Converting Surface Velocity to Mean Velocity in Radar Flow Measurement

A radar surface-velocity sensor measures motion in a limited area near the water surface. Discharge calculations require cross-sectional mean velocity. The conversion coefficient is not a universal constant: it changes with channel shape, depth, roughness, bends, wind, backwater and the sensor's sampled region. A reliable system uses site reference measurements to establish and maintain that relationship.

Understand what the radar samples

Document beam footprint, incidence angle, mounting height and the location of the velocity measurement relative to the channel. The strongest returned motion may come from ripples, floating material or rain rather than the representative current. Confirm minimum signal and surface-roughness requirements for the selected instrument.

Keep the footprint away from eddies, wall effects, hydraulic jumps, gates and falling water. A stable mounting angle is essential because geometry changes affect the derived velocity.

Measure the cross-sectional reference

Use a recognized reference method appropriate to channel size and flow, such as a current-meter traverse, acoustic Doppler measurement or another accepted procedure. Measure enough verticals and depths to calculate cross-sectional mean velocity, not just one convenient point.

Synchronize the reference measurement with radar velocity and water level. Rapidly changing flow can make readings taken several minutes apart incomparable. Record gate, pump, wind, rainfall and backwater conditions.

Develop the coefficient

For each valid test, divide reference mean velocity by the simultaneous radar surface velocity. Evaluate the coefficient across the operating depth and velocity range. A single factor may be adequate in a stable uniform channel; other sites need a depth-dependent curve or separate regimes.

Do not force a coefficient through poor data. Exclude tests only using documented quality criteria such as insufficient radar return, non-uniform reference traverse or rapidly changing stage—not because the result is inconvenient.

Recognize changing flow profiles

Bed roughness, vegetation, sediment and ice alter the vertical velocity profile. Bends create lateral asymmetry, while downstream control can change flow distribution at the same depth. Wind affects the surface most strongly and may bias low velocities.

Define validity limits for direction, minimum velocity, depth, backwater and environmental conditions. When limits are exceeded, transmit a quality flag or estimated status instead of presenting normal precision.

Validate discharge end to end

Mean velocity must be combined with cross-sectional area from the same hydraulic section and timestamp. Verify sensor velocity, coefficient selection, water level, geometry, area, units and final discharge through the complete data path.

Repeat reference measurements after sensor relocation, channel work or persistent reconciliation drift. Track coefficient versions and effective dates so historical results can be reproduced.

Worked calibration example

At a straight lined channel, six reference traverses show coefficients clustered near 0.84 at medium and high depth but lower values at very shallow flow, where wall and bed effects dominate. The project uses a depth-dependent rule and declares results below the tested depth as estimated.

A later sediment survey changes the area model but not automatically the velocity coefficient. Both components are rechecked separately before the revised discharge calculation is accepted.

Engineering checklist

• Map the radar velocity footprint.
• Use synchronized cross-sectional reference measurements.
• Test coefficients across depth and velocity range.
• Define wind, backwater and low-signal limits.
• Combine velocity and area at matching timestamps.
• Version coefficients and repeat calibration after change.

Frequently asked questions

Is 0.85 always the correct coefficient?

No. It is sometimes used as an initial assumption, but the accepted coefficient should come from site evidence.

Can one reference test calibrate the full range?

No. Flow profile and coefficient can change with depth and hydraulic condition.

What if the surface is nearly smooth?

Radar return may be weak; the system should flag insufficient signal rather than report unsupported velocity.

Related measurement solutions

Radar open-channel flow meters · Radar water level meters