Single-beam or multibeam for shallow-water survey?

The three options, honestly

Single-beam echosounder (SBES). One narrow beam, straight down, one depth per ping. You get a profile along whatever line the vessel drives. Cheap, simple, and genuinely well suited to very shallow water and to repeat measurements along fixed transects. The data is easy to interpret and fast to edit - a real advantage when one person is doing both the survey and the processing. What you do not get is anything between the lines. You interpolate, and you hope nothing interesting sits in the gap.

Multibeam echosounder (MBES). A fan of beams across track, tens to hundreds of depth points per ping, producing a continuous swath as the vessel moves. Run overlapping lines and you have complete coverage of the bottom rather than a sample of it. More data, more processing, higher cost - and the only honest answer when the deliverable is a volume, a surface, or “show me everything down there”.

Contracting a crewed survey vessel. Still the right answer for deep water, open coast, or anything requiring a certified hydrographic deliverable to IHO order. For a harbour basin, a quarry lake or a 40-hectare inshore site, mobilising a launch and crew has become hard to justify against a USV that one person carries to the water.

What it costs, and what you get

Prices are SepcoTech’s own EUR figures, excluding VAT and shipping, and are read from the same price list the product pages quote.

Both units run on the same free software (Cerulean SonarView), which matters more than it sounds - it means the processing workflow does not change if you start with one and add the other.

The 10× figure is worth unpacking, because it is not marketing. A single-beam at 5° in 10 m of water insonifies a patch under a metre across. An 80° swath in the same depth covers roughly 17 m. You are not driving ten times faster; you are driving roughly a tenth as many lines.

What the swath and the angular resolution actually buy you

Two numbers decide whether a multibeam is useful to you, and they are not the frequency.

Swath width sets your line spacing, and therefore your survey time. The Surveyor 240-16 transmits 80° across track. In shallow water that is the binding constraint on productivity - 80° in 5 m of water is about 8.4 m of coverage, so you are still driving a lot of lines in very shallow ground. Swath scales with depth, which is why multibeam economics improve as the water deepens and why the single-beam argument is strongest under about 5 m.

Angular resolution sets how fine the detail is across that swath. Conventional beamforming on a 16-element array gives roughly 7°. The Surveyor instead uses angle-of-arrival estimation, which resolves to about 1°. In practical terms that is the difference between knowing there is something there and knowing its shape.

Note what that does not mean. Sixteen receive channels is not 256 beams. The Surveyor detects 10–20 points per ping, not hundreds. Point density is built up by slowing the vessel and running overlapping passes, not by the sensor alone. A survey-grade system from an established hydrographic vendor will give you far more points per ping - and cost accordingly.

The specification that trips people up: range

You will see three different range figures for the same instrument, and they are all true.

• 300 m is the housing depth rating. It is how deep the unit survives, not how deep it sees. An AUV can carry it to 300 m; it will not map the bottom from the surface at 300 m.
• 100 m is the maximum suggested range setting. It is a ceiling on the control, not a promise of detection.
• 50 m is where Cerulean says points are detected in most situations, falling to around 30 m from the target plane in compromised conditions - soft bottom, heavy pitch and roll, acoustic interference.

Plan on the 30–50 m figures. Several reseller listings quote “depths up to 100 metres”, which repeats the setting ceiling as if it were a capability, and at least one still carries an older revision of Cerulean’s own table showing 50 m as the maximum suggested range. If a supplier quotes you a number, ask which of the three they mean.

Range also costs you ping rate, and that is physics rather than a product limitation. A ping cannot be transmitted until the last one has come back, so the maximum rate falls as you extend range: about 20 Hz at short range near 20 m, dropping to roughly 6 pings per second at the longest range settings. Slower pings over the same ground means fewer points, which means slower survey speed to keep density up. Every echosounder ever built behaves this way. Budget for it in your line plan.

The honest caveat: sloped ground and position latency

This one is worth knowing before you buy rather than after.

On a small USV, the sonar does not know where it is. It takes position and heading from the vessel, and on a Blue Robotics BlueBoat that arrives through the ArduPilot EKF. In March 2026, a user testing a Surveyor with an external GNSS reported data that looked good on flat bottoms but showed discrepancies on slopes. Cerulean’s co-founder Larry Lukis confirmed the cause on the Blue Robotics forum: position arriving through the EKF can carry an apparent delay of a couple of hundred milliseconds. On flat ground it is invisible. On a steep slope it shows up as an offset between the up-slope and down-slope passes over the same seabed.

Cerulean added fixed-time-delay compensation in SonarView, and the user who reported it confirmed it corrected data already collected. So the practical advice is narrow: if your sites are dredging basins and harbour floors, this never affects you. If they are quarry walls, channel banks or steep-sided pits, make sure you are on a current SonarView build and check the compensation setting before you mobilise.

Two more things the datasheet will not tell you, both from Cerulean’s own installation guide:

• A BlueBoat has no spare Ethernet port. You need to add a switch. Factor it into the quote rather than discovering it on the dock.
• QGroundControl will override the vehicle’s MAVLink message rates unless you enable “All Streams Controlled By Vehicle Settings”. Cerulean’s documentation is blunt that skipping this step badly degrades data quality. It is a two-minute fix that costs a day of survey if missed.

The question everyone asks next: do I need RTK and a sound velocity profiler?

RTK: effectively yes, for survey work. Your depth accuracy is capped by your position accuracy, and a standard GNSS fix is not good enough for a deliverable. Users running the Surveyor for dredging report an RTK correction stream - an Emlid receiver with the NTRIP extension on BlueOS is a common setup - and describe the resulting data as reliable enough to work alongside a survey-grade system on the same site.

Sound velocity profiler: it depends, and often no. Sound velocity errors bend your beams and skew depths toward the swath edges. In deep or stratified water you need a profile. In shallow fresh water at a roughly constant temperature, one contractor reported getting good results by fitting a temperature sensor to the sonar itself - the Surveyor has a spare penetrator for exactly this - and setting sound velocity manually. SonarView defaults to 1500 m/s; fresh water can sit nearer 1445 m/s, and the difference is not small. Note that many survey contracts require an SVP reading regardless of whether the physics demands it.

In three lines

• Depth along known lines, very shallow water, one operator, minimum training: single-beam. The Sounder S500 at around €780 does this well and mounts on an ROV, AUV or ASV.
• Volumes, surfaces, obstruction searches, change detection over time: multibeam. The Surveyor 240-16 at around €4,375 puts full-coverage bathymetry on a USV a single person can carry.
• Deep water, open coast, or a certified IHO-order deliverable: contract a crewed survey vessel. The compact systems are genuinely good; they are not a substitute for that.

Frequently asked questions

Working out which one you need

We supply both, along with the USVs and ROVs they mount to, so we have no reason to push you toward the more expensive answer. Tell us the water depth, the site, and what the deliverable has to look like, and we will name the system - and say so if the cheaper one covers it.

SepcoTech A/S is an authorised Cerulean Sonar distributor, supplying worldwide in EUR with no hidden import costs and support in your time zone. The hardware, warranty and documentation are Cerulean’s. Getting it specified, cabled and working on your platform is ours.