Mapping a Forest With Sound: Echosounders and the 3D Shape of Kelp

A satellite sees the roof of a kelp forest. A diver swims through a doorway of it. But a kelp forest is a building — it has floors, volume and load-bearing structure — and almost none of that structure is visible from above or fully grasped from within a single transect. The instrument that can read the whole volume in one pass is the echosounder: a downward-looking sonar that turns the water column between canopy and seafloor into numbers. This is the physics of how kelp announces itself to sound, how surveys turn echoes into canopy maps, and — just as important — the biases that keep sonar readings honest only when the people running them know what the numbers can't see.

If you want the wider map of acoustic survey tools, our primer on sonar reef mapping covers how the seafloor itself is drawn. This article is about the upside-down half of the problem: the living column of water above the rock.

Why kelp echoes at all

Sound returns from water only where it meets something that differs acoustically from water. Flesh differs so little that a smooth-skinned fish is a modest target. Kelp's signature is threefold:

• Tissue itself. Frond tissue is denser and more elastic than seawater. It scatters sound weakly, but a dense stand of fronds scatters enough to register on a sensitive echosounder as a textured band between surface and bottom.
• Pneumatocysts — the gas floats. Many kelp species buoy their fronds with millions of tiny gas bladders. Gas is acoustically loud: a small bubble near the resonant frequency of the sonar ring can scatter sound tens of decibels more strongly than an equivalent parcel of tissue. To an echosounder at the right frequency, a forest of pneumatocyst-bearing kelp is a forest of ring-mounted bells.
• Angle and density. A canopy lying flat at slack tide presents a thin target to a vertical beam; a canopy standing up in a current presents its whole blade face. Kelp's acoustic target strength changes with the shape of the forest, not just its mass — which is both the method's richness and its favourite trap.

Different species sing differently. Giant kelp's fronds, buoyed throughout the water column by small pneumatocysts, return a full-column signal from surface to bottom attachment. A bull kelp with one large float tends to ring once, near the surface. A bark kelp with no floats at all returns only tissue scatter — far quieter, far easier to miss. Species-blind acoustic surveys inherit these blind spots whole.

The four ways to listen

Downward-looking echosounder: the workhorse transect

A single-beam or split-beam echosounder on a pole over the side of a small boat, driven in tight parallel lines across the forest, builds the classic canopy survey. Every ping returns a vertical echo profile — one trace of "what is at each depth beneath the boat." Stitched along the boat's GPS track, thousands of traces become a cross-section of the forest: canopy top, canopy base, bottom depth, and the textured density band between them.

Split-beam models add position within the beam, so that individual targets (a school of anchovies inside the canopy; a diving cormorant) can be separated from the kelp mass.

Upward-looking sonar: the forest's own weather station

An echosounder bolted to the reef and pointed up can log canopy presence and height for months — storm after storm, tide after tide. Where a boat transect gives you a snapshot, a moored upward-looker gives you a time series: canopy collapse in a heatwave in real hours, canopy return in the spring in real days. Restoration programs plant these the way meteorologists plant rain gauges; our Drifter vessels drop them on seeded reefs to read the forest's growth between visits.

Multibeam water-column data: the forest in full geometry

A multibeam sonar built for seafloor mapping also records echoes arriving in the water above the bottom. Processing that water-column data is heavier engineering — every ping is a three-dimensional fan of returns — but the payoff is a true 3D grid of scatter strength: the forest as a volume, with canopy surface texture, understory gaps and fringing schools of fish all in one geometry.

Sidescan: the extent, not the height

Where is the kelp bed? A sidescan towfish answers the horizontal question with aerial-photograph clarity — a kelp bed's textured grain looks different in sidescan from bare rock, sand or urchin pavement. Sidescan tells you the forest's extent superbly and its height not at all; paired with an echosounder transect it gives you both.

From pings to canopy height

Turning echo traces into ecological numbers is a calibration exercise, not a computation:

1. Find the bottom. Every trace's bottom echo anchors the scale. Multibeam systems do this at survey-grade precision; small-boat echosounders use the strong bottom return as the reference height from the seafloor.
2. Detect the canopy edge. A threshold above the bottom — in echo strength and in height — flags each trace as canopy or no-canopy. The resulting binary, line by line, is the forest's footprint.
3. Measure the depth at which canopy echoes begin. Canopy height above bottom, trace by trace, becomes a canopy surface model — the roofline.
4. Ground-truth against divers. A diver measuring a kelp stand's actual height and density at a known position calibrates what "echo strength X" means in stems per square metre or kilograms of wet kelp. The entire quantitative chain rests on this step, which is why acoustic surveys and diver transects travel together rather than competing.

What falls out of the process is a suite of numbers that means something: canopy cover (what fraction of reef has kelp overhead), canopy height (the forest's stature, averaged and gridded), canopy rugosity (how textured the roofline is — a wave-damping parameter all its own), and — through species-specific allometric equations — standing biomass, the number climate and harvest accounting both need.

What can fool the instrument

Every technology announces its honesty by how well it names its biases. Four matter most in kelp acoustics:

• Acoustic shadowing. A dense surface canopy absorbs and blocks what passes below it. In a heavy forest the echosounder may correctly draw the roof and miss the mid-water fronds underneath, understating standing biomass.
• Tide and swell geometry. Canopy angle with the incoming sound changes the apparent strength. A trace taken at flood stream and one taken at slack water are different measurements of the same forest. Disciplined surveys either fix the tide window or model the angle effect.
• The fish problem. A school of small fish inside the canopy is acoustically loud and easily mistaken for kelp — or, worse, aggregated with the kelp signal. Split-beam position data and time-of-day discipline (fish move at dawn and dusk through the canopy; kelp stays home) filter most of it, but the residual error is real.
• Species blindness. Quiet-kelp species — those without gas floats — slip under acoustic attention. Some of the most ecologically important understory kelps are nearly sonar-invisible, which is why satellite canopy mapping and echosounder work are complements, not competitors: the satellite sees the surface of all species; the sounder sees structure but hears flotation loudest.

Why three dimensions matter

A canopy map is not a portrait; it is an engineering drawing. The volume and texture of a kelp forest drive three things every restoration program cares about:

• Wave damping. The forest's drag on moving water is a function of its frontal area in three dimensions — the physics of how kelp calms the sea needs the canopy's vertical geometry, which only acoustics (or patient divers) can supply at reef scale.
• Habitat volume. Fish, octopus and a hundred invertebrate species use the forest as a three-dimensional space. A flat two-hectare canopy of short kelp and a standing canopy of the same area are different cities, and the difference shows up first in the acoustic traces.
• Carbon and harvest accounting. Whether a project's ledger is about sequestration pathways or a sustainable harvest, biomass is the number every model consumes. Echosounder-derived estimates, calibrated against diver-measured quadrats, are the only repeatable way to census standing stock over hectares rather than metres.

When our canopy analysts talk about "reading the forest in 3D" inside the Canopy Ledger platform, this is the packet of measurements they mean — the echo-derived volume that turns satellite pixels into a building you can quantify.

The future: listening harder

The next decade of kelp acoustics is less about new instruments and more about new ears. Machine-learning classifiers trained on labelled echo traces are learning to separate kelp species by their acoustic texture, to flag fish contamination automatically, and to fuse sonar with drone photogrammetry of the surface into unified above-and-below models. The forest will remain a building that light cannot enter and divers cannot census — but sound, arriving ping by ping, keeps adding rooms to the map.

Key takeaways

• Kelp echoes because of gas — pneumatocysts are acoustically loud, so float-bearing and floatless species return very different signals to the same sounder.
• Four tools, four questions — downward echosounders measure canopy height; upward-lookers log change through time; multibeam water-column data builds true 3D structure; sidescan maps bed extent.
• Calibration against diver measurements is the entire budget of truth — without it, echo strength is a colour, not a number.
• Tide, fish schools and acoustic shadowing are the standing biases — disciplined survey design controls them; undisciplined surveys bake them into the map.
• 3D structure is the variable that matters downstream — wave damping, habitat volume and biomass all depend on the forest's vertical geometry, not just its footprint.

FAQ

Can sonar identify kelp species?

Partially, and getting better. Float architecture gives strong species-level hints — a single large float reads differently from a column of small ones — and machine-learning classifiers trained on labelled traces are beginning to separate common species by echo texture. But for now, species confirmation still comes from divers or imagery, with acoustics providing the architecture.

Does echosounding disturb the kelp or its inhabitants?

At survey power levels and durations, no measurable disturbance is documented for kelp; the pings are brief and the sound levels far below anything associated with harm to fish or invertebrates. Marine mammals are the caution case for any loud underwater source, which is why survey protocols include marine-mammal watches and exclusion zones, same as for fish survey acoustics.

How accurate is a biomass estimate from acoustics?

Calibrated against diver-measured density on the same reef and day, canopy height estimates run within centimetres and biomass within roughly the error bars of the diver calibration itself. Uncalibrated, an acoustic number is a relative index — reliable for tracking change, unreliable as an absolute figure.

Why not just use drones for all this?

Drones see the canopy surface superbly — and nothing underneath it. The mid-water structure, the canopy's depth, and the bottom attachment that anchors every height measurement are all invisible to a camera. Drone photogrammetry and acoustics together are the whole survey; either alone is half a building.