RTK for Bathymetric Surveys: Tides, Datums, and the Base on the Bank

How RTK transforms shallow-water bathymetry: RTK tides replacing gauges, ellipsoid-referenced surveying with USVs and echo sounders, chart datum transformations, and one base serving the boat, the drone, and the shore crew.

The vertical problem that defines hydrography

Bathymetry is a vertical discipline wearing a horizontal costume: the deliverable is depth below a defined water datum, and the survey’s hardest errors live in knowing where the water — and the boat on it — stands at every instant. Classically that meant tide gauges, staff readings, and squat/settlement models stacked on top of positioning. RTK collapsed much of the stack: with centimetre 3D positioning on the sounder’s antenna, the water surface becomes something you measure continuously rather than model — the technique the industry calls RTK tides, and the reason a GNSS base on the bank is now core hydrographic kit.

RTK tides and ellipsoid-referenced surveying

The modern chain: the echo sounder measures antenna-to-seabed (via draft and sound velocity); RTK measures the antenna’s ellipsoidal height to centimetres; subtraction yields seabed ellipsoidal heights directly — ellipsoid-referenced surveying (ERS) — with heave largely observable in the RTK trace and tide gauges demoted to verification. The final products transform once to the required vertical datum: chart datum / LAT for navigation products, orthometric for engineering, using the separation models hydrographic authorities publish. The discipline’s golden rule matches land surveying’s: transform at export, document the model and epoch, never nudge (the datum bookkeeping).

Small craft changed the client list: USVs and the drone pairing

Uncrewed surface vessels — sub-2 m catamarans with single-beam or compact multibeam echo sounders and an RTK receiver on the mast — brought bathymetry to ponds, tailings dams, intake channels, and rivers that crewed launches never economically touched. The natural pairing on such jobs: the USV maps the wet part while a drone maps the dry part — banks, structures, stockpiles at the water’s edge — both consuming the same base’s RTCM stream, so the terrestrial surface and the seabed merge into one seamless topo-bathymetric model without a registration step. Mining water balances, dredge progress, and reservoir capacity work consume exactly this merged product (the volumetrics discipline transfers directly).

The base on the bank: setup realities

Waterside sites amplify the base-coordinates question: known monuments are rare on riverbanks and reservoir rims, cellular fades in gorges, and the vertical component — which bathymetry lives on — is unforgiving of averaged setups. A self-converging PPP base answers the profile: absolute ~1.5 cm horizontal / 3 cm vertical in about three minutes on any open bank, RTCM over UHF to the USV and drone (water gives generous radio horizons), RINEX logging for PPK insurance on legs behind bluffs (the methods compared). One practical waterside habit: place the base above any plausible splash and wake line — IP68 forgives, but antennas prefer dry sky.

Verification with wet checkpoints

Hydrography’s QA analogue to checkpoints: bar checks and reference surfaces. Sound a repeat line at each session’s start and compare; where possible, sound over a hard reference (a boat ramp toe, a lock sill of surveyed elevation) and reconcile against the rover-measured value from the same base — tying the wet and dry measurement chains together through one datum. Crossline analysis — flying survey lines against perpendicular check lines and statistically comparing intersections — remains the internal-consistency standard. File the numbers as you would land checkpoints; dredging disputes are settled by exactly this paper.

Where the limits sit

RTK tides need the link: beyond radio/NTRIP reach, workflows fall back to PPK-processed trajectories (the base’s RINEX again) or traditional tide modelling. Sound velocity remains bathymetry’s own atmosphere — profile it; no GNSS fixes a wrong SV cast. Vegetation-choked and turbid margins defeat acoustics and photogrammetry alike — topo-bathy LiDAR or pole soundings bridge the gap. And multipath off the water is real for the shore rover — checkpoint on set-back hard stands. None of these dent the core gain: the vertical control problem that once dominated shallow-water surveying has become, with an absolute base on the bank, a three-minute setup step.

A pond-to-port scaling note

The same architecture scales with dignity: a farm dam capacity survey (USV + drone + one base, half a day), a dredge-progress program on a river terminal (weekly epochs, volumes by subtraction on the shared datum), and coastal engineering baselines (ERS throughout, chart-datum products at export). What changes is sensor cost and mobilization; what does not change is the shape of the solution — one absolute correction source on the bank, every wet and dry sensor drinking from it, one frame from seabed to stockpile (the architecture underneath).

Kit notes for the drone-shop entering hydro

Mapping firms adding water work discover most of their kit transfers. The base, the rover, the datum discipline, and the checkpoint habit move over unchanged; the genuinely new purchases are the USV with its sounder, a sound-velocity probe, and the hydrographic processing seat. Two habits to import on day one: profile sound velocity at every session (the wet atmosphere), and keep the antenna-to-transducer offsets in a signed configuration file — the wet equivalent of antenna height, with the same bust potential (the dry version of the lesson). Firms report the learning curve is shorter than feared precisely because the positioning half — usually the hard half — arrives already solved.

One-line takeaway

Put an absolute base on the bank and bathymetry inherits everything land surveying learned: RTK tides replace gauges, wet and dry data share one frame, and the vertical problem that defined the discipline becomes a three-minute setup step.

Choosing the first sounder: a positioning-person's guide

For mapping firms crossing over, the sounder decision parallels familiar territory. Single-beam echo sounders are the Mavic of the water — affordable, simple, ideal for ponds, dams, and profile lines. Compact multibeams are the LiDAR — full-coverage swaths, richer QA (crosslines mean something), and a processing seat to match. The positioning requirement is identical for both: centimetre RTK at the antenna from the bank base, offsets documented, sound velocity profiled. Start single-beam on capacity and dredge-progress work where clients accept profiles; graduate to multibeam when coverage specifications — not curiosity — demand it. The base you already own serves both without change.

One-line takeaway

RTK moved hydrography's hardest problem onto dry land: converge an absolute base on the bank and the tides, the datum, and the wet-dry seam all reduce to one broadcast every sensor on site can drink.

Frequently asked questions

What is an RTK tide?

Using the vessel antenna’s centimetre RTK height to measure the water surface continuously, replacing much of traditional tide-gauge control in shallow-water surveys.

Can a drone base station serve a bathymetric survey?

Yes — the USV’s receiver consumes the same RTCM 3.x stream as the drone, putting seabed and terrestrial data on one datum from one base.

What vertical datum do bathymetric deliverables use?

Navigation products use chart datum/LAT; engineering usually orthometric. Survey in ellipsoidal heights and transform at export with the published separation models.

How accurate is USV bathymetry with RTK?

Position centimetres from RTK; depth accuracy then hinges on sound-velocity profiling and sounder quality — well-run shallow surveys deliver decimetre-or-better total vertical uncertainty.

What happens beyond radio range on big water?

Log raw data and post-process the trajectory (PPK) against the base’s RINEX, or fall back to tide modelling — the base’s logs keep every option open.