The Chaos of the Drammen-Fjord Interface
October in the Drammensfjord is a brutal time to be on the water. The wind cuts through your gear, and the water turns a murky, sediment-laden brown as the autumn rains push the Drammen river’s discharge into overdrive. I remember pushing off from the quay, watching that violent intersection where the river’s freshwater plume slams into the denser saltwater pushing in from the Oslofjord. It isn't a gentle mixing; it's a battleground of densities.
Most practitioners treat estuarine waters as well-mixed. That’s a rookie mistake. In the Drammensfjord, if you assume a homogeneous water column, you’ve ruined your data before the sensor even hits the seabed. This system is a textbook example of high stratification. We are dealing with a narrow corridor where the bathymetry is erratic—deep pockets that suddenly give way to shallow banks. This creates a funnel effect, accelerating flow and kicking up suspended solids that make visibility virtually zero.
The Vertical Velocity Shear Problem
When the data from our October 2023 deployment came back, it confirmed my suspicions: the vertical velocity shear is staggering. We observed surface currents sprinting downstream toward the fjord, while the bottom flows were moving inland. This is the classic salt wedge in action, but the intensity here is surprising. The pycnocline—the razor-thin boundary separating the fresh surface layer from the saline depths—is incredibly aggressive.
In some of our acoustic bins, the velocity vector flipped 180 degrees over a vertical distance of less than two meters. If you aren't sampling at high frequencies, you're just guessing. You miss the shear, you miss the transport volume, and you end up with a model that doesn't reflect reality. The sheer volatility of the interface means that any averaging over long time steps effectively erases the most critical physics of the system.
The Sound Speed Nightmare
The real headache in Drammensfjord isn't just the current; it's the acoustics. Because we are operating in a highly stratified environment, the speed of sound is a moving target. Sound speed depends on temperature, pressure, and salinity. In a salt wedge, salinity gradients are extreme. A static sound-speed setting is useless here; it's a recipe for distance errors in your ADCP pings.
During the autumn rains, the river discharge pushes the salt wedge significantly further downstream than historical averages suggest. As the pycnocline shifts, the refraction of the acoustic signal changes. If you don't account for the varying sound velocity profile (SVP) in real-time, your depth bins shift. You think you're measuring the bottom flow at 10 meters, but you're actually looking at 11 or 12. In a narrow channel like this, a one-meter error in bin depth can lead to a massive miscalculation of total discharge.
Dealing with Suspended Solids
Drammensfjord is notorious for its sediment load. When the river is high, the water is thick with suspended solids. For an acoustic Doppler current profiler, this is a double-edged sword. You need backscatter to get a reading, but too much sediment—or the wrong kind of organic matter—can attenuate the signal before it ever returns to the transducer.
We found that the 'noise' in the data wasn't just electronic; it was physical. The turbulence at the salt wedge interface creates micro-bubbles and sediment plumes that scatter the signal. I've seen colleagues struggle with 'blanking distance' issues here, where the first few meters of data are garbage because the surface turbulence is so chaotic. You have to tune your sampling strategy to ignore the noise without losing the critical data at the interface.
Tidal Influence and the 'Plug' Effect
The tidal range in the Drammen area is modest, but its impact on the salt wedge is disproportionate. During the flood tide, the denser saltwater pushes inland, wedging itself under the freshwater. This creates a 'plug' effect. The saltwater doesn't just slide in; it displaces the freshwater, forcing a surge of surface water downstream.
If you look at the coordinates around the narrowest points of the fjord, you can see where the flow accelerates. The interaction between the tidal oscillation and the river's discharge creates a complex harmonic. It's not a simple sine wave. You get skewed tidal curves and residual currents that linger long after the tide has turned. This is where the real physics happens, and it's where most off-the-shelf models fail because they can't handle the non-linear interaction of the wedge.
Practical Lessons from the Field
If you're planning a deployment in these waters, stop relying on historical salinity maps. They are outdated the moment the first autumn storm hits. You need concurrent CTD (Conductivity, Temperature, Depth) casts to calibrate your acoustic data. Without a real-time SVP, your velocity profiles are essentially educated guesses.
I also suggest paying close attention to the mooring tension. The currents in the Drammensfjord can be deceptively strong at the bottom during a flood tide. If your mooring isn't weighted correctly, the drag from the salt wedge will tilt your sensor, introducing a cosine error into your horizontal velocity components. A 5-degree tilt might not seem like much, but when you're trying to quantify precise mass transport in a narrow estuary, it's enough to throw the whole budget off.
Ultimately, the Drammensfjord is a reminder that the ocean isn't a uniform body of water. It's a series of layers, each with its own agenda. To quantify these currents, you have to stop treating the water as a single mass and start treating it as a stack of competing forces.
Dr. Alistair Vance, estuarine dynamics and salt wedge modeling. With over 20 years of field experience, Dr. Vance specializes in high-resolution acoustic profiling of stratified coastal waters.
Wrestling with the Salt Wedge of Drammensfjord