-detail imagery, less precise depth	Shipwreck, cable, and seabed texture surveys
Synthetic aperture sonar	Narrow swath, signal-processing gain	Very high-resolution imagery independent of range	Mine detection, archaeological surveys, small-object identification

Multibeam systems dominate large-area bathymetric mapping today, while side-scan and synthetic aperture sonar trade depth precision for far sharper imagery of objects and textures on the seafloor.

Multibeam data from repeated, overlapping ship passes gets processed with specialized hydrographic software that converts each beam's return angle and travel time, corrected using the sound velocity profile described above, into a precise three-dimensional point in space. Stitching thousands of these swaths together over weeks or months of survey time produces a continuous bathymetric map, the underwater equivalent of a topographic map on land. The deepest confirmed point in the ocean, Challenger Deep in the Mariana Trench, was measured at approximately 10,935 meters (35,876 feet) below sea level by the 2021 Five Deeps Expedition using modern multibeam sonar — a figure that replaced far rougher early 20th-century estimates made with wire-and-weight soundings, a method that could take hours to complete a single measurement and carried substantial uncertainty from cable drift and stretch.

Side-scan and synthetic aperture sonar: painting a picture, not just a depth

Multibeam sonar is excellent at answering "how deep is the water here," but it is not the tool of choice when the question is "what does an object on the seafloor actually look like." For that job, surveyors turn to side-scan sonar, which points its beams outward and slightly downward on either side of a towed sensor rather than straight down, and builds an image based on how strongly different surfaces reflect sound back toward the receiver. A hard, angular object like a shipwreck's hull reflects sound very differently than soft sediment or a rippled sand bottom, and side-scan sonar renders those differences as a grayscale image that looks remarkably like an aerial photograph, even though not a single photon was involved in producing it.

This distinction matters practically: a multibeam survey could pass directly over a sunken ship and record it only as an unusual bump in the depth data, while a side-scan pass over the same wreck can produce a recognizable image showing the hull's outline, deck structures, and debris field. Side-scan sonar has been the standard tool for shipwreck discovery, submarine cable and pipeline inspection, and mapping geological seabed features like sediment ripples and rock outcrops that a straight-down depth reading alone would simply miss.

Synthetic aperture sonar pushes that imaging capability further by using signal processing rather than raw pulse strength to build resolution. Instead of relying purely on how tightly a physical transducer can focus a beam, synthetic aperture sonar combines the returns from many pulses collected as the sensor moves along its track, mathematically synthesizing the resolution of a much larger physical array. The result is imagery sharp enough to identify small, human-scale objects on the seafloor, such as sea mines or small archaeological artifacts, at ranges and resolutions that conventional side-scan sonar cannot match. That capability is exactly what search teams reached for when conventional bathymetry alone wasn't enough to resolve a target with confidence.

Why this matters: mapping the search for MH370

The most extensive real-world demonstration of modern sonar's capability came, unintentionally, from a tragedy. When Malaysia Airlines Flight 370 disappeared over the southern Indian Ocean on March 8, 2014, investigators faced a region of seafloor that was almost entirely unmapped at any useful resolution. Between 2014 and 2017, the search effort commissioned what became the largest bathymetric survey ever conducted in that part of the Indian Ocean, acquiring roughly 710,000 square kilometers (274,000 square miles) of shipboard multibeam data, an area comparable in size to France or the state of Texas, plus more than 120,000 square kilometers (46,000 square miles) of higher-resolution data gathered by side-scan, multibeam, and synthetic aperture sonar mounted on towed sleds and autonomous underwater vehicles (Geoscience Australia, "MH370 - Data release"; ATSB, "Mapping the ocean floor — Bathymetric survey - MH370").

The improvement in mapping resolution the search produced was dramatic by any standard. Before the search, satellite-derived estimates of that region's seafloor averaged roughly 100 kilometers (62 miles) of horizontal resolution, essentially a rough guess at where major features sat. The multibeam survey pushed horizontal resolution down to about 0.1 kilometers (100 meters, or 328 feet), a thousand-fold improvement, and vertical resolution improved more than twenty-fold, from over 100 meters of uncertainty down to roughly 5 meters (16 feet) (ScienceDirect, "Malaysia Airlines flight MH370 search data reveal geomorphology and seafloor processes in the remote southeast Indian Ocean," 2017). The aircraft itself was never located within the primary search zone, but the sonar data produced an enduring scientific legacy: researchers used it to identify previously unknown seafloor volcanoes, fault scarps, and a set of long ridge-and-swale features nicknamed "corrugations," offering new insight into how oceanic crust deforms far from any plate boundary (Eos.org, "Search for MH370 Revealed Ocean Crust Waves"). It is a useful reminder that a targeted survey built to answer one urgent question often ends up permanently expanding the map for everyone who comes after.

What sonar still can't do

Sonar's limitations are worth stating plainly, because they explain why more than 70 percent of the ocean floor remains unmapped at high resolution even after more than a century of the technology's development. Multibeam and side-scan surveys require a ship, towed sensor, or autonomous vehicle to physically travel over the area being mapped, moving at only a few knots while collecting data, which makes covering the full 361 million square kilometers (139 million square miles) of ocean an enormously slow and expensive undertaking compared with satellite imaging of dry land. Deep water also attenuates and scatters sound over long ranges, so surveying the deepest trenches at high resolution requires either operating close to the seafloor with an autonomous underwater vehicle or accepting a coarser resolution from a surface ship's longer-range beams.

Satellite altimetry offers a partial workaround by inferring rough seafloor topography from tiny variations in sea-surface height caused by the gravitational pull of underwater mountains and trenches, but that method's resolution is far coarser than direct sonar surveying and cannot resolve individual small features like a shipwreck or an isolated seamount. That gap between the roughly 27 percent mapped in genuine sonar-survey detail and the satellite-estimated remainder is exactly the gap Seabed 2030 is organized around closing, by coordinating and centralizing sonar survey data from research institutions, navies, and commercial shipping companies worldwide into a single freely available global grid.

The takeaway

Sonar maps the seafloor by timing how long a pulse of sound takes to travel down, bounce off the bottom, and return, then converting that round-trip time into depth using the actual, measured speed of sound in that specific water column. The physics behind that trick dates to Paul Langevin's piezoelectric quartz transducers, built under wartime urgency between 1915 and 1918, and to the parallel, Titanic-driven echo-sounding patents filed within weeks of the 1912 disaster. Multibeam and side-scan sonar scaled that single-ping principle into fan-shaped swaths and photograph-like imagery, technology precise enough to confirm the ocean's deepest point and, in the search for MH370, to remap an area the size of France in unprecedented detail. Roughly a quarter of the global seafloor has now been mapped this way, and closing the rest of that gap remains a matter of ship-hours and funding rather than any missing physics. Related reading: how gyroscopes resist falling over, how gravitational lensing magnifies galaxies, and how astronomers weigh galaxies.

FAQ

Q: Why does sonar use sound instead of light to map the ocean floor?

Seawater absorbs and scatters visible light within about 200 meters and destroys radio-frequency radar within centimeters, but sound travels through water as a mechanical pressure wave with far less attenuation, letting sonar pulses reach and return from depths of thousands of meters that light and radar could never survive.

Q: How does sonar calculate ocean depth?

It measures the round-trip travel time of a sound pulse sent to the seafloor and back, then applies d = (v × t) / 2, multiplying the travel time by the speed of sound in water (roughly 1,500 meters per second, though this varies with temperature, salinity, and pressure) and dividing by two to isolate one-way depth.

Q: Who invented sonar and why?

French physicist Paul Langevin built the first working piezoelectric quartz sonar system between 1915 and 1918 to detect German U-boats during the First World War, successfully locating the submarine Messidor at 500 meters on May 4, 1918. A separate, civilian-motivated thread of echo-sounding patents, filed by Lewis Fry Richardson and others within weeks of the 1912 Titanic sinking, pursued the same physics for iceberg and depth detection.

Q: What is the difference between single-beam and multibeam sonar?

Single-beam sonar records one depth reading directly beneath the vessel per ping, requiring a dense grid of passes to map an area. Multibeam sonar sends a fan of dozens to hundreds of simultaneous beams, mapping a wide swath of seafloor with every single pass instead of one narrow line.

Q: How much of the seafloor has been mapped?

As of the GEBCO Seabed 2030 project's 2025 release, roughly 27.3 percent of the global ocean floor has been mapped to modern high-resolution sonar standards, with most of the remainder still relying on coarser, satellite-derived depth estimates.

Q: How was the deepest point in the ocean measured?

Challenger Deep in the Mariana Trench, measured at approximately 10,935 meters by the 2021 Five Deeps Expedition, was surveyed using modern multibeam sonar, which offers far greater precision than the hours-long, cable-based wire-and-weight soundings used in early 20th-century surveys.

Q: What sonar technology was used in the search for MH370?

The 2014-2017 search used shipboard multibeam sonar for broad bathymetric coverage across roughly 710,000 square kilometers of the southern Indian Ocean, plus side-scan and synthetic aperture sonar towed on underwater vehicles for higher-resolution imaging, improving horizontal seafloor resolution in the region a thousandfold.