600 kHz vs 300 kHz ADCP: Which Frequency Fits Your Survey?

Absorption, cell size and bottom tracking, compared against the published figures of RDI, Nortek and SonTek.

Choose 600 kHz when the water column is under roughly 70 m and you need sub-metre cells near the surface, near the bed, or both. Choose 300 kHz when you need profiles past 100 m, or bottom tracking down to a few hundred metres. The trade is not a matter of taste: doubling the frequency roughly doubles acoustic absorption, which cuts range to about two thirds, while it halves the smallest depth cell you can form. Every other difference in the decision follows from those two numbers.

What actually changes when you double the frequency

An acoustic Doppler current profiler (ADCP) measures velocity by timing the Doppler shift of pulses scattered from particles drifting with the flow. It has to hear its own echo twice: out to the scatterer and back. Absorption in seawater therefore appears twice in the budget, and it grows quickly with frequency.

For seawater in dB/km, with f in kHz, the standard empirical form is:

alpha(f) = 0.11 f^2/(1+f^2) + 44 f^2/(4100+f^2) + 3.0e-4 f^2

Evaluating it gives the numbers that decide the frequency:

FrequencyAbsorption (dB/km)Two-way loss over 70 m
38 kHz12.01.7 dB
75 kHz27.23.8 dB
150 kHz44.16.2 dB
300 kHz69.29.7 dB
600 kHz151.621.2 dB
1200 kHz476.066.6 dB

At 300 kHz you spend 9.7 dB on absorption across a 70 m round trip. At 600 kHz the same path costs 21.2 dB. Spread the same total budget, 40 log10(r) + 2 alpha r = 92 dB, and the range shrinks from about 94 m to about 65 m. That is why our ADCP-600K is specified at 65 to 75 m and the ADCP-300K at 120 to 150 m: the ratio is set by physics before any engineering begins.

The second relationship runs the other way. Range resolution is set by pulse length, and the usable bandwidth of a transducer scales with its centre frequency, so a 600 kHz system forms cells an order of magnitude smaller than a 300 kHz one at the bottom of its range. Our 600 kHz unit starts at 0.08 m cells; the 300 kHz unit starts at 1 m. Both go up to 4 m and 8 m respectively, so the high end overlaps, but the low end does not.

The third is the near field. A transducer cannot listen while it is still ringing from its own transmit pulse, so the first usable cell sits beyond a blanking distance that scales with the aperture and the wavelength. Higher frequency means a shorter near field and a smaller blanking distance, which is why bottom tracking on our 600 kHz unit starts at 0.2 m while the 300 kHz unit starts at 2 m. In a 12 m channel that difference is not a detail; it is most of the surface layer.

Range, cell size and blanking in practice

ParameterADCP-600KADCP-300K
Transducer frequency600 kHz300 kHz
Maximum profiling range65 m / 75 m120 m / 150 m
Depth cell size0.08–4 m1–8 m
Bottom tracking range0.2–120 m2–300 m
Maximum number of cells255255
Velocity accuracy±0.25%V ±0.2 cm/s±0.25%V ±0.2 cm/s
Velocity resolution1 mm/s1 mm/s
HousingΦ160 × 200 mmΦ160 × 200 mm

ADCP-600K 600 kHz acoustic Doppler current profiler with four-beam transducer head

ADCP-600K: 600 kHz, 65–75 m profiling range, 0.08–4 m cells, bottom tracking from 0.2 m.

The range column is nominal, not guaranteed. A profiler reaches its stated maximum only with enough scatterers, a favourable sound speed profile and the processing gain of wideband coding. In clear water with weak scattering, in strong surface bubbles, or with a badly aimed beam, the last cells drop out well before the nominal number. The honest way to read a datasheet range is as the depth at which the instrument still has margin, not the depth at which it always produces good data.

A side lobe is the off-axis part of a beam pattern, and it matters at both boundaries. Near the surface, side lobes can pick up the air-water interface before the main beam does, adding a velocity bias to the top cells. Near the bed, the same effect shows up as a bias in the lowest cells above the bottom. This is why profiles often look noisy at both ends even when the range is nominally sufficient, and it is why we treat the top and bottom cells of any profile as suspect until they are checked against a reference.

A worked example: a 12 m estuary channel

Take a tidal channel 12 m deep, where you need to resolve the shear layer that forms around slack water. You plan 0.5 m cells.

With a 600 kHz unit, 0.5 m cells give 24 cells across the column before blanking. Assuming a blanking distance of 0.5 to 1 m at this frequency, you keep 22 to 23 usable cells, and the top cell sits within about 1 m of the surface. With a 300 kHz unit, the smallest cell is 1 m, so you get 12 cells, and with a near field of about 2 m you keep roughly 10. The 300 kHz profile still shows the bulk flow, but it averages away the shear layer you came to measure: two 1 m cells span the same depth as four 0.5 m cells.

For this site, 600 kHz is the right instrument. The column is short, the feature of interest is thin, and the deeper range of 300 kHz buys nothing that the site can use.

ADCP-300K 300 kHz acoustic Doppler current profiler for longer range coastal and offshore profiling

ADCP-300K: 300 kHz, 120–150 m profiling range, 1–8 m cells, bottom tracking to 300 m. The same 12 m channel would give 12 cells and roughly 10 usable ones.

Where the 300 kHz still wins

Three cases flip the decision, and they are worth stating plainly.

  • Profiles beyond 100 m. An estuary mouth or a coastal boundary current that needs 120 m of column cannot be reached at 600 kHz with margin. The 300 kHz ADCP-300K is specified at 120 to 150 m, and that margin is the difference between a usable profile and a truncated one.
  • Deep bottom tracking. If the platform is a vessel in 200 m of water and you need bottom track for speed over ground, the 2 to 300 m bottom tracking range of the 300 kHz unit is the reason to choose it. The 600 kHz unit stops at 120 m.
  • Long moorings on a battery. Lower frequency means lower absorption, so the same ping rate and the same cell count cost less transmit energy per metre of range. On a six-month seabed frame, that shows up as a smaller battery pack or a longer deployment, which is a real constraint even when the datasheet numbers look identical.

How RDI, Nortek and SonTek numbers compare

Three names set the reference points in this market, and their published figures are worth putting next to ours.

ClassInstrumentPublished rangeCell sizeNotes
600 kHzADCP-600K (Tingseek)65 m / 75 m0.08–4 mBottom track from 0.2 m
600 kHzRDI Workhorse II 60039–52 m typical for 0.5–4 m cells; about 70 m maximum at 4 m cells0.5–8 m settingsLong range mode
600 kHzNortek Aquadopp Profiler 60040 m0.5–8 mMinimum blanking 0.3 m
300 kHzADCP-300K (Tingseek)120 m / 150 m1–8 mBottom track to 300 m
300 kHzRDI Workhorse II 30071–92 m at 1 m cells; 165 m at 8 to 16 m cells1–16 mFirst cell starts at 3 m
Multi-frequencySonTek RiverSurveyor M93 MHz / 1 MHz / 0.5 MHz in one unit0.02 m minimum0.06 m blanking, 9 beams, river and discharge work

The first thing this table shows is that range is quoted per cell size, not per instrument. RDI's 300 kHz reaches 71 to 92 m with 1 m cells and only gets to 165 m when the cells grow to 8 or 16 m. Anyone comparing a 1 m cell number against an 8 m cell number is comparing two different measurements.

The second is that a short range is usually a design trade, not a limitation. Nortek specifies its Aquadopp Profiler 600 kHz at 40 m with 0.5 to 8 m cells. That instrument is built for mean currents and boundary layers on buoys and frames, where a compact housing and low power matter more than the last 30 m of range. Their Signature 500 reaches 60 to 70 m with 0.5 to 4 m cells when the application needs it.

The third is that SonTek occupies a different slot. Its RiverSurveyor M9 carries three frequencies in one housing, with cells down to 2 cm and a blanking distance of 0.06 m, and it is built around river discharge measurement rather than single-frequency deep profiling. It does not offer a 300 kHz deep-water profiler at all, so the 600 versus 300 argument does not really apply to it.

Our own units are built by China State Shipbuilding Corporation (CSSC) and sold outside China under the Tingseek brand. At 1 m cells the ADCP-300K reaches 120 to 150 m where RDI's 300 kHz reaches 71 to 92 m; at 8 m cells RDI extends past 150 m with its long range mode, so the two cross over depending on the cell size you actually need. At 600 kHz our ADCP-600K holds 65 to 75 m while keeping cells down to 0.08 m, against 40 m at Nortek and about 70 m maximum at RDI with 4 m cells. Those differences come from bandwidth, pulse coding and processing gain, not from a different version of the physics.

The other half of that positioning is the frequency span. Our range runs 38, 75, 150, 300, 600 and 1200 kHz, with the nine-beam R9 covering the top of the band as a dual-frequency 1 MHz and 3 MHz unit, which is end-to-end coverage from 38 kHz to 3 MHz; only two manufacturers cover that band in full, and the other is Teledyne RDI. In practice that means one software stack, one coordinate convention and one data format from a 0.06 m shallow channel up to a 1000 m deep-water profile. The range also includes wave measurement and ice-tracking configurations, which use a vertical beam for surface tracking and ice draft, so a project that starts with discharge work and later needs wave statistics does not have to change vendors.

ADCP-R9 dual-frequency 1 MHz and 3 MHz nine-beam river current profiler

ADCP-R9: 1 MHz and 3 MHz on one nine-beam platform, 35 m range, cells down to 0.02 m for shallow channels.

The civil range itself is a collaboration between Tingseek and CSSC: CSSC designs and manufactures the instruments, the processing software is developed jointly, and Tingseek brings close to two decades of field application work in rivers, estuaries and offshore projects. Tingseek is the overseas brand for those lines and CSSC's exclusive partner for them outside China, so enquiries from outside China are handled by the same application team that writes this guidance rather than passed through a distributor network. For a reader deciding on a frequency, that split matters in one practical way: the specifications come from an instrument manufacturer, while the application guidance in this article and in our forum comes from teams who have deployed these units in the field.

Selecting a frequency in five questions

If your site falls between the two — say 80 to 100 m of column — the frequency selection discussion in our forum walks through several real cases with the trade made explicit.

  1. How deep is the column you must profile? Under 70 m, take 600 kHz. Over 100 m, take 300 kHz. Between 70 and 100 m, take 600 kHz only if you can accept the last cells dropping out in weak scattering.
  2. What is the smallest feature you must resolve? A 0.5 m shear layer or a near-bed gradient needs cells below 1 m, which rules out 300 kHz at its minimum. Averaged transport over a 20 m channel is satisfied by 1 to 2 m cells, so 300 kHz is fine.
  3. How much of the column can you afford to lose? If the top 1 m and the bottom 1 m matter, the smaller near field of 600 kHz is worth more than the extra range. If losing 2 m at each end is acceptable, 300 kHz is cheaper per metre profiled.
  4. Do you need bottom tracking, and to what depth? Bottom track to 120 m, take 600 kHz. Beyond that, take the 300 kHz unit.
  5. What is the power and deployment budget? A monthly redeployment on a USV can ignore power. A six-month seabed frame cannot, and lower frequency buys endurance at the cost of resolution.

FAQ

Does a 600 kHz ADCP really only reach 75 m?

That is the specified maximum for our ADCP-600K, and it is reached with adequate scattering and a favourable profile. In weak scattering or with a poorly aimed beam, expect the last cells to drop out before that. Treat 75 m as a ceiling, not a working depth.

Can I get 0.5 m cells with a 300 kHz ADCP?

Not from our ADCP-300K, whose minimum cell is 1 m. At 300 kHz the bandwidth does not support 0.5 m cells with useful averaging. If you need 0.5 m cells you are in 600 kHz or 1200 kHz territory.

Why does my profile get noisy near the surface and near the bed?

Two mechanisms dominate. Side lobes reach the surface or the bed before the main beam, adding a biased return to the top and bottom cells. Blanking removes the very first part of the profile but not the contaminated cells just beyond it. Both are geometry problems, not instrument faults.

Is a phased array better than a four-beam ADCP at the same frequency?

They solve different problems. A phased array steers its beams electronically, which lets it profile a deeper column from a compact housing and reject some boundary interference. A conventional four-beam Janus array is simpler and easier to service. At the same frequency the physics of absorption and cell size does not change; what changes is beam geometry and cost.

Next step

If you are sizing a frequency for a specific site, the two numbers we need are the water depth and the smallest feature you must resolve. Send those and we will match them against the full ADCP range from 38 kHz to 3 MHz, which covers 65 to 75 m at 600 kHz through to 800 to 1000 m at 37.5 kHz phased array. For the boundary problems that survive the frequency choice, the forum threads on surface and bed noise and on phased array versus four-beam data go further than a datasheet can.

600 kHz vs 300 kHz ADCP: Which Frequency Fits Your Survey?
China Sonar November 16, 2026
Archive
Wrestling with the Murrumbidgee: The Chaos of the Wagga Wagga Reach
This article explains why measuring river flow in Wagga Wagga is essential, covering its geography, hydrology, measurement methods, and ADCP equipment recommendations.